Pure liquids are those which are chemically pure and do not contain any other
impurity even in traces of 1 in 109, and are structurally simple. Examples of such
simple pure liquids are /x-hexane (CgH^), n-heptane (C/Hj^) and other paraffin
hydrocarbons. By using simple and pure liquids, it is easier to separate out the various
factors that influence conduction and breakdown in them. On the other hand, the
commercial liquids which are insulating liquids like oils which are not chemically
pure, normally consist of mixtures of complex organic molecules which cannot be
easily specified or reproduced in a series of experiments
Purification
The main impurities in liquid dielectrics are dust, moisture, dissolved gases and ionic impurities. Various methods employed for purification are filtration (through
mechanical filters, spray filters, and electrostatic filters), centrifuging, degassing and distillation, and chemical treatment (adding ion exchange materials such as alumina, fuller's earth, etc. and filtering). Dust particles when present become charged and reduce the breakdown strength of the liquid dielectrics, and they can be removed by careful filtration. Liquid will normally contain moisture and dissolved gases in small quantities. Gases like oxygen and carbon dioxide significantly affect the breakdown strength of the liquids, and hence it is necessary to control the amount of gas present.
This is done by distillation and degassing. Ionic impurity in liquids, like water vapour which easily dissociates, leads to very high conductivity and heating of the liquid depending on the applied electric field. Water is removed using drying agents or by vacuum drying. Sometimes, liquids are shaken with concentrated sulphuric acid to remove wax and residue and washed with caustic soda and distilled water. A commonly used closed-cycle liquid purification system to prepare liquids as per the above requirements is shown in Fig. 3.1. This system provides for cycling the liquid.
The liquid from the reservoir flows through the distillation column where ionic impurities are removed. Water is removed by drying agents or frozen out in the low-temperature bath. The gases dissolved in the liquid are removed by passing them through the cooling tower and/or pumped out by the vacuum pumps. The liquid then passes through the filter where dust particles are removed. The liquid thus purified is thenused in the test cell. The used liquid then flows back into the reservior. The vacuum system thus helps to remove the moisture and other gaseous impurities.
Saturday, September 26, 2009
Transformer Oil
As already mentioned, transformer oil is the most commonly used liquid dielectric in
power apparatus. It is an almost colourless liquid consisting a mixture of hydrocarbons
which include paraffins, iso-paraffins, naphthalenes and aromatics. When in
service, the liquid in a transformer is subjected to prolonged heating at high temperatures
of about 950C, and consequently it undergoes a gradual ageing process. With
time the oil becomes darker due to the formation of acids and resins, or sludge in the
liquid. Some of the acids are corrosive to the solid insulating materials and metal parts
in the transformer. Deposits of sludge on the transformer core, on the coils and inside
the oil ducts reduce circulation of oil and thus its heat transfer capability gets
considerably reduced. Complete specifications for the testing of transformer oils are
given in IS 1866 (1983), IEC 296 (1969) and IEC 474 (1974).
Electrical Properties
The electrical properties that are essential in determining the dielectric performance
of a liquid dielectric are
(a) its capacitance per unit volume or its relative permittivity
(b) its resistivity
(c) its loss tangent (tan 5) or its power factor which is an indication of the power
loss under a.c. voltage application
(d) its ability to withstand high electric stresses.
Permittivities of most of the petroleum oils vary from 2.0 to 2.6 while those of
askerels vary between 4.5 and 5.0 and those of silicone oils from 2.0 to 73 (see Table
3.1). In case of the non-polar liquids, the permittivity is independent of frequency but in the case of polar liquids, such as water, it changes with frequency. For example, the permittivity of water is 78 at 50 Hz and reduces to about 5.0 at 1 MHz.
Resistivities of insulating liquids used for high voltage applications should be more
than 1016 ohm-metre and most of the liquids in their pure state exhibit this property.
Power Factor of a liquid dielectric under a.c. voltage will determine its performance
under load conditions. Power factor is a measure of the power loss and is an
important parameter in cable and capacitor systems. However, in the case of transformers, the dielectric loss in the oil is negligible when compared to copper and iron losses. Pure and dry transformer oil will have a very low power factor varying between 1(T4 at 2O0C and 10~3 at 9O0C at a frequency of 50 Hz.
Dielectric Strength is the most important parameter in the choice of a given liquid
dielectric for a given application. The dielectric strength depends on the atomic and
molecular properties of the liquid itself. However, under practical conditions the
dielectric strength depends on the material of the electrodes, temperature, type of
applied voltage, gas content in the liquid etc., which change the dielectric strength by changing the molecular properties of the liquid. The above factors which control the breakdown strength and lead to electrical breakdown of the liquid dielectrics are discussed in subsequent sections.
power apparatus. It is an almost colourless liquid consisting a mixture of hydrocarbons
which include paraffins, iso-paraffins, naphthalenes and aromatics. When in
service, the liquid in a transformer is subjected to prolonged heating at high temperatures
of about 950C, and consequently it undergoes a gradual ageing process. With
time the oil becomes darker due to the formation of acids and resins, or sludge in the
liquid. Some of the acids are corrosive to the solid insulating materials and metal parts
in the transformer. Deposits of sludge on the transformer core, on the coils and inside
the oil ducts reduce circulation of oil and thus its heat transfer capability gets
considerably reduced. Complete specifications for the testing of transformer oils are
given in IS 1866 (1983), IEC 296 (1969) and IEC 474 (1974).
Electrical Properties
The electrical properties that are essential in determining the dielectric performance
of a liquid dielectric are
(a) its capacitance per unit volume or its relative permittivity
(b) its resistivity
(c) its loss tangent (tan 5) or its power factor which is an indication of the power
loss under a.c. voltage application
(d) its ability to withstand high electric stresses.
Permittivities of most of the petroleum oils vary from 2.0 to 2.6 while those of
askerels vary between 4.5 and 5.0 and those of silicone oils from 2.0 to 73 (see Table
3.1). In case of the non-polar liquids, the permittivity is independent of frequency but in the case of polar liquids, such as water, it changes with frequency. For example, the permittivity of water is 78 at 50 Hz and reduces to about 5.0 at 1 MHz.
Resistivities of insulating liquids used for high voltage applications should be more
than 1016 ohm-metre and most of the liquids in their pure state exhibit this property.
Power Factor of a liquid dielectric under a.c. voltage will determine its performance
under load conditions. Power factor is a measure of the power loss and is an
important parameter in cable and capacitor systems. However, in the case of transformers, the dielectric loss in the oil is negligible when compared to copper and iron losses. Pure and dry transformer oil will have a very low power factor varying between 1(T4 at 2O0C and 10~3 at 9O0C at a frequency of 50 Hz.
Dielectric Strength is the most important parameter in the choice of a given liquid
dielectric for a given application. The dielectric strength depends on the atomic and
molecular properties of the liquid itself. However, under practical conditions the
dielectric strength depends on the material of the electrodes, temperature, type of
applied voltage, gas content in the liquid etc., which change the dielectric strength by changing the molecular properties of the liquid. The above factors which control the breakdown strength and lead to electrical breakdown of the liquid dielectrics are discussed in subsequent sections.
LIQUIDS AS INSULATORS
Liquid dielectrics, because of their inherent properties, appear as though they would
be more useful as insulating materials than either solids or gases. This is because both
liquids and solids are usually 103 times denser than gases and hence, from Paschen's
law it should follow that they possess much higher dielectric strength of the order of
107 V/cm. Also, liquids, like gases, fill the complete volume to be insulated and
simultaneously will dissipate heat by convection. Oil is about 10 times more efficient
than air or nitrogen in its heat transfer capability when used in transformers. Although
liquids are expected to give very high dielectric strength of the order of 10 MV/cm,
in actual practice the strengths obtained are only of the order of 100 kV/cm.
Liquid dielectrics are used mainly as impregnants in high voltage cables and
capacitors, and for filling up of transformers, circuit breakers etc. Liquid dielectrics
also act as heat transfer agents in transformers and as arc quenching media in circuit
breakers. Petroleum oils (Transformer oil) are the most commonly used liquid
dielectrics. Synthetic hydrocarbons and halogenated hydrocarbons are also used for
certain applications. For very high temperature application, silicone oils and
fluorinated hydrocarbons are also employed. In recent times, certain vegetable oils
and esters are also being tried. However, it may be mentioned that some of the isomers
of poly-chlorinated diphenyls (generally called askerels) have been found to be very
toxic and poisonous, and hence, their use has been almost stopped. In recent years, a
synthetic ester fluid with the trade name 'Midel' has been developed as a replacement
for askerels.
Liquid dielectrics normally are mixtures of hydrocarbons and are weakly
polarised. When used for electrical insulation purposes they should be free from
moisture, products of oxidation and other contaminants. The most important factor
that affects the electrical strength of an insulating oil is the presence of water in the
form of fine droplets suspended in the oil. The presence of even 0.01% water in
transformer oil reduces its electrical strength to 20% of the dry oil value. The
dielectric strength of oil reduces more sharply, if it contains fibrous impurities in
addition to water.
be more useful as insulating materials than either solids or gases. This is because both
liquids and solids are usually 103 times denser than gases and hence, from Paschen's
law it should follow that they possess much higher dielectric strength of the order of
107 V/cm. Also, liquids, like gases, fill the complete volume to be insulated and
simultaneously will dissipate heat by convection. Oil is about 10 times more efficient
than air or nitrogen in its heat transfer capability when used in transformers. Although
liquids are expected to give very high dielectric strength of the order of 10 MV/cm,
in actual practice the strengths obtained are only of the order of 100 kV/cm.
Liquid dielectrics are used mainly as impregnants in high voltage cables and
capacitors, and for filling up of transformers, circuit breakers etc. Liquid dielectrics
also act as heat transfer agents in transformers and as arc quenching media in circuit
breakers. Petroleum oils (Transformer oil) are the most commonly used liquid
dielectrics. Synthetic hydrocarbons and halogenated hydrocarbons are also used for
certain applications. For very high temperature application, silicone oils and
fluorinated hydrocarbons are also employed. In recent times, certain vegetable oils
and esters are also being tried. However, it may be mentioned that some of the isomers
of poly-chlorinated diphenyls (generally called askerels) have been found to be very
toxic and poisonous, and hence, their use has been almost stopped. In recent years, a
synthetic ester fluid with the trade name 'Midel' has been developed as a replacement
for askerels.
Liquid dielectrics normally are mixtures of hydrocarbons and are weakly
polarised. When used for electrical insulation purposes they should be free from
moisture, products of oxidation and other contaminants. The most important factor
that affects the electrical strength of an insulating oil is the presence of water in the
form of fine droplets suspended in the oil. The presence of even 0.01% water in
transformer oil reduces its electrical strength to 20% of the dry oil value. The
dielectric strength of oil reduces more sharply, if it contains fibrous impurities in
addition to water.
VACUUM INSULATION
Introduction
The idea of using vacuum for insulation purposes is very old. According to the
Townsend theory, the growth of current in a gap depends on the drift of the charged
particles. In the absence of any such particles, as in the case of perfect vacuum, there
should be no conduction and the vacuum should be a perfect insulating medium.
However, in practice, the presence of metallic electrodes and insulating surfaces
within the vacuum complicate the issue and, therefore, even in vacuum, a sufficiently
high voltage will cause a breakdown.
In recent years a considerable amount of work has been done to determine the
electrical properties of high vacuum. This is mainly aimed at adopting such a medium
for a wide range of applications in devices such as vacuum contractors and interrupters,
high frequency capacitors and relays, electrostatic generators, microwave tubes,
etc. The contractors and circuit breakers using vacuum as insulation are finding
increasing applications in power systems.
What Is Vacuum?
A vacuum system which is used to create vacuum is a system in which the pressure is
maintained at a value much below the atmospheric pressure. In vacuum systems the
pressure is always measured in terms of millimetres of mercury, where one standard
atmosphere is equal to 760 millimetres of mercury at a temperature OfO0C. The term
*'millimetres of mercury" has been standardised as "Torr" by the International
Vacuum Society, where one millimetre of mercury is taken as equal to one Torn
Vacuum may be classified as
High vacuum : 1 x 1(T3 to 1 x IGT6 Ton-
Very high vacuum : 1 x IQT6 to 1 x 1(T8 Ton-
Ultra high vacuum : 1 x 10~9 torr and below.
For electrical insulation purposes, the range of vacuum generally used is the* 'high
vacuum", in the pressure range of 10~3 Torr to 1(T6 Torr.
Vacuum Breakdown
In the Townsend type of discharge in a gas described earlier, electrons get multiplied due to various ionisalion processes and an electron avalanche is formed. In a high vacuum, even if the electrodes are separated by, say, a few centimetres, an electron crosses the gap without encountering any collisions. Therefore, the current growth prior to breakdown cannot be due to the formation of electron avalanches. However, if a gas is liberated in the vacuum gap, then, breakdown can occur in the manner described by the Townsend process. Thus, the various breakdown mechanisms in high vacuum aim at establishing the way in which the liberation of gas can be brought about in a vacuum gap.
During the last 70 years or so, many different mechanisms for breakdown in
vacuum have been proposed. These can be broadly divided into three categories
(a) Particle exchange mechanism
(ft) Field emission mechanism
(c) Clump theory
(a) Particle Exchange Mechanism
In this mechanism it is assumed that a charged particle would be emitted from one
electrode under the action of the high electric field, and when it impinges on the other electrode, it liberates oppositely charged particles. These particles are accelerated by the applied voltage back to the first electrode where they release more of the original type of particles. When this process becomes cumulative, a chain reaction occurs which leads to the breakdown of the gap.
The particle-exchange mechanism involves electrons, positive ions, photons and
the absorbed gases at the electrode surfaces. Qualitatively, an electron present in the vacuum gap is accelerated towards the anode, and on impact releases A positive ions and C photons. These positive ions are accelerated towards the cathode, and on impact
each positive ion liberates B electrons and each photon liberates D electrons. This is
shown schematically in Fig. 2.24. The breakdown will occur if the coefficients of
production of secondary electrons exceeds unity. Mathematically, the condition for
breakdown can be written as
(AB + CD) > I (2.32)
Later, Trump and Van de Graaff measured these coefficients and showed that
they were too small for this process to take place. Accordingly, this theory was
modifie4 to allow for the presence of negative ions and the criterion for breakdown
then becomes
(AB + EF) >l (2.33)
Where A and B are the same as before and E and F represent the coefficients for
negative and positive ion liberation by positive and negative ions. It was experimentally
found that the values of the product EF were close enough to unity for copper,
aluminium and stainless steel electrodes to make this mechanism applicable at
voltages above 250 kV.
(b) Field Emission Theory
(O Anode Heating Mechanism
This theory postulates that electrons produced at small micro-projections on the
cathode due to field emission bombard the anode causing a local rise in temperature
and release gases and vapours into the vacuum gap. These electrons ionise the atoms
of the gas and produce positive ions. These positive ions arrive at the cathode, increase the primary electron emission due to space charge formation and produce secondary electrons by bombarding the surface. The process continues until a sufficient number of electrons are produced to give rise to breakdown, as in the case of a low pressure Townsend type gas discharge.
Cathode Heating Mechanism
This mechanism postulates that near the breakdown voltages of the gap, sharp
points on the cathode surface are responsible for the existence of the pre-breakdown
current, which is generated according to the field emission process described below.
This current causes resistive heating at the tip of a point and when a critical current density is reached, the tip melts and explodes, thus initiating vacuum discharge. This mechanism is called field emission as shown schematically in Fig. 2.26. Thus, the initiation of breakdown depends on the conditions and the properties of the cathode surface. Experimental evidence shows that breakdown takes place by this process when the effective catliode electric field is of the order of 106 to 107V/cm.
Clump Mechanism
Basically this theory has been developed on the following assumptions (Fig. 2.27):
(O A loosely bound particle (clump) exists on one of the electrode surfaces.
(11) On the application of a high voltage, this particle gets charged, subsequently
gets detached from the mother electrode, and is accelerated across the gap.
(///) The breakdown occurs due to a discharge in the vapour or gas released by the
impact of the particle at the target electrode.
Cranberg was the first to propose this theory. He initially assumed that breakdown
will occur when the energy per unit area, W9 delivered to the target electrode by a
clump exceeds a value C, a constant, characteristic of a given pair of electrodes. The quantity W is the product of gap voltage (V) and the charge density on the clump. The latter is proportional to the electric field E at the electrode of origin. The criterion for breakdown, therefore, is
VE = C
In case of parallel plane electrodes the field E = VId9 where d is the distance
between the electrodes. So the generalised criterion for breakdown becomes
V=(Cd)^ (2.35)
where C is another constant involving C and the electrode surface conditions.
Cranberg presented a summary of the experimental results which satisfied this
breakdown criterion with reasonable accuracy. He stated that the origin of the clump
was the cathode and obtained a value for the constant C as 60 x 101^ V2/cm (for iron
particles). However the equation was later modified as V = C d01, where a varies
between 0.2 and 1.2 depending on the gap length and the electrode material, with a
maximum at 0.6. The dependence of V on the electrode material, comes from the
observations of markings on the electrode surfaces. Craters were observed on the
anode and melted regions on the cathode or vice-versa after a single breakdown.
Summary
Although there has been a large amount of work done on vacuum breakdown
phenomena, so far, no single theory has been able to explain all the available
experimental measurements and observations. Since experimental evidence exists for
all the postulated mechanisms, it appears that each mechanism would depend, to a
great extent, on the conditions under which the experiments were performed. The
most significant experimental factors which influence the breakdown mechanism are:
gap length, geometry and material of the electrodes, surface uniformity and treatment
of the surface, presence of extraneous particles and residual gas pressure in the
vacuum gap. It was observed that the correct choice of electrode material, and the use of thin insulating coatings in long gaps can increase the breakdown voltage of a
vacuum gap. On the other hand, an increase of electrode area or the presence of
particles in the vacuum gap will reduce the breakdown voltage.
The idea of using vacuum for insulation purposes is very old. According to the
Townsend theory, the growth of current in a gap depends on the drift of the charged
particles. In the absence of any such particles, as in the case of perfect vacuum, there
should be no conduction and the vacuum should be a perfect insulating medium.
However, in practice, the presence of metallic electrodes and insulating surfaces
within the vacuum complicate the issue and, therefore, even in vacuum, a sufficiently
high voltage will cause a breakdown.
In recent years a considerable amount of work has been done to determine the
electrical properties of high vacuum. This is mainly aimed at adopting such a medium
for a wide range of applications in devices such as vacuum contractors and interrupters,
high frequency capacitors and relays, electrostatic generators, microwave tubes,
etc. The contractors and circuit breakers using vacuum as insulation are finding
increasing applications in power systems.
What Is Vacuum?
A vacuum system which is used to create vacuum is a system in which the pressure is
maintained at a value much below the atmospheric pressure. In vacuum systems the
pressure is always measured in terms of millimetres of mercury, where one standard
atmosphere is equal to 760 millimetres of mercury at a temperature OfO0C. The term
*'millimetres of mercury" has been standardised as "Torr" by the International
Vacuum Society, where one millimetre of mercury is taken as equal to one Torn
Vacuum may be classified as
High vacuum : 1 x 1(T3 to 1 x IGT6 Ton-
Very high vacuum : 1 x IQT6 to 1 x 1(T8 Ton-
Ultra high vacuum : 1 x 10~9 torr and below.
For electrical insulation purposes, the range of vacuum generally used is the* 'high
vacuum", in the pressure range of 10~3 Torr to 1(T6 Torr.
Vacuum Breakdown
In the Townsend type of discharge in a gas described earlier, electrons get multiplied due to various ionisalion processes and an electron avalanche is formed. In a high vacuum, even if the electrodes are separated by, say, a few centimetres, an electron crosses the gap without encountering any collisions. Therefore, the current growth prior to breakdown cannot be due to the formation of electron avalanches. However, if a gas is liberated in the vacuum gap, then, breakdown can occur in the manner described by the Townsend process. Thus, the various breakdown mechanisms in high vacuum aim at establishing the way in which the liberation of gas can be brought about in a vacuum gap.
During the last 70 years or so, many different mechanisms for breakdown in
vacuum have been proposed. These can be broadly divided into three categories
(a) Particle exchange mechanism
(ft) Field emission mechanism
(c) Clump theory
(a) Particle Exchange Mechanism
In this mechanism it is assumed that a charged particle would be emitted from one
electrode under the action of the high electric field, and when it impinges on the other electrode, it liberates oppositely charged particles. These particles are accelerated by the applied voltage back to the first electrode where they release more of the original type of particles. When this process becomes cumulative, a chain reaction occurs which leads to the breakdown of the gap.
The particle-exchange mechanism involves electrons, positive ions, photons and
the absorbed gases at the electrode surfaces. Qualitatively, an electron present in the vacuum gap is accelerated towards the anode, and on impact releases A positive ions and C photons. These positive ions are accelerated towards the cathode, and on impact
each positive ion liberates B electrons and each photon liberates D electrons. This is
shown schematically in Fig. 2.24. The breakdown will occur if the coefficients of
production of secondary electrons exceeds unity. Mathematically, the condition for
breakdown can be written as
(AB + CD) > I (2.32)
Later, Trump and Van de Graaff measured these coefficients and showed that
they were too small for this process to take place. Accordingly, this theory was
modifie4 to allow for the presence of negative ions and the criterion for breakdown
then becomes
(AB + EF) >l (2.33)
Where A and B are the same as before and E and F represent the coefficients for
negative and positive ion liberation by positive and negative ions. It was experimentally
found that the values of the product EF were close enough to unity for copper,
aluminium and stainless steel electrodes to make this mechanism applicable at
voltages above 250 kV.
(b) Field Emission Theory
(O Anode Heating Mechanism
This theory postulates that electrons produced at small micro-projections on the
cathode due to field emission bombard the anode causing a local rise in temperature
and release gases and vapours into the vacuum gap. These electrons ionise the atoms
of the gas and produce positive ions. These positive ions arrive at the cathode, increase the primary electron emission due to space charge formation and produce secondary electrons by bombarding the surface. The process continues until a sufficient number of electrons are produced to give rise to breakdown, as in the case of a low pressure Townsend type gas discharge.
Cathode Heating Mechanism
This mechanism postulates that near the breakdown voltages of the gap, sharp
points on the cathode surface are responsible for the existence of the pre-breakdown
current, which is generated according to the field emission process described below.
This current causes resistive heating at the tip of a point and when a critical current density is reached, the tip melts and explodes, thus initiating vacuum discharge. This mechanism is called field emission as shown schematically in Fig. 2.26. Thus, the initiation of breakdown depends on the conditions and the properties of the cathode surface. Experimental evidence shows that breakdown takes place by this process when the effective catliode electric field is of the order of 106 to 107V/cm.
Clump Mechanism
Basically this theory has been developed on the following assumptions (Fig. 2.27):
(O A loosely bound particle (clump) exists on one of the electrode surfaces.
(11) On the application of a high voltage, this particle gets charged, subsequently
gets detached from the mother electrode, and is accelerated across the gap.
(///) The breakdown occurs due to a discharge in the vapour or gas released by the
impact of the particle at the target electrode.
Cranberg was the first to propose this theory. He initially assumed that breakdown
will occur when the energy per unit area, W9 delivered to the target electrode by a
clump exceeds a value C, a constant, characteristic of a given pair of electrodes. The quantity W is the product of gap voltage (V) and the charge density on the clump. The latter is proportional to the electric field E at the electrode of origin. The criterion for breakdown, therefore, is
VE = C
In case of parallel plane electrodes the field E = VId9 where d is the distance
between the electrodes. So the generalised criterion for breakdown becomes
V=(Cd)^ (2.35)
where C is another constant involving C and the electrode surface conditions.
Cranberg presented a summary of the experimental results which satisfied this
breakdown criterion with reasonable accuracy. He stated that the origin of the clump
was the cathode and obtained a value for the constant C as 60 x 101^ V2/cm (for iron
particles). However the equation was later modified as V = C d01, where a varies
between 0.2 and 1.2 depending on the gap length and the electrode material, with a
maximum at 0.6. The dependence of V on the electrode material, comes from the
observations of markings on the electrode surfaces. Craters were observed on the
anode and melted regions on the cathode or vice-versa after a single breakdown.
Summary
Although there has been a large amount of work done on vacuum breakdown
phenomena, so far, no single theory has been able to explain all the available
experimental measurements and observations. Since experimental evidence exists for
all the postulated mechanisms, it appears that each mechanism would depend, to a
great extent, on the conditions under which the experiments were performed. The
most significant experimental factors which influence the breakdown mechanism are:
gap length, geometry and material of the electrodes, surface uniformity and treatment
of the surface, presence of extraneous particles and residual gas pressure in the
vacuum gap. It was observed that the correct choice of electrode material, and the use of thin insulating coatings in long gaps can increase the breakdown voltage of a
vacuum gap. On the other hand, an increase of electrode area or the presence of
particles in the vacuum gap will reduce the breakdown voltage.
Conduction and Breakdown in Gases(GASES AS INSULATING MEDIA)
The simplest and the most commonly found dielectrics are gases. Most of the
electrical apparatus use air as the insulating medium, and in a few cases other gases
such as nitrogen (N^, carbon dioxide (CO^, freon (CC12F2) and sulphur hexafluoride
(SF$) are also used.
Various phenomena occur in gaseous dielectrics when a voltage is applied.
When the applied voltage is low, small currents flow between the electrodes
and the insulation retains its electrical properties. On the other hand, if the
applied voltages are large, the current flowing through the insulation increases
very sharply, and an electrical breakdown occurs. A strongly conducting spark
formed during breakdown practically produces a short circuit between the
electrodes. The maximum voltage applied to the insulation at the moment of
breakdown is called the breakdown voltage. In order to understand the breakdown
phenomenon in gases, a study of the electrical properties of gases and the processes
by which high currents are produced in gases is essential.
The electrical discharges in gases are of two types, i.e. (O non-sustaining
discharges, and (U) self-sustaining types. The breakdown in a gas, called spark
breakdown is the transition of a non-sustaining discharge into a self-sustaining
discharge. The build-up of high currents in a breakdown is due to the process
known as ionization in which electrons and ions are created from neutral atoms
or molecules, and their migration to the anode and cathode respectively leads
to high currents. At present two types of theories, viz. (i) Townsend theory,
and (U) Streamer theory are known which explain the mechanism for breakdown
under different conditions. The various physical conditions of gases, namely,
pressure, temperature, electrode field configuration, nature of electrode surfaces,
and the availability of initial conducting particles are known to govern the ionization
processes.
electrical apparatus use air as the insulating medium, and in a few cases other gases
such as nitrogen (N^, carbon dioxide (CO^, freon (CC12F2) and sulphur hexafluoride
(SF$) are also used.
Various phenomena occur in gaseous dielectrics when a voltage is applied.
When the applied voltage is low, small currents flow between the electrodes
and the insulation retains its electrical properties. On the other hand, if the
applied voltages are large, the current flowing through the insulation increases
very sharply, and an electrical breakdown occurs. A strongly conducting spark
formed during breakdown practically produces a short circuit between the
electrodes. The maximum voltage applied to the insulation at the moment of
breakdown is called the breakdown voltage. In order to understand the breakdown
phenomenon in gases, a study of the electrical properties of gases and the processes
by which high currents are produced in gases is essential.
The electrical discharges in gases are of two types, i.e. (O non-sustaining
discharges, and (U) self-sustaining types. The breakdown in a gas, called spark
breakdown is the transition of a non-sustaining discharge into a self-sustaining
discharge. The build-up of high currents in a breakdown is due to the process
known as ionization in which electrons and ions are created from neutral atoms
or molecules, and their migration to the anode and cathode respectively leads
to high currents. At present two types of theories, viz. (i) Townsend theory,
and (U) Streamer theory are known which explain the mechanism for breakdown
under different conditions. The various physical conditions of gases, namely,
pressure, temperature, electrode field configuration, nature of electrode surfaces,
and the availability of initial conducting particles are known to govern the ionization
processes.
Friday, September 25, 2009
Google payment
What Is Pay Per Click (PPC)?
To start with me must understand the market. I am sure like most people you use Google for your searching. While PPC is not limited to Google it is certainly dominated by them. As a consumer myself I use Google to search for information and I am sure you do the same. Research has proven that we are not alone, with the majority of consumer searches performed on Google, often to search for information, products and services. The reason, because Google finds exactly what we are looking for in a matter of seconds and it gives very relevant and clear results. Clearly Google's algorithm is far ahead of the competition in it's ability to find the right information first time. As a business owner who wants to market your product and has the need to get your products in front of as many people as possible, paying Google for one its Google advertising product, AdWords pay per click is a sure fire way to increase web traffic. The issue is not whether Google can help your business gain more clicks and hits because they have proven results. The real question for a business to ask of the PPC service is will I get a return on my investment?
Google PPC places relevant ads along the right-hand side of its organic search listings. If you have ever used Google's search engine, you will have seen the AdWords advertisements listed under "Sponsored Links" on the right side of the page.
Managed Pay-Per-Click Campaign
Google AdWords campaigns can more than quadruple monthly traffic. Google AdWords will only appear to people who enter keyphrases or keywords that are relevant to your company's offerings and which you have setup your campaign to target. A managed PPC campaign will achieve far greater return on investment than a self managed PPC campaign, purely because expert online marketing company will know more of the hints, tips and tricks.
To further increase traffic, a professional PPC company will use the Google advertising network. Google pays companies for displaying AdWords advertisements on their websites, this is called affiliate marketing and is part of the Google advertising network. For example, if your company specializes in discounted computers then you would pay Google for advertising space on all searches for "discounted computers". Google would not only display the AdWords advert on its own site but also display the ad on informational sites dedicated to discounted computers.
What Do, We Do?
Google AdWords should not be your entire online marketing campaign, it should certainly be part of it, as it is an excellent addition to an overall marketing strategy. Google AdWords is a pay-per-click service, meaning that you only pay when the advertising results in traffic. The driving force of the Adwords system are the choices of keywords and keyphrases for use in the advertising campaign. Why, because these are the searches that potential customers make and result in the page where your ad is displayed.
Before you start using Google AdWords as a marketing tool, I highly recommend watching the Getting Started With Google Adwords video which is available after you log into your Adwords account. I would also suggest reading all of the Google support guides before you start bidding on keywords and phrases. Reason? Because you can spend a lot of money if you don't know what you're doing.
Here at Kanga we pride ourselves on being a pay per click authority and we know the google PPC Adwords system inside out. We manage professional PPC campaigns for our clients that deliver in terms of traffic and ROI. We give you a great value for money pay per click campaign. Contact us to find out how we can help your business today!
To start with me must understand the market. I am sure like most people you use Google for your searching. While PPC is not limited to Google it is certainly dominated by them. As a consumer myself I use Google to search for information and I am sure you do the same. Research has proven that we are not alone, with the majority of consumer searches performed on Google, often to search for information, products and services. The reason, because Google finds exactly what we are looking for in a matter of seconds and it gives very relevant and clear results. Clearly Google's algorithm is far ahead of the competition in it's ability to find the right information first time. As a business owner who wants to market your product and has the need to get your products in front of as many people as possible, paying Google for one its Google advertising product, AdWords pay per click is a sure fire way to increase web traffic. The issue is not whether Google can help your business gain more clicks and hits because they have proven results. The real question for a business to ask of the PPC service is will I get a return on my investment?
Google PPC places relevant ads along the right-hand side of its organic search listings. If you have ever used Google's search engine, you will have seen the AdWords advertisements listed under "Sponsored Links" on the right side of the page.
Managed Pay-Per-Click Campaign
Google AdWords campaigns can more than quadruple monthly traffic. Google AdWords will only appear to people who enter keyphrases or keywords that are relevant to your company's offerings and which you have setup your campaign to target. A managed PPC campaign will achieve far greater return on investment than a self managed PPC campaign, purely because expert online marketing company will know more of the hints, tips and tricks.
To further increase traffic, a professional PPC company will use the Google advertising network. Google pays companies for displaying AdWords advertisements on their websites, this is called affiliate marketing and is part of the Google advertising network. For example, if your company specializes in discounted computers then you would pay Google for advertising space on all searches for "discounted computers". Google would not only display the AdWords advert on its own site but also display the ad on informational sites dedicated to discounted computers.
What Do, We Do?
Google AdWords should not be your entire online marketing campaign, it should certainly be part of it, as it is an excellent addition to an overall marketing strategy. Google AdWords is a pay-per-click service, meaning that you only pay when the advertising results in traffic. The driving force of the Adwords system are the choices of keywords and keyphrases for use in the advertising campaign. Why, because these are the searches that potential customers make and result in the page where your ad is displayed.
Before you start using Google AdWords as a marketing tool, I highly recommend watching the Getting Started With Google Adwords video which is available after you log into your Adwords account. I would also suggest reading all of the Google support guides before you start bidding on keywords and phrases. Reason? Because you can spend a lot of money if you don't know what you're doing.
Here at Kanga we pride ourselves on being a pay per click authority and we know the google PPC Adwords system inside out. We manage professional PPC campaigns for our clients that deliver in terms of traffic and ROI. We give you a great value for money pay per click campaign. Contact us to find out how we can help your business today!
Monday, September 21, 2009
The Most Expensive Keywords
By knowing the most expensive keywords on the internet, you can create websites and web pages based on these keywords. On these sites and pages, you can serve expensive ads and/or promote affiliate offers that reap high bounties when clicked or completed.
The challenge lies in driving traffic to these sites. One way to drive traffic is through search engine optimization, whereby you create a website that ranks highly on the search engines and receive “free” traffic. Or, if you already have an existing well-ranked site, you can create new pages linked from your existing pages. A second way to drive traffic is through online advertising.
With regards to advertising, one key is to understand correlations between keywords. By determining correlations between inexpensive and expensive keywords, you can advertise on the inexpensive keyword (via pay-per-click (PPC) advertising on search engines) to drive traffic to your website that serves ads based on expensive keywords. For instance, advertising on the keyword “painting” ($0.20 keyword), and sending visitors to your site about the related topic “home remodeling” ($2.00 keyword) allows you to profit from the wide spread between the keyword prices of the two correlated keywords.
Likewise there are many opportunities to run banner and other ads through online advertising networks to drive traffic to your website. These ads can run on both “painting” and “home remodeling” websites. The cost of CPM (cost per impression) advertising typically does not reflect the price of the keywords. That is, a banner ad on a “painting” website typically costs the same as a banner ad on a “home remodeling” website. This sharply contrasts the price of PPC advertising on search engines which vary widely based on the keyword.
There are two other ways to profit from knowing the most expensive keywords. The first doesn’t even require you to own or operate a website. In this method, you advertise using inexpensive means (e.g., via pay-per-click advertising on inexpensive keywords or CPM advertising) and direct traffic to high paying affiliate offers. The second method is to set up your own search engine. There are several companies that allow you to easily set up a search engine that serves PPC ads syndicated from Google, Overture or other PPC search engine networks. Once you create your engine, you can drive traffic directly to your PPC results, and get paid every time someone clicks on one of the results. Likewise, on your search engine’s home page, you can include many links to expensive keywords, so that when someone clicks on it, the engine shows expensive ads based on those keywords.
Understanding keyword pricing is critical to the online advertising, PPC search engine and affiliate marketing businesses – businesses that combine for over $25 billion a year in revenues. With a market of this size, generating substantial profits, if you have the right information, is really not that challenging
The challenge lies in driving traffic to these sites. One way to drive traffic is through search engine optimization, whereby you create a website that ranks highly on the search engines and receive “free” traffic. Or, if you already have an existing well-ranked site, you can create new pages linked from your existing pages. A second way to drive traffic is through online advertising.
With regards to advertising, one key is to understand correlations between keywords. By determining correlations between inexpensive and expensive keywords, you can advertise on the inexpensive keyword (via pay-per-click (PPC) advertising on search engines) to drive traffic to your website that serves ads based on expensive keywords. For instance, advertising on the keyword “painting” ($0.20 keyword), and sending visitors to your site about the related topic “home remodeling” ($2.00 keyword) allows you to profit from the wide spread between the keyword prices of the two correlated keywords.
Likewise there are many opportunities to run banner and other ads through online advertising networks to drive traffic to your website. These ads can run on both “painting” and “home remodeling” websites. The cost of CPM (cost per impression) advertising typically does not reflect the price of the keywords. That is, a banner ad on a “painting” website typically costs the same as a banner ad on a “home remodeling” website. This sharply contrasts the price of PPC advertising on search engines which vary widely based on the keyword.
There are two other ways to profit from knowing the most expensive keywords. The first doesn’t even require you to own or operate a website. In this method, you advertise using inexpensive means (e.g., via pay-per-click advertising on inexpensive keywords or CPM advertising) and direct traffic to high paying affiliate offers. The second method is to set up your own search engine. There are several companies that allow you to easily set up a search engine that serves PPC ads syndicated from Google, Overture or other PPC search engine networks. Once you create your engine, you can drive traffic directly to your PPC results, and get paid every time someone clicks on one of the results. Likewise, on your search engine’s home page, you can include many links to expensive keywords, so that when someone clicks on it, the engine shows expensive ads based on those keywords.
Understanding keyword pricing is critical to the online advertising, PPC search engine and affiliate marketing businesses – businesses that combine for over $25 billion a year in revenues. With a market of this size, generating substantial profits, if you have the right information, is really not that challenging
Friday, September 11, 2009
RENEWABLE ENERGY SOURCES - WIND ENERGY
INTRODUCTION
Renewable energy sources capture their energy from existing flows of energy, from on-going natural processes, such as sunshine, wind, flowing water, biological processes, and geothermal heat flows. Renewable energy resources may be used directly, or used to create other more convenient forms of energy.
Wind is commercially and operationally the most viable renewable energy resource and accordingly, emerging as one of the largest source in terms of the renewable energy sector. Wind power is the kinetic energy of wind, or the extraction of this energy by wind turbines. In 2004, wind power became the least expensive form of new power generation, dipping below the cost per kilowatt-hour of coal-fired plants. Wind power is growing faster than any other form of electrical generation, at about 37%, up from 25% growth in 2002. In the late-1990s, the cost of wind power was about five times what it is in 2005, and that downward trend is expected to continue as larger multi-megawatt turbines are mass-produced.
An estimated 1 to 3 percent of the energy from the Sun is converted into wind energy. This is about 50 to 100 times more energy than what is converted into biomass by all the plants on earth through photosynthesis.
India now ranks as a "wind superpower" with an installed wind power capacity of 1167 MW and about 5 billion units of electricity have been fed to the national grid so far.In progress are wind resource assessment programme, wind monitoring, wind mapping, covering 800 stations in 24 states with 193 wind monitoring stations in operations. Altogether 13 states of India have a net potential of about 45000 MW.
Why Wind Energy
The project is environment friendly.
Good wind potential to harness wind energy.
A permanent shield against ever increasing power prices. The cost per kwh reduces over a period of time as against rising cost for conventional power projects.
The cheapest source of electrical energy. (on a levelled cost over 20 years.)
Least equity participation required, as well as low cost debt is easily available to wind energy projects.
A project with the fastest payback period.
A real fast track power project, with the lowest gestation period; and a modular concept.
Operation and Maintenance costs are low.
No marketing risks, as the product is electrical energy.
A project with no investment in manpower
The wind power generation in the country is influenced to a great extent by the wind speed and wind power density prevalent at a particular potential location at any given point of time. The wind speed is affected to a large extent by the strong southwesterly monsoons, starting in May-June, and at the same time by the weaker northeastern monsoons in the winter months. It has been generally observed that 60-70% of the total wind power generation in the country takes place during June- October when the southwest monsoons are prevalent through out the country. According to a latest study, locations having an annual mean wind power density greater than 150 watts/ square meter at 30 meter hub height have been found to be suitable for development of wind power projects.
PRESENT SCENARIO OF WIND ENERGY
Exploitation of wind energy has been in place from time immemorial but the development of technology for tapping the same for generation of grid quality electricity is of a recent origin. India has been quick to make a foray in this area. It has made its mark as one of the top ranking countries in the world in wind power generation. With an installed generation capacity of 1702.30 MW as on 31.3.2005 of wind power, India now ranks 5th in the world after Germany, USA, Denmark and Spain in wind power generation. According to a recent estimate, the gross wind power generation potential in the country is estimated at 45,195 MW at 50 meter. Hub Height. Hub height is defined as the height from the Ground Level at which the hub of the propeller blades of the wind energy generator is situated.
The state wise potential and installed capacity is given in the table below:
State Gross
Potential in MW Total Installed Capacity in MW
Demonstration
Projects (MW) Private Sector
Projects (MW) Total Capacity (MW)
Andhra Pradesh 8275 5.40 87.20 92.60
Gujarat 9675 17.30 149.60 166.90
Karnataka 6620 2.60 93.60 96.20
Kerala 875 2.00 0.00 2.00
Madhya Pradesh 5500 0.60 22.00 22.60
Maharashtra 3650 6.40 392.80 399.20
Orissa 1700 6.40 18.70 25.10
Rajasthan 5400 19.40 875.60 895.00
Tamil Nadu 3050 1.10 0.00 1.10
West Bengal 450 1.60 0.00 1.60
Total 45195 62.80 1639.50 1702.30
CONTRIBUTION OF TAMIL NADU
Tamil Nadu has pioneered the effective usage of wind resources and the State Government is keen on encouraging wind energy production, Windmills in Tamil Nadu produced 55 per cent of the total wind-generated electricity in the country, the total power capacity of 1,664 MW was available from winds mills in the private sector and 9 MW of power from the windmills of tamilnadu electricity board (TNEB). The State had contributed 57 per cent of the total 2,909 MW of harnessed energy in the country. As of 2005, Tamil Nadu was one of the few states with surplus power enabling the electricity authority to sell it to neighbouring states.
PROMOTIONAL POLICIES AND SUBSIDIES
Govt. of India and state govts have developed suitable policies and guidelines for providing technical help, financial support and various other incentives for development of wind power in the country. These include R&D activities for design and development of low cost indigenous wind energy harnessing technologies, dissemination of the developed technologies through demonstration projects, setting up of the commercial wind farms through central and state government subsidy, providing financial incentives to potential entrepreneurs etc.
The various incentives that are being provided by the central and the state governments are as per the details given below:
From Central Government
Income Tax Holiday
Accelerated Depreciation
Concessional Custom Duty/ Duty Free Import
Capital/ Interest Subsidy
From State Governments
Energy buyback, power wheeling and banking facilities
Sales tax concession benefits
Electricity tax exemption
Demand cut concession offered to industrial consumers who establish power generating units from renewable energy sources
Capital Subsidy
The table given below depicts the initiatives provided by some of the state governments towards development of commercial wind power projects.
Andhra
Pradesh Karnataka Madhya
Pradesh Maharashtra Rajasthan Tamil Nadu
Wheeling 2% of energy 2% of energy 2% of energy 2% of energy 2% of energy 2% of energy
Banking
12 months 2% p.m. for 12 months - 12 Months 12 Months 12 Months
Third Party Sale Not allowed Allowed Allowed Allowed Allowed Not Allowed
Capital Subsidy
20%
Max. Rs. 25.00 Lakh Max. Rs. 25.00
Lakh for backward areas Same as other industries 30% Max. Rs. 30.00 Lakh - -
ESSENTIAL REQUIREMENTS OF WINDMILLS
An area where a number of wind electric generators are installed is known as a wind farm. The essential requirements for establishment of a wind farm for optimal exploitation of the wind are :
1 High wind resource at particular site
2 Adequate land availability
3 Suitable terrain and good soil condition
4 Proper approach to site
5 Suitable power grid nearby
6 Techno-economic selection of WEGs
7 Scientifically prepared layout
Features of wind turbine generators
TECHNICAL DESCRIPTION
Components of wind turbine generators are as follows:
DESIGN:
During assembling, the gearbox with all the attached nacelle components is easily mounted on the top of the yaw system with no need for any additional mounting of cables,pipes etc. except for the nacelle top cover and the rotor.
The design of the gearbox housing is simple and robust, with a supporting structure which transmits all dynamic stress and forces from the rotor directly into the tower structure.
All parts and components are manufactured of certified quality steel. Every contact face is machined in CNC milling machines prior to assembly. The gearbox housing is machined in one single operation to prevent any misalignments and ensuring full compatibility between individual units. In fact, there is no need for alignment at all during the assembly, thus making maintenance easier, simpler and less costly.
This leads to the following advantages:
A very compact, simple and robust construction
A functional and straight forward lay-out
Very safe and user-friendly working conditions
Long lifetime and low operational and maintenance costs
GEARBOX HOUSING:
The very high efficiency of the transmission is reflected by the extreme-low levels of noise emission, vibration and temperature rise during operation.
The gearbox housing, is welded instead of cast since this gives a more ductile and sturdy gear construction. The integrated gearbox is designed to transmit all static and dynamic forces directly into the tower construction. Also, the gearbox acts as the structural support for the entire nacelle and incorporated parts there.
At the front of the gearbox a parking disc, brake is mounted on the high speed shaft. On the rear side, the generator is flange mounted with a flexible coupling to the high speed shaft. The electrical oil pressure pump and the hydraulic pump unit are placed on the side of the gearbox.
The gearbox has a heavy steel bottom plate, which is connected to the yaw system through 5 heavy duty silent blocks (damper supports). The system effectively eliminates the transmission of noise and vibrations from the nacelle to the tower.
SHAFT AND COG WHEELS:
The integrated gearbox has 3 stages with 4 parallel shafts or 2 stages with 3 parallel shafts depending on the generator pole number. The efficiency of the individual steps are very high. The low speed shaft is hollow and manufactured in one piece with a flange for direct mounting of the rotor. Hydraulic pressure is lead through the shaft for control of the blade tip airbrakes.
BEARINGS:
The bearings on all shafts inside the gearbox are of the spherical roller bearing type. Mounting this type of bearings on all shafts in the gearbox secures that parallelism of shafts as well as build-in axial tolerances are maintained during operation.
BRAKE SYSTEM:
The brake system is of the negative fail-safe type and consists of 3 independent blade tip airbrakes and a parking brake with primary and secondary functions controlled by a hydraulic pump unit. An active hydraulic pressure keeps the blade tips in an operational situation, but as soon as a stop command is encountered or at any loss of electrical power from the grid, the oil pressure drops which instantly releases the airbrakes.
BLADE TIP AIRBRAKES:
Each blade is equipped with a turnable blade tip which in the activated position is turned 88 degrees out of the rotor plane and thereby acting as an efficient airbrake. The blade tip is mounted on a stainless steel rod embedded in the main blade. During normal operation, the blade tip is kept in the passive (operational) position by a stainless steel wire connected to a hydraulic cylinder under pressure. There are 3 of these cylinders, one for each blade, located in each blade root at the hub.
Once the hydraulic pressure is released, all 3 blade tips will be turned independently into the braking position by the centrifugal forces arising from the rotation and a very effective decrease of the rotor speed will take place. Even with only two activated airbrakes, the rotation will be kept well below the normal operational speed.
By re-setting the hydraulic pressure, each blade tip will be turned back into the operating position and the parking brake will be released.
PARKING BRAKE:
The parking brake is a fail-safe unit with a built-in spring-activated braking force. The brake disc is mounted on the high speed shaft. An active hydraulic pressure keeps the brake caliber in the open (operational) position.
The brake pads are self-adjusting, i.e. they keep a constant clearance between brake disc and brake pads. A pad wear sensor indicates time for replacement on the wind turbine control panel.
HYDRAULIC PUMP UNIT AND HYDRAULIC OIL:
The hydraulic pressure from the pump unit controls the blade tip airbrakes and the parking brake. A filter unit secures clean hydraulic oil in the system. Only a high quality hydraulic oil is used which has an appropriate viscosity, even at extreme low temperatures.
PRIMARY BRAKE SYSTEM:
The primary brake system is controlled by the wind turbine computer and the hydraulic pump unit. The first step when encountering a stop command is to release all 3 blade tip airbrakes. Instantly the airbrakes will decrease the rotor 3peed. At 40% of the nominal rotor speed or, at the latest after15 seconds, the parking brake will be activated as the second step and bring the rotor to a complete stop.
GENERATOR:
The generator is flange mounted to a machined bracket on the rear of the gearbox. A recess on the flange secures a well defined positioning and a perfect alignment with the high speed shaft for the mounting of the elastic coupling through which the rotor torque is transmitted.
TYPE:
The generator is a highly efficient asynchronous IP54 totally enclosed machine designed and built especially to wind turbine applications. This means that the generator reaches its maximum efficiency factor already at 50% full load.
INSULATION CLASS AND MAXIMUM OUTPUT:
The name plate rating of 150 kW refers to "Class B" insulation and temperature raise. However the generator is built with "Class F" insulation which adds an extra allowable temperature rise of 20°C (36 OF). This secures against unnecessary stops due to high ambient temperatures or other operational conditions.
The maximum power output is not depending on the wind speed
only. Air pressure, air temperature and air turbulence have influence on the energy contents of the wind. This means that identical wind turbines could have widely different power out-puts at apparently equal wind speeds. It is normal to experience variations, of up to +/-15% in the maximum power output in relation to the standard power curve, i.e. at standard atmospherical conditions: Air temperature 15°C (59 OF), air pressure 1013 hPa and max. 15% turbulence intensity.
SENSOR:
To monitor the generator temperature, PT l00 temperature sensors and thermistors have been build into the windings. The PT lOO sensor is used for directly read-out of the temperature on the controller display as well as for shut-down of the turbine when exceeding a certain, user-defined temperature limit. The thermistor has a higher, build-in maximum limit, which forces the turbine to a shut-down when exceeded. Both sensors are doubled with a spare set.
COOLING EQUIPMENT:
The generator is surrounded by a ventilation duct of steel plate and is effectively cooled by a fan which forces an air flow to pass the cooling fins of the generator. The nacelle cabin is divided in a way which prevents re-circulation of the cooling air.
CABLE TWIST SENSOR:
On the underside of the yaw mounting plate a cable twist sensor is located in junction with a cogwheel in order to count the net yawing angle. At three full 360 degrees turns to one side (net), the controller automatically brings the rotor to a complete stop, un-twists the cable by counter yawing and re-starts the wind turbine.
PERFORMANCE
The figure below shows the power delivered by the windmill at the two velocity of rotation settings. The lower 80 rpm value is used to start up the windmill and for wind speeds up to 8 m/s, whereas the 120 rpm setting is used for wind speeds between 8 and 14 m/s. The design power of 90 kW is reached at a wind speed of 13 m/s. For comparison, the chart also contains the power curve of a windmill using conventional airfoil sections, which result in a undesired power output at higher wind speeds.
CONCLUSION
Inspite of the availability of various financial incentives and availability of technological know-how, the development of wind power is very tardy in the country. Urgent efforts are required for the design and development of low cost, simple to use wind turbines . Suitable extension mechanism has to be devised wherein the benefits of development of wind power can be disseminated to the rural communities, village panchayats so that collective organizational skills can be developed. By doing these, wind energy, the most valuable, pollution free energy resource can be saved and can be made available for the benefits of the future generation. benefits of the renewable energy over the non renewable energy and the availability and technical description of wind power plants in India.
REFERENCES
1] Rakesh Bakshi, “Wind Energy in India,” IEEE Power Engineering Review, pp 16-18, September 2002
[2] Bhim Singh, “Induction generators–A prospective,” Int. J on Electric machines and power systems, vol.23, 1995, pp. 163-177
[3] C. F. Wagner, “Self-excitation of induction Motors,” Trans. Amer. Inst.Elect. Eng., vol. 58, pp. 47–51, Feb. 1939.
[4] C. Grantham, D. Sutanto, and B. Mismail, “Steady-state and transient analysis of self-excited induction generators,” Proc. Inst. Elect. Eng., pt. B, vol. 136, no. 2, pp. 61–68, Mar. 1989.
[5] J. M. Elder, J. T. Boys, and J. L. Woodward, “Self-excited induction machine as a small low-cost generator,” Proc. Inst. Elect. Eng., pt. C, vol. 131, no. 2, pp. 33–41, Mar. 1984.
[6] M. H. Salama and P. G. Holmes, “Transient and steady-state load performance of stand-alone self-excited induction generator,” Proc. IEE—Elect. Power Applicat. vol. 143, no. 1, pp. 50–58, Jan. 1996.
[7] Dawit Seyoum Colin Grantham, and Muhammed Fazlur Rahman, “The dynamic characteristics of an isolated self-excited induction generator driven by a wind turbine”, Trans. Inst. Elect. Eng., Industry Applicat, vol. 39, no. 4, July/August 2003
[8] Rohin M Hilloowala and Adel M Sharaf, “A Rule-based fuzzy controller for a PWM inverter in a stand-alone wind energy conversion scheme”, Trans. Inst Elect. Eng. Industry Applicat, vol. 32, no. 1, pp. 57-65, Jan/Feb 1996.
Renewable energy sources capture their energy from existing flows of energy, from on-going natural processes, such as sunshine, wind, flowing water, biological processes, and geothermal heat flows. Renewable energy resources may be used directly, or used to create other more convenient forms of energy.
Wind is commercially and operationally the most viable renewable energy resource and accordingly, emerging as one of the largest source in terms of the renewable energy sector. Wind power is the kinetic energy of wind, or the extraction of this energy by wind turbines. In 2004, wind power became the least expensive form of new power generation, dipping below the cost per kilowatt-hour of coal-fired plants. Wind power is growing faster than any other form of electrical generation, at about 37%, up from 25% growth in 2002. In the late-1990s, the cost of wind power was about five times what it is in 2005, and that downward trend is expected to continue as larger multi-megawatt turbines are mass-produced.
An estimated 1 to 3 percent of the energy from the Sun is converted into wind energy. This is about 50 to 100 times more energy than what is converted into biomass by all the plants on earth through photosynthesis.
India now ranks as a "wind superpower" with an installed wind power capacity of 1167 MW and about 5 billion units of electricity have been fed to the national grid so far.In progress are wind resource assessment programme, wind monitoring, wind mapping, covering 800 stations in 24 states with 193 wind monitoring stations in operations. Altogether 13 states of India have a net potential of about 45000 MW.
Why Wind Energy
The project is environment friendly.
Good wind potential to harness wind energy.
A permanent shield against ever increasing power prices. The cost per kwh reduces over a period of time as against rising cost for conventional power projects.
The cheapest source of electrical energy. (on a levelled cost over 20 years.)
Least equity participation required, as well as low cost debt is easily available to wind energy projects.
A project with the fastest payback period.
A real fast track power project, with the lowest gestation period; and a modular concept.
Operation and Maintenance costs are low.
No marketing risks, as the product is electrical energy.
A project with no investment in manpower
The wind power generation in the country is influenced to a great extent by the wind speed and wind power density prevalent at a particular potential location at any given point of time. The wind speed is affected to a large extent by the strong southwesterly monsoons, starting in May-June, and at the same time by the weaker northeastern monsoons in the winter months. It has been generally observed that 60-70% of the total wind power generation in the country takes place during June- October when the southwest monsoons are prevalent through out the country. According to a latest study, locations having an annual mean wind power density greater than 150 watts/ square meter at 30 meter hub height have been found to be suitable for development of wind power projects.
PRESENT SCENARIO OF WIND ENERGY
Exploitation of wind energy has been in place from time immemorial but the development of technology for tapping the same for generation of grid quality electricity is of a recent origin. India has been quick to make a foray in this area. It has made its mark as one of the top ranking countries in the world in wind power generation. With an installed generation capacity of 1702.30 MW as on 31.3.2005 of wind power, India now ranks 5th in the world after Germany, USA, Denmark and Spain in wind power generation. According to a recent estimate, the gross wind power generation potential in the country is estimated at 45,195 MW at 50 meter. Hub Height. Hub height is defined as the height from the Ground Level at which the hub of the propeller blades of the wind energy generator is situated.
The state wise potential and installed capacity is given in the table below:
State Gross
Potential in MW Total Installed Capacity in MW
Demonstration
Projects (MW) Private Sector
Projects (MW) Total Capacity (MW)
Andhra Pradesh 8275 5.40 87.20 92.60
Gujarat 9675 17.30 149.60 166.90
Karnataka 6620 2.60 93.60 96.20
Kerala 875 2.00 0.00 2.00
Madhya Pradesh 5500 0.60 22.00 22.60
Maharashtra 3650 6.40 392.80 399.20
Orissa 1700 6.40 18.70 25.10
Rajasthan 5400 19.40 875.60 895.00
Tamil Nadu 3050 1.10 0.00 1.10
West Bengal 450 1.60 0.00 1.60
Total 45195 62.80 1639.50 1702.30
CONTRIBUTION OF TAMIL NADU
Tamil Nadu has pioneered the effective usage of wind resources and the State Government is keen on encouraging wind energy production, Windmills in Tamil Nadu produced 55 per cent of the total wind-generated electricity in the country, the total power capacity of 1,664 MW was available from winds mills in the private sector and 9 MW of power from the windmills of tamilnadu electricity board (TNEB). The State had contributed 57 per cent of the total 2,909 MW of harnessed energy in the country. As of 2005, Tamil Nadu was one of the few states with surplus power enabling the electricity authority to sell it to neighbouring states.
PROMOTIONAL POLICIES AND SUBSIDIES
Govt. of India and state govts have developed suitable policies and guidelines for providing technical help, financial support and various other incentives for development of wind power in the country. These include R&D activities for design and development of low cost indigenous wind energy harnessing technologies, dissemination of the developed technologies through demonstration projects, setting up of the commercial wind farms through central and state government subsidy, providing financial incentives to potential entrepreneurs etc.
The various incentives that are being provided by the central and the state governments are as per the details given below:
From Central Government
Income Tax Holiday
Accelerated Depreciation
Concessional Custom Duty/ Duty Free Import
Capital/ Interest Subsidy
From State Governments
Energy buyback, power wheeling and banking facilities
Sales tax concession benefits
Electricity tax exemption
Demand cut concession offered to industrial consumers who establish power generating units from renewable energy sources
Capital Subsidy
The table given below depicts the initiatives provided by some of the state governments towards development of commercial wind power projects.
Andhra
Pradesh Karnataka Madhya
Pradesh Maharashtra Rajasthan Tamil Nadu
Wheeling 2% of energy 2% of energy 2% of energy 2% of energy 2% of energy 2% of energy
Banking
12 months 2% p.m. for 12 months - 12 Months 12 Months 12 Months
Third Party Sale Not allowed Allowed Allowed Allowed Allowed Not Allowed
Capital Subsidy
20%
Max. Rs. 25.00 Lakh Max. Rs. 25.00
Lakh for backward areas Same as other industries 30% Max. Rs. 30.00 Lakh - -
ESSENTIAL REQUIREMENTS OF WINDMILLS
An area where a number of wind electric generators are installed is known as a wind farm. The essential requirements for establishment of a wind farm for optimal exploitation of the wind are :
1 High wind resource at particular site
2 Adequate land availability
3 Suitable terrain and good soil condition
4 Proper approach to site
5 Suitable power grid nearby
6 Techno-economic selection of WEGs
7 Scientifically prepared layout
Features of wind turbine generators
TECHNICAL DESCRIPTION
Components of wind turbine generators are as follows:
DESIGN:
During assembling, the gearbox with all the attached nacelle components is easily mounted on the top of the yaw system with no need for any additional mounting of cables,pipes etc. except for the nacelle top cover and the rotor.
The design of the gearbox housing is simple and robust, with a supporting structure which transmits all dynamic stress and forces from the rotor directly into the tower structure.
All parts and components are manufactured of certified quality steel. Every contact face is machined in CNC milling machines prior to assembly. The gearbox housing is machined in one single operation to prevent any misalignments and ensuring full compatibility between individual units. In fact, there is no need for alignment at all during the assembly, thus making maintenance easier, simpler and less costly.
This leads to the following advantages:
A very compact, simple and robust construction
A functional and straight forward lay-out
Very safe and user-friendly working conditions
Long lifetime and low operational and maintenance costs
GEARBOX HOUSING:
The very high efficiency of the transmission is reflected by the extreme-low levels of noise emission, vibration and temperature rise during operation.
The gearbox housing, is welded instead of cast since this gives a more ductile and sturdy gear construction. The integrated gearbox is designed to transmit all static and dynamic forces directly into the tower construction. Also, the gearbox acts as the structural support for the entire nacelle and incorporated parts there.
At the front of the gearbox a parking disc, brake is mounted on the high speed shaft. On the rear side, the generator is flange mounted with a flexible coupling to the high speed shaft. The electrical oil pressure pump and the hydraulic pump unit are placed on the side of the gearbox.
The gearbox has a heavy steel bottom plate, which is connected to the yaw system through 5 heavy duty silent blocks (damper supports). The system effectively eliminates the transmission of noise and vibrations from the nacelle to the tower.
SHAFT AND COG WHEELS:
The integrated gearbox has 3 stages with 4 parallel shafts or 2 stages with 3 parallel shafts depending on the generator pole number. The efficiency of the individual steps are very high. The low speed shaft is hollow and manufactured in one piece with a flange for direct mounting of the rotor. Hydraulic pressure is lead through the shaft for control of the blade tip airbrakes.
BEARINGS:
The bearings on all shafts inside the gearbox are of the spherical roller bearing type. Mounting this type of bearings on all shafts in the gearbox secures that parallelism of shafts as well as build-in axial tolerances are maintained during operation.
BRAKE SYSTEM:
The brake system is of the negative fail-safe type and consists of 3 independent blade tip airbrakes and a parking brake with primary and secondary functions controlled by a hydraulic pump unit. An active hydraulic pressure keeps the blade tips in an operational situation, but as soon as a stop command is encountered or at any loss of electrical power from the grid, the oil pressure drops which instantly releases the airbrakes.
BLADE TIP AIRBRAKES:
Each blade is equipped with a turnable blade tip which in the activated position is turned 88 degrees out of the rotor plane and thereby acting as an efficient airbrake. The blade tip is mounted on a stainless steel rod embedded in the main blade. During normal operation, the blade tip is kept in the passive (operational) position by a stainless steel wire connected to a hydraulic cylinder under pressure. There are 3 of these cylinders, one for each blade, located in each blade root at the hub.
Once the hydraulic pressure is released, all 3 blade tips will be turned independently into the braking position by the centrifugal forces arising from the rotation and a very effective decrease of the rotor speed will take place. Even with only two activated airbrakes, the rotation will be kept well below the normal operational speed.
By re-setting the hydraulic pressure, each blade tip will be turned back into the operating position and the parking brake will be released.
PARKING BRAKE:
The parking brake is a fail-safe unit with a built-in spring-activated braking force. The brake disc is mounted on the high speed shaft. An active hydraulic pressure keeps the brake caliber in the open (operational) position.
The brake pads are self-adjusting, i.e. they keep a constant clearance between brake disc and brake pads. A pad wear sensor indicates time for replacement on the wind turbine control panel.
HYDRAULIC PUMP UNIT AND HYDRAULIC OIL:
The hydraulic pressure from the pump unit controls the blade tip airbrakes and the parking brake. A filter unit secures clean hydraulic oil in the system. Only a high quality hydraulic oil is used which has an appropriate viscosity, even at extreme low temperatures.
PRIMARY BRAKE SYSTEM:
The primary brake system is controlled by the wind turbine computer and the hydraulic pump unit. The first step when encountering a stop command is to release all 3 blade tip airbrakes. Instantly the airbrakes will decrease the rotor 3peed. At 40% of the nominal rotor speed or, at the latest after15 seconds, the parking brake will be activated as the second step and bring the rotor to a complete stop.
GENERATOR:
The generator is flange mounted to a machined bracket on the rear of the gearbox. A recess on the flange secures a well defined positioning and a perfect alignment with the high speed shaft for the mounting of the elastic coupling through which the rotor torque is transmitted.
TYPE:
The generator is a highly efficient asynchronous IP54 totally enclosed machine designed and built especially to wind turbine applications. This means that the generator reaches its maximum efficiency factor already at 50% full load.
INSULATION CLASS AND MAXIMUM OUTPUT:
The name plate rating of 150 kW refers to "Class B" insulation and temperature raise. However the generator is built with "Class F" insulation which adds an extra allowable temperature rise of 20°C (36 OF). This secures against unnecessary stops due to high ambient temperatures or other operational conditions.
The maximum power output is not depending on the wind speed
only. Air pressure, air temperature and air turbulence have influence on the energy contents of the wind. This means that identical wind turbines could have widely different power out-puts at apparently equal wind speeds. It is normal to experience variations, of up to +/-15% in the maximum power output in relation to the standard power curve, i.e. at standard atmospherical conditions: Air temperature 15°C (59 OF), air pressure 1013 hPa and max. 15% turbulence intensity.
SENSOR:
To monitor the generator temperature, PT l00 temperature sensors and thermistors have been build into the windings. The PT lOO sensor is used for directly read-out of the temperature on the controller display as well as for shut-down of the turbine when exceeding a certain, user-defined temperature limit. The thermistor has a higher, build-in maximum limit, which forces the turbine to a shut-down when exceeded. Both sensors are doubled with a spare set.
COOLING EQUIPMENT:
The generator is surrounded by a ventilation duct of steel plate and is effectively cooled by a fan which forces an air flow to pass the cooling fins of the generator. The nacelle cabin is divided in a way which prevents re-circulation of the cooling air.
CABLE TWIST SENSOR:
On the underside of the yaw mounting plate a cable twist sensor is located in junction with a cogwheel in order to count the net yawing angle. At three full 360 degrees turns to one side (net), the controller automatically brings the rotor to a complete stop, un-twists the cable by counter yawing and re-starts the wind turbine.
PERFORMANCE
The figure below shows the power delivered by the windmill at the two velocity of rotation settings. The lower 80 rpm value is used to start up the windmill and for wind speeds up to 8 m/s, whereas the 120 rpm setting is used for wind speeds between 8 and 14 m/s. The design power of 90 kW is reached at a wind speed of 13 m/s. For comparison, the chart also contains the power curve of a windmill using conventional airfoil sections, which result in a undesired power output at higher wind speeds.
CONCLUSION
Inspite of the availability of various financial incentives and availability of technological know-how, the development of wind power is very tardy in the country. Urgent efforts are required for the design and development of low cost, simple to use wind turbines . Suitable extension mechanism has to be devised wherein the benefits of development of wind power can be disseminated to the rural communities, village panchayats so that collective organizational skills can be developed. By doing these, wind energy, the most valuable, pollution free energy resource can be saved and can be made available for the benefits of the future generation. benefits of the renewable energy over the non renewable energy and the availability and technical description of wind power plants in India.
REFERENCES
1] Rakesh Bakshi, “Wind Energy in India,” IEEE Power Engineering Review, pp 16-18, September 2002
[2] Bhim Singh, “Induction generators–A prospective,” Int. J on Electric machines and power systems, vol.23, 1995, pp. 163-177
[3] C. F. Wagner, “Self-excitation of induction Motors,” Trans. Amer. Inst.Elect. Eng., vol. 58, pp. 47–51, Feb. 1939.
[4] C. Grantham, D. Sutanto, and B. Mismail, “Steady-state and transient analysis of self-excited induction generators,” Proc. Inst. Elect. Eng., pt. B, vol. 136, no. 2, pp. 61–68, Mar. 1989.
[5] J. M. Elder, J. T. Boys, and J. L. Woodward, “Self-excited induction machine as a small low-cost generator,” Proc. Inst. Elect. Eng., pt. C, vol. 131, no. 2, pp. 33–41, Mar. 1984.
[6] M. H. Salama and P. G. Holmes, “Transient and steady-state load performance of stand-alone self-excited induction generator,” Proc. IEE—Elect. Power Applicat. vol. 143, no. 1, pp. 50–58, Jan. 1996.
[7] Dawit Seyoum Colin Grantham, and Muhammed Fazlur Rahman, “The dynamic characteristics of an isolated self-excited induction generator driven by a wind turbine”, Trans. Inst. Elect. Eng., Industry Applicat, vol. 39, no. 4, July/August 2003
[8] Rohin M Hilloowala and Adel M Sharaf, “A Rule-based fuzzy controller for a PWM inverter in a stand-alone wind energy conversion scheme”, Trans. Inst Elect. Eng. Industry Applicat, vol. 32, no. 1, pp. 57-65, Jan/Feb 1996.
POWER TRANSMISSION THROUGH FRACTIONAL FREQUENCY SYSTEM
ABSTRACT
The fractional frequency transmission system FFTS is a very promising long -distance transmission approach, which uses lower frequency (50/3) to reduce the electrical length of the AC power line, and thus its transmission capacity can be increased several fold . This paper introduces the primary experiments results of FFTS.
The experiment uses the phase-controlled cycloconverter as the frequency changer, stepping up 50/3Hz electricity to 50Hz electricity and supplying it to the utility grid. Thus, a new flexible ac transmission system device is successfully established in this experiment. The synchronizing process of 50/3Hz transmission system with 50Hz utility system is introduced in this paper. The experiment results show that a 1200km/500kv transmission line can transmit more than 2000MW electric power when employing the FFTS. The experiment also illustrates that there is no essential difficulty to realize FFTS in engineering practice
INTRODUCTION
Increasing transmission distance and capacity is always the motivation to advance power industry technologies. In the history of the ac transmission system, increasing distance and capacity mainly depends on raising voltage of transmission lines. At present, the highest voltage level of the AC power transmission line in operation is 750kv.To further up grade, the voltage level encounters difficulties of material and environment issues.
The high voltage direct current ( HVDC) transmission that has no stability limit program once become another approach to increasing electricity transmission capacity.
However, the current converters at two ends of HVDC are very expensive. In addition, up to now, the HVDC practices have been limited to the point to point transmission. It is still difficult to operate a multiterminal HVDC system.
The flexible AC transmission system (FACTS) has been used to improve power system performance and has become a research field. The FACTS exploits power electronic techniques to regulate the parameters of the ac transmission, which can raise transmission capacity to some degree.
This paper introduces the experimental installation of FFTS and primary experiment results. The experiment uses the phase controlled cycloconverter as the frequency changer, stepping up 50/3Hz electricity to 50Hz and supplying it to the utility grid. Thus, a new FACTS device is successfully established I this experiment. The experiment results show that a 1200km/500kv transmission line can transmit more than 200MW electric power when employing FFTS. The experiment also illustrates that there is no essential difficulty to realize FFTS in engineering practice.
The structure of this paper is as follows. The next section briefly introduces the principle of FACTS. In that section discusses the components of the experimental FFTS.
This approach can multiply increase transmission capacity and remarkably improve operating performance. The feasibility arid efficiency of the FFTS is investigated in this paper by the computer simulation method.
The fractional frequency transmission system (FFTS) is a very promising long distance transmission approach, which uses lower frequency(50/3Hz) to reduce the electrical length of the AC power line, and thus, its transmission capacity can be increased several fold. This paper introduces the experimental installation of FFTS and primary experiment results. The experiment uses the phase-controlled cycloconverter as the frequency changer, stepping up 50/3Hz electricity to 50 Hz electricity.
Thus, a new flexible ac transmission system device is successfully established in this experiment. The synchronizing process of 50/3Hz transmission system with 50Hz utility system is introduced in this paper, the experiment also illustrates that there is no essential difficulty to realize FFTS in engineering practice.
2. Flexible AC Transmission systems (facts)
The need for more efficient electricity systems management has given rise to innovative technologies in power generation and transmission. The combined cycle power station is good example of a new development in power generation and flexible AC transmission system FACTS as they are generally known, are new devices that improve transmission systems. Worldwide transmission systems are undergoing continuous changes and restructuring .They are becoming more heavily loaded and are being operated in ways not originally envisioned .Transmission systems must be flexible to react to more diverse generation and load patterns. In addition the economical utilization of transmission system assets is of vital importance to enable utilities in industrialized countries to remain competitive and to survive.
In developing countries, the optimized use of transmission systems investments is also important to support industry, create employment and utilize efficiency scarce economics resources. Flexible AC transmission systems (FACTS) are a technology that responds to these needs. It significantly alters the way transmission systems are developed and controlled together with improvements in asset utilization, system flexibility and system performance
FRACTIONAL FREQUENCY TRANSMISSION SYSTEM (FFTS)
The fractional frequency transmission system FFTS is a very promising long distance transmission approach, which uses lower frequency (50/3) to reduce the electrical length of the AC power line, and thus its transmission capacity can be increased several fold
The AC electricity supplied by utilities has two basic parameters: voltage and frequency. After the transformer was invented, different voltage levels could be used flexibly in generating, transmitting, and consuming electricity to guarantee efficiency for different segments of the power systems. In the history of electrical transmission, besides of 50_60 Hz, many frequencies were used, such as 25, 50/3and 133 Hz. In 1896, the first two generators and the transmission line from Niagara to Buffalo, NY were put into the operation A 25 Hz electric system had been chosen as the winning design
How ever, since 50_60Hz was selected as the standard, changing frequency apparently become taboo. The reason for this might consist in that o transform frequency is more difficult than to transform voltage. As new materials and power electronic techniques continuously advance, different kinds of large_ frequency changers are developed rapidly.
This trend may possibly lead to more reasonably selecting different frequencies for electricity transmission and utilization. For instance, the low frequency electricity can be used to transmit large power for longer distance, and the high frequency electricity can be used more efficiently to drive electric tools
Generally speaking, there are three factors limiting transmission capability, i.e., the thermal limit, stability limit, and voltage drop limit. For the long-distance ac transmission, the thermal limitation is not a significant impediment. Its load ability mainly depends on the stability limit and voltage drop limit [6]. The stability limit of an ac transmission line can be approximately evaluated by
Pmax=V2/X
Where V is the normal voltage, and X is the Reactance of the transmission line.
We can see from the above equation that transmission capacity is proportional to the square of the normal voltage and inversely proportional to the reactance of the transmission line. The voltage drop ∆v% can be evaluated by
∆v%=QX/V2 *100
Where Q is the reactive power flow of transmission line. Thus, the voltage drop is inversely proportional to the square of voltage and proportional to the reactance of the transmission line. Therefore, in order to raise transmission capability, we can either increase the voltage level or decrease the reactance of the transmission line. The reactance is proportional to power frequency f.
X=2πfl
Where L is the total inductance of the transmission line. Hence, decreasing the electricity frequency can proportionally increase transmission capability. The FFTS uses fractional frequency to reduce the reactance of the transmission system; thus, its transmission capacity can be increased several fold. For instance, when frequency is 50/3Hz, the theoretically transmission capability can be raised three times.
.
The principle of FFTS can also view from another perspective. It is well known that the velocity of electricity transmission is approximately equal to the light velocity, 300000 km/s. When electricity frequency is 50Hz, the wavelength is 6000km; for 50/3Hz, the wave length enlarges to 18000km. Thus, when frequency is 50 Hz, a transmission line of 1200km corresponds to one fifteenth of the wave length. Therefore, the “ELECTRICAL LENGTH” decrease to one third. This is the essential reason why the FFTS can increase transmission capability several fold and remarkably improve is performance.
The basic structure of FFTS is illustrated in fig (). The hydropower generator in the figure generates ac power of fractional frequency (say 50/3 Hz), which is then stepped up by a transformer and transmitted to the receiving end of the transmission line where the fractional frequency ac power is stepped up to the industrial frequency.
The hydropower generator can easily generate low-frequency electric power because its rotating speed is usually very low. To generate low- frequency power, the only change for the generator is to reduce its pole number .this change has little influence on cost and efficiency of the hydropower unit. For the transformer, since the electric power that has to be stepped up is of low frequency, the core section area and the coil turn number must be increased.
Therefore, the cost of the transformer in FFTS is higher than that of the conventional transformer. The conventional transmission line can be used in FFTS Without any change.
The frequency changer is the key equipment in FFTS, which can be either the saturable transformer [7] or the power electronics ac-ac frequency changer, such as the cycloconverter [8]. The ferromagnetic frequency changer has advantages of simpler structure, lower cost, and more reliable operation, while the electronic type is superior in higher efficiency and more flexible in installation.
CYCLOCONVERTER
In industrial applications, two forms of electrical energy are used: direct current (dc) and alternating current (ac). Usually constant voltage constant frequency single phase or three-phase ac is readily available. However, for different applications, different forms, magnitude and/or frequencies are required. There are four different conversions between dc and ac power sources.
These conversions are done by circuits called power converters. The converters are classified as:
1-rectifiers: from single – phase or three-phase ac to variable voltage dc
2-choppers: from DC to variable voltage DC
3-inverters: from DC to variable magnitude and variable frequency, single phase or three phase AC.
4-cycloconverter: from single-phase or three-phase ac to variable magnitude and variable frequency, single-phase or three- phase
Traditionally, ac-ac conversion using semiconductor switches is done in two different ways: 1-in two stages (ac-dc and then dc-ac)as in dc link converters or 2-in one stage (ac-ac) cycloconverters . Cycloconverters are used in high power applications driving induction and synchronous motors. They are usually phase-controlled and they traditionally use to their ease of phase commutation.
There are other newer forms of cycloconversion such as ac-ac matrix converters and high frequency AC-AC (HFAC-AC) converters and these use self-controlled switches. These converters, however, are not popular yet.
Simulation and analysis
Operation of cycloconverter
The three phase to three phase cycloconverter is used to convert the input source frequency 50/3 to the required output frequency50 Hz. The operation of phase A is given below
Phase A operation
The output frequency is three time higher than that of input frequency. The model input and output waveforms are given below. From the figure time axis is separated by electrical degrees with respect to input voltage
1. 0o< ωt<30o the most positive phase in C and most negative phase in B. Hence enter switch is 3 and leaving switch is 5.
2. 0o< ωt<60o The most positive phase in A and most negative phase in B. hence the entering switch is 1 and leaving switch is 5.
3. 0o< ωt<90o In this period the output voltage is in negative cycle .so the download switch is operated to get the negative cycle. The most positive phase is A and the most negative phase is B. hence the entering switch is 10 and leaving switch is 9.
4. 0o< ωt<120o the most positive phase is A and most negative phase is C .hence the entering switch is 10 and leaving switch is 9.
5. 120o< ωt<150o during this period the output voltage is again in positive cycle. So the upward switch is to be operated to get the positive cycle at the load. The most positive phase is A and the most negative phase is C .hence the entering switch is 1 and the leaving switch is 6.
6.150o< ωt<180o: The most positive phase is B and the most negative phase is C.
Hence the entering switch is 2and leaving switch is 6.
* 180< ωt<210: The most positive phase is B and the most negative phase isC.
Hence the entering switch is 11 and leaving switch is 9.
* 210< ωt<240: The most positive phase is B and the most negative phase isA.
Hence the entering switch is 11 and leaving switch is 7.
* 240< ωt<270: The most positive phase is B and the most negative phase is A. Hence the entering switch is 2 and leaving switch is 4.
* 270< ωt<300: The most positive phase is C and the most negative phase is A. Hence the entering switch is 3 and leaving switch is 4.
* 300< ωt<330: The most positive phase is C and the most negative phase is A.
Hence the entering switch is 12 and leaving switch is 7.
* 330< ωt<360: The most positive phase is C and the most negative phase is B.
Hence the entering switch is 12 and leaving switch is 8.
Similarly the other phases B,C are switched upward/downward Corresponding to the positive and negative cycles.
VERIFICATION MODEL OF FFTS
The verification model is used to check the frequency and corresponding voltage levels, verification model also indicates the phase system. Thus the real power flow through the transmission line can be measured by knowing the voltage and current parameters of the three phase system. The model is designed for 1200km and the voltage of the system is varied in steps.
The value of inductance, capacitance and resistance is chosen from the default values of the distributed line parameters, the value of neutral resistance is 100 ohm; the 3 phase supply is taken from standard AC source available in the SIM POWER SYSTEM.
The verification model is just used to note the drop across the transmission line of 1200kms and to note the voltage level at different frequencies. The model employees voltmeters and multimeters and output scope from the inbuilt tool box in MATLAB6.0
The AC voltage source is kept at 500kv and the phase sequence is established as R phase 0 degree, Y phase 120 phase shift &B phase- 120 degree phase shift .it can also be noted that the three phase source is common grounded.
VOLTAGE LEVEL Vph(V) for 50/3 Hz Iph(A) for
50/3 Hz Vph(V) for
50 Hz Iph(A) for
50Hz
1.1 kv 1000 10 968.32 9.68
3.3 kv 3149.28 31.49 2915 27.93
6.6 kv 6336.795 63.36 5888 58.9
11 kv 10571 105.71 9813.2 98.2
CONCLUSION
We proposed FFTS, in this he main idea of which is multiplying raising transmission capacity by reducing power frequency. The experiment employs the cycloconverter as the frequency changer to step 50/3 Hz power to 50Hz power and then supply it to the utility grid. Thus a new FACTS device is successfully established. The result of the experiment demonstrates that a 1200km/500kv transmission line can transmit electrical power to by using FFTS. Comparing with 50 Hz AC transmission line, the transmission capability increases by 1.5 times. It demonstrates the great potential of applying this new FACTS device.
Comparing with HVDC, the FFTS can save an electronic converter terminal ,thus reducing investment. In addition, usually HVDC can be used only for point to point transmission, but FFTS can easily form a network-like conventional ac system. Nowadays, it is mature to transform power frequency by the electronic converter(e.g., the cycloconverter). Therefore, FFTS on/ under 750kv can be completed without any special technical difficulty.
The fractional frequency transmission system (FFTS) is a very promising long – distance transmission approach, which uses lower frequency (50/3Hz)to reduce the electrical length of the ac power line, and thus, its transmission capacity can be increased several fold.
This paper introduces the experimental installation of FFTS and primary experiment results .as new materials and power electronic techniques continuously advance different kinds of large-frequency changes are developed rapidly. This trend may possibly lead to more reasonably selecting different frequencies for electricity transmission and utilization. For instance, the lower frequency electricity can be used to transmit larger power for longer distance, and the higher frequency can be used more efficiently to drive the electric tools.
The fractional frequency transmission system FFTS is a very promising long -distance transmission approach, which uses lower frequency (50/3) to reduce the electrical length of the AC power line, and thus its transmission capacity can be increased several fold . This paper introduces the primary experiments results of FFTS.
The experiment uses the phase-controlled cycloconverter as the frequency changer, stepping up 50/3Hz electricity to 50Hz electricity and supplying it to the utility grid. Thus, a new flexible ac transmission system device is successfully established in this experiment. The synchronizing process of 50/3Hz transmission system with 50Hz utility system is introduced in this paper. The experiment results show that a 1200km/500kv transmission line can transmit more than 2000MW electric power when employing the FFTS. The experiment also illustrates that there is no essential difficulty to realize FFTS in engineering practice
INTRODUCTION
Increasing transmission distance and capacity is always the motivation to advance power industry technologies. In the history of the ac transmission system, increasing distance and capacity mainly depends on raising voltage of transmission lines. At present, the highest voltage level of the AC power transmission line in operation is 750kv.To further up grade, the voltage level encounters difficulties of material and environment issues.
The high voltage direct current ( HVDC) transmission that has no stability limit program once become another approach to increasing electricity transmission capacity.
However, the current converters at two ends of HVDC are very expensive. In addition, up to now, the HVDC practices have been limited to the point to point transmission. It is still difficult to operate a multiterminal HVDC system.
The flexible AC transmission system (FACTS) has been used to improve power system performance and has become a research field. The FACTS exploits power electronic techniques to regulate the parameters of the ac transmission, which can raise transmission capacity to some degree.
This paper introduces the experimental installation of FFTS and primary experiment results. The experiment uses the phase controlled cycloconverter as the frequency changer, stepping up 50/3Hz electricity to 50Hz and supplying it to the utility grid. Thus, a new FACTS device is successfully established I this experiment. The experiment results show that a 1200km/500kv transmission line can transmit more than 200MW electric power when employing FFTS. The experiment also illustrates that there is no essential difficulty to realize FFTS in engineering practice.
The structure of this paper is as follows. The next section briefly introduces the principle of FACTS. In that section discusses the components of the experimental FFTS.
This approach can multiply increase transmission capacity and remarkably improve operating performance. The feasibility arid efficiency of the FFTS is investigated in this paper by the computer simulation method.
The fractional frequency transmission system (FFTS) is a very promising long distance transmission approach, which uses lower frequency(50/3Hz) to reduce the electrical length of the AC power line, and thus, its transmission capacity can be increased several fold. This paper introduces the experimental installation of FFTS and primary experiment results. The experiment uses the phase-controlled cycloconverter as the frequency changer, stepping up 50/3Hz electricity to 50 Hz electricity.
Thus, a new flexible ac transmission system device is successfully established in this experiment. The synchronizing process of 50/3Hz transmission system with 50Hz utility system is introduced in this paper, the experiment also illustrates that there is no essential difficulty to realize FFTS in engineering practice.
2. Flexible AC Transmission systems (facts)
The need for more efficient electricity systems management has given rise to innovative technologies in power generation and transmission. The combined cycle power station is good example of a new development in power generation and flexible AC transmission system FACTS as they are generally known, are new devices that improve transmission systems. Worldwide transmission systems are undergoing continuous changes and restructuring .They are becoming more heavily loaded and are being operated in ways not originally envisioned .Transmission systems must be flexible to react to more diverse generation and load patterns. In addition the economical utilization of transmission system assets is of vital importance to enable utilities in industrialized countries to remain competitive and to survive.
In developing countries, the optimized use of transmission systems investments is also important to support industry, create employment and utilize efficiency scarce economics resources. Flexible AC transmission systems (FACTS) are a technology that responds to these needs. It significantly alters the way transmission systems are developed and controlled together with improvements in asset utilization, system flexibility and system performance
FRACTIONAL FREQUENCY TRANSMISSION SYSTEM (FFTS)
The fractional frequency transmission system FFTS is a very promising long distance transmission approach, which uses lower frequency (50/3) to reduce the electrical length of the AC power line, and thus its transmission capacity can be increased several fold
The AC electricity supplied by utilities has two basic parameters: voltage and frequency. After the transformer was invented, different voltage levels could be used flexibly in generating, transmitting, and consuming electricity to guarantee efficiency for different segments of the power systems. In the history of electrical transmission, besides of 50_60 Hz, many frequencies were used, such as 25, 50/3and 133 Hz. In 1896, the first two generators and the transmission line from Niagara to Buffalo, NY were put into the operation A 25 Hz electric system had been chosen as the winning design
How ever, since 50_60Hz was selected as the standard, changing frequency apparently become taboo. The reason for this might consist in that o transform frequency is more difficult than to transform voltage. As new materials and power electronic techniques continuously advance, different kinds of large_ frequency changers are developed rapidly.
This trend may possibly lead to more reasonably selecting different frequencies for electricity transmission and utilization. For instance, the low frequency electricity can be used to transmit large power for longer distance, and the high frequency electricity can be used more efficiently to drive electric tools
Generally speaking, there are three factors limiting transmission capability, i.e., the thermal limit, stability limit, and voltage drop limit. For the long-distance ac transmission, the thermal limitation is not a significant impediment. Its load ability mainly depends on the stability limit and voltage drop limit [6]. The stability limit of an ac transmission line can be approximately evaluated by
Pmax=V2/X
Where V is the normal voltage, and X is the Reactance of the transmission line.
We can see from the above equation that transmission capacity is proportional to the square of the normal voltage and inversely proportional to the reactance of the transmission line. The voltage drop ∆v% can be evaluated by
∆v%=QX/V2 *100
Where Q is the reactive power flow of transmission line. Thus, the voltage drop is inversely proportional to the square of voltage and proportional to the reactance of the transmission line. Therefore, in order to raise transmission capability, we can either increase the voltage level or decrease the reactance of the transmission line. The reactance is proportional to power frequency f.
X=2πfl
Where L is the total inductance of the transmission line. Hence, decreasing the electricity frequency can proportionally increase transmission capability. The FFTS uses fractional frequency to reduce the reactance of the transmission system; thus, its transmission capacity can be increased several fold. For instance, when frequency is 50/3Hz, the theoretically transmission capability can be raised three times.
.
The principle of FFTS can also view from another perspective. It is well known that the velocity of electricity transmission is approximately equal to the light velocity, 300000 km/s. When electricity frequency is 50Hz, the wavelength is 6000km; for 50/3Hz, the wave length enlarges to 18000km. Thus, when frequency is 50 Hz, a transmission line of 1200km corresponds to one fifteenth of the wave length. Therefore, the “ELECTRICAL LENGTH” decrease to one third. This is the essential reason why the FFTS can increase transmission capability several fold and remarkably improve is performance.
The basic structure of FFTS is illustrated in fig (). The hydropower generator in the figure generates ac power of fractional frequency (say 50/3 Hz), which is then stepped up by a transformer and transmitted to the receiving end of the transmission line where the fractional frequency ac power is stepped up to the industrial frequency.
The hydropower generator can easily generate low-frequency electric power because its rotating speed is usually very low. To generate low- frequency power, the only change for the generator is to reduce its pole number .this change has little influence on cost and efficiency of the hydropower unit. For the transformer, since the electric power that has to be stepped up is of low frequency, the core section area and the coil turn number must be increased.
Therefore, the cost of the transformer in FFTS is higher than that of the conventional transformer. The conventional transmission line can be used in FFTS Without any change.
The frequency changer is the key equipment in FFTS, which can be either the saturable transformer [7] or the power electronics ac-ac frequency changer, such as the cycloconverter [8]. The ferromagnetic frequency changer has advantages of simpler structure, lower cost, and more reliable operation, while the electronic type is superior in higher efficiency and more flexible in installation.
CYCLOCONVERTER
In industrial applications, two forms of electrical energy are used: direct current (dc) and alternating current (ac). Usually constant voltage constant frequency single phase or three-phase ac is readily available. However, for different applications, different forms, magnitude and/or frequencies are required. There are four different conversions between dc and ac power sources.
These conversions are done by circuits called power converters. The converters are classified as:
1-rectifiers: from single – phase or three-phase ac to variable voltage dc
2-choppers: from DC to variable voltage DC
3-inverters: from DC to variable magnitude and variable frequency, single phase or three phase AC.
4-cycloconverter: from single-phase or three-phase ac to variable magnitude and variable frequency, single-phase or three- phase
Traditionally, ac-ac conversion using semiconductor switches is done in two different ways: 1-in two stages (ac-dc and then dc-ac)as in dc link converters or 2-in one stage (ac-ac) cycloconverters . Cycloconverters are used in high power applications driving induction and synchronous motors. They are usually phase-controlled and they traditionally use to their ease of phase commutation.
There are other newer forms of cycloconversion such as ac-ac matrix converters and high frequency AC-AC (HFAC-AC) converters and these use self-controlled switches. These converters, however, are not popular yet.
Simulation and analysis
Operation of cycloconverter
The three phase to three phase cycloconverter is used to convert the input source frequency 50/3 to the required output frequency50 Hz. The operation of phase A is given below
Phase A operation
The output frequency is three time higher than that of input frequency. The model input and output waveforms are given below. From the figure time axis is separated by electrical degrees with respect to input voltage
1. 0o< ωt<30o the most positive phase in C and most negative phase in B. Hence enter switch is 3 and leaving switch is 5.
2. 0o< ωt<60o The most positive phase in A and most negative phase in B. hence the entering switch is 1 and leaving switch is 5.
3. 0o< ωt<90o In this period the output voltage is in negative cycle .so the download switch is operated to get the negative cycle. The most positive phase is A and the most negative phase is B. hence the entering switch is 10 and leaving switch is 9.
4. 0o< ωt<120o the most positive phase is A and most negative phase is C .hence the entering switch is 10 and leaving switch is 9.
5. 120o< ωt<150o during this period the output voltage is again in positive cycle. So the upward switch is to be operated to get the positive cycle at the load. The most positive phase is A and the most negative phase is C .hence the entering switch is 1 and the leaving switch is 6.
6.150o< ωt<180o: The most positive phase is B and the most negative phase is C.
Hence the entering switch is 2and leaving switch is 6.
* 180< ωt<210: The most positive phase is B and the most negative phase isC.
Hence the entering switch is 11 and leaving switch is 9.
* 210< ωt<240: The most positive phase is B and the most negative phase isA.
Hence the entering switch is 11 and leaving switch is 7.
* 240< ωt<270: The most positive phase is B and the most negative phase is A. Hence the entering switch is 2 and leaving switch is 4.
* 270< ωt<300: The most positive phase is C and the most negative phase is A. Hence the entering switch is 3 and leaving switch is 4.
* 300< ωt<330: The most positive phase is C and the most negative phase is A.
Hence the entering switch is 12 and leaving switch is 7.
* 330< ωt<360: The most positive phase is C and the most negative phase is B.
Hence the entering switch is 12 and leaving switch is 8.
Similarly the other phases B,C are switched upward/downward Corresponding to the positive and negative cycles.
VERIFICATION MODEL OF FFTS
The verification model is used to check the frequency and corresponding voltage levels, verification model also indicates the phase system. Thus the real power flow through the transmission line can be measured by knowing the voltage and current parameters of the three phase system. The model is designed for 1200km and the voltage of the system is varied in steps.
The value of inductance, capacitance and resistance is chosen from the default values of the distributed line parameters, the value of neutral resistance is 100 ohm; the 3 phase supply is taken from standard AC source available in the SIM POWER SYSTEM.
The verification model is just used to note the drop across the transmission line of 1200kms and to note the voltage level at different frequencies. The model employees voltmeters and multimeters and output scope from the inbuilt tool box in MATLAB6.0
The AC voltage source is kept at 500kv and the phase sequence is established as R phase 0 degree, Y phase 120 phase shift &B phase- 120 degree phase shift .it can also be noted that the three phase source is common grounded.
VOLTAGE LEVEL Vph(V) for 50/3 Hz Iph(A) for
50/3 Hz Vph(V) for
50 Hz Iph(A) for
50Hz
1.1 kv 1000 10 968.32 9.68
3.3 kv 3149.28 31.49 2915 27.93
6.6 kv 6336.795 63.36 5888 58.9
11 kv 10571 105.71 9813.2 98.2
CONCLUSION
We proposed FFTS, in this he main idea of which is multiplying raising transmission capacity by reducing power frequency. The experiment employs the cycloconverter as the frequency changer to step 50/3 Hz power to 50Hz power and then supply it to the utility grid. Thus a new FACTS device is successfully established. The result of the experiment demonstrates that a 1200km/500kv transmission line can transmit electrical power to by using FFTS. Comparing with 50 Hz AC transmission line, the transmission capability increases by 1.5 times. It demonstrates the great potential of applying this new FACTS device.
Comparing with HVDC, the FFTS can save an electronic converter terminal ,thus reducing investment. In addition, usually HVDC can be used only for point to point transmission, but FFTS can easily form a network-like conventional ac system. Nowadays, it is mature to transform power frequency by the electronic converter(e.g., the cycloconverter). Therefore, FFTS on/ under 750kv can be completed without any special technical difficulty.
The fractional frequency transmission system (FFTS) is a very promising long – distance transmission approach, which uses lower frequency (50/3Hz)to reduce the electrical length of the ac power line, and thus, its transmission capacity can be increased several fold.
This paper introduces the experimental installation of FFTS and primary experiment results .as new materials and power electronic techniques continuously advance different kinds of large-frequency changes are developed rapidly. This trend may possibly lead to more reasonably selecting different frequencies for electricity transmission and utilization. For instance, the lower frequency electricity can be used to transmit larger power for longer distance, and the higher frequency can be used more efficiently to drive the electric tools.
REAL- TIME ROLE COORDINATION FOR AMBIENT INTELLIGENCE
ABSTRACT
We propose group communication for agent coordination within” active rooms” and other pervasive computing scenarios featuring strict real-time requirements, inherently unreliable communication, and a large but continuously changing set of context-aware autonomous systems. Messages are exchanged over multicast channels, which may remind of chat rooms in which everybody hears everything being said. The issues that have to be faced (e.g., changing users’ preferences and locations; performance constraints; redundancies of sensors and actuators; agents on mobile devices continuously joining and leaving) require the ability of dynamically selecting the “best” agents for providing a service in a given context. Our approach is based on the idea of implicit organization, which refers to the set of all agents willing to play a given role on a given channel. An implicit organization is a special form of team with no explicit formation phase and a single role involved. No middle agent is required. Asset of protocols, designed for unreliable group communication, are used to negotiate a coordination policy, and for team coordination. We sketch a general computational model for an agent participating to an implicit organization.
INTRODUCTION
We use a form of group communication, called channeled multicast for coordinating agents. Channeled multicast often reduces the amount of communication needed when more than two agents are involved in a task, and allows overhearing of the activity of other agents. Overhearing, in turn, enables the collection of contextual information, pro-active assistance , monitoring , even partial recovery of message losses in specific situations, and is exploited by the protocols described in this paper. Our current implementation of channeled multicast is based on IP multicast, thus it features almost instantaneous message distribution on a local area network but suffers from occasional message losses. Objective of this paper is to describe our current work on an agent coordination technique based on channeled multicast. Rather than addressing the well-known problems of task decomposition or sub goal negotiation, we focus on achieving robustness and tolerance to failure in a setting where agents can be redundant, communication is unreliable, hardware can be switched off, and so on. Such an environment can evolve faster than the agents execute their task. We want to avoid centralized or static solutions like mediators, facilitators or brokers; rather, we aim at a fully distributed and flexible system, without necessarily looking for optimality.
IMPLICIT ORGANIZATIONS
In multi-agent systems, we define a role as a communication-based API, or abstract agent interface (AAI), i.e. one or more protocols aimed at obtaining a cohesive set of functions from an agent. An agent may play more than one role, simultaneously or at different times depending on its capabilities and the context. We call implicit organization a set of agents tuned on the same channel to play the same role but different ways and willing to coordinate their actions. This situation is commonly managed by putting a broker or some other form of middle agent supervising the organization. By contrast, our objective is to explore advantages and disadvantages of an approach based on unreliable group communication, in a situation where agents can come and go fairly quickly, their capabilities can change or evolve over time, and it is not necessarily known a-priori which agent can achieve a specific goal without first trying it out. An implicit organization is a special case of team. An implicit organization is in charge of defining its own control policy, which means how a sub-team is formed within the organization in order to a hive a specific goal and how the intentions of this sub-team are established.
ROLE BASED COMMUNICATION
In this initial work, we assume that any request – by which we mean any REQUEST and QUERY generates a commitment by an implicit organization to perform the necessary actions and answer appropriately.Thus, in principle the interactions between commanding agents and implicit organizations are straightforward, and can be summarized in the simple UML sequence diagram below. A generic Requester agent addresses its request to a role R on a channel; the corresponding implicit organization replies appropriately.
Plain Competition
This policy is nothing more than “everybody races against everybody else and the first to finish wins”. It is by far the easiest policy of all: no pre-work nor post-work coordination is required, while the on-going coordination consists in overhearing the reply sent by who finishes first. In summary, the policy works as follows: when a role receives a request, any agent able to react starts working on the job immediately. When an agent finishes, it sends back its results.The other agents overhear this answer and stop working on the same job. Coordinate enters do/notify(Work, ABORT). A race condition is possible: two or more agents finish at the same time and send their answers. This is not a problem since the requester must accept whatever answer comes first and ignore the others.
Simple Collaboration
This policy consists of collaboration among all participants for synthesizing a common answer from the results obtained by each agent independently. This policy does not require any pre-work coordination. The policy works as follows. As in Plain Competition, all agents able to react start working on the job as soon as the request is received. The first agent to finish advertises his results with an INFORM to the role, and moves to the post-work phase. Within a short timeout all other members of the sub team must react by sending, to the role again, either an INFORM with their own results, or an INFORM that says that they are still working followed by an INFORM with the results when finally done. The first agent collects all these messages and synthesizes the common result, in a goal-dependent way
Multicast Contract Net
This policy is a simplification of the well-known Contract Net protocol where the Manager is the agent sending a request to a role, and the award is determined by the bidder themselves, since everybody knows everybody else’s bid. Thus, effectively this policy contemplates coordination only in the pre-work phase, while neither on-going nor post-work are required. This policy has three parameters: the winning criteria (lowest or highest bid), a currency for the bid (which can be any string), and a timeout within which bids must be sent. The policy works as follows. As soon as a request arrives to the role, all participating agents send their bid to the role. Since everybody receives everybody else’s offer, each agent can easily compute which one is the winner. At the expiration of the timeout for the bid, the winning agent declares its victory to the role with an INFORM repeating its successful bid, and starts working. Some degraded cases must be handled. The first case happens when two or more agents send the same winning bid .The second case happens because of a race condition when the timeout for the bid expires, or because of the loss of messages; as a consequence, it may happen that two or more agents believe to be winners. This is solved by an additional, very short wait after declaring victory, during which each agent believing to be the winner listens for contradictory declarations from others. In spite of these precautions, it may happen that two or more agents believe to be the winners and attempt to achieve the goal independently. The winner declaration mechanism, however, reduces its probability to the square of the probability of losing single messages, since at least two consecutive messages (the bid from the real winner and its victory declaration) must be lost by the others.
Master Slave
This policy has many similarities with the Multicast Contract Net; the essential difference is that a master decides which agent is delegated to achieve a goal, rather than a bidding phase. The master is elected by the policy negotiation protocol. Typically, agents that support this policy either propose themselves as masters, or accept any other agent but refuse to be master themselves; this is because the master is necessarily part of any goal-specific sub-team, i.e. it must always be able to react to any request and must have an appropriate logic for the selection of the slave delegated to achieve a goal. The policy works as follows. On reception of a request, all agents of the sub team send an INFORM to the role declaring their availability. The master agent collects all declarations, and issues an INFORM to the role nominating the slave, which acknowledges by repeating the same INFORM. Message loss is recovered by the master handling a simple timeout between its declaration and the reply, and repeating the INFORM if necessary. Of course, there is no way that two agents end up believing to be slaves at the same time.
NEGOTIATING A POLICY
We discuss here how an implicit organization establishes its own coordination policy. The negotiation protocol works in two phases: first, the set of policy instances common to all agents is computed; then, one is selected. The protocol can be restarted at any time.In the normal case, i.e. when the protocol is not restarted mid-way, the total number of messages sent on a channel for a negotiation is equal to the number of agents in the implicit organization multiplied by a tunable parameter plus one. Some simple examples are the following:
name = MulticastContractNet
Param = Currency constraint = (one of Dollar, Euro)
Suggested = (Euro)
Param = Winning Criteria value = lowest
Param = Bid Timeout constraint = (in 100...2000 )
Suggested = (1000)
Name = Master Slave
Param = Master constraint = (not one of Agent1, Agent2)
Suggested = (Agent3, Agent4)
We show in the algorithm below how the negotiation is triggered. As said above, the negotiation is done in two straightforward steps. In the first, a mutual belief about the policy instances common to the entire organization is established. This can be easily achieved by having each agent sending INFORMs to the role on what it is able to support, and intersecting the contents of all received messages. The second steps consists in having one agent – called the oracle – selecting one policy among those common to everybody and notifying the organization of its decision, by means of a simple ASSERT sent to the role. The oracle can be any agent, either a member or external to the organization .
The algorithm presented below:
(1) allows for any external agent (such as a network monitor or an application agent interested in enforcing certain policies) to intervene just after the common policies have been established;
(2) provides a default oracle election mechanism if an external one is not present; and,
(3) handles conflicting oracles by forcing a renegotiation. For the negotiation protocol to work in an unreliable environment, one additional information and a few more messages to those already mentioned are needed.
All messages concerning policy negotiation are marked with a Negotiation Sequence Number (NSN). A NSN is the identifier of a negotiation process, and is unique during the lifetime of an organization. NSNs form an ordered set; an increment(nsn) function returns a NSN that is strictly greater than its input nsn. In our current implementation, a NSN is simply an integer; the first agent joining an organization sets the NSN of the first negotiation to zero. Goal of the NSN is to help in guaranteeing coherence of protocols
Implementing the protocol
This section describes, as pseudo-code, the negotiation algorithm performed by each agent of an implicit organization.A few primitives are used to send messages (INFORM, ASSERT, QUERY) to the role for which the policy is being negotiated; for simplicity, we assume that the beliefs being transmitted or queried are expressed in a Prolog-like language. For readability, the role is never explicitly mentioned in the following, since the algorithm works for a single role at the time. Each agent has three main beliefs related to the policies: the set of policies it supports, the set of policies believed to be common to all agents in the organization, and the current policy, which is the result of the negotiation.The three main beliefs described above and the states of the negotiation are represented by the following variables:
negotiationState: {UNKNOWN, NEGOTIATING, DECIDED};
commonPolicies: set of Policy; NegotiatingA ents: set of Agent_Identifier;
supportedPolicies: set of Policy; currentPolicy: Policy;
myNSN: Negotiation_Sequence_Number;
myself: Agent_Identifier;
When the agent starts playing a role, it needs to discover the current situation of the organization, in particular its current NSN. This is done by sending a query to the role about the current policy. If nothing happens, after a while the agent assumes to be alone, and forces a new negotiation to start; potential message losses are recovered during the rest of the algorithm.
On Start () {
Set negotiationState = UNKNOWN;
Set myNSN = MIN_VALUE;
Set myself = getOwnAgentIdentifier ();
REQUEST (CURRENT_POLICY(?,?));
suspend until timeout;
if ((currentPolicy == nil)
AND (negotiationState == UNKNOWN))
negotiate( increment(myNSN),
supportedPolicies,
set_of (myself) );
}on_Request ( CURRENT_POLICY (input_NSN, input_policy) ) {
if ((input_NSN == ? ) AND (input_policy == ? )
AND (currentPolicy != nil))
INFORM ( CURRENT_POLICY(myNSN,currentPolicy) );
else...... }
The negotiation process mainly consists of an iterative intersection of the policies supported by all agents, which any agent can start by sending an INFORM with its own supported policies and a NSN higher than the one of the last negotiation (see negotiate () Later on). Conversely, if the agent receives an INFORM on the common policies whose NSN is greater than the one known to the agent, it infers that a new negotiation has started, and joins it. The iterative step consists of intersecting the contents of all INFORMs on the common policies that are received during the negotiation. If the resulting common policies set is empty, i.e. no policy can be agreed upon, the agent notifies a failure condition, waits for some time to allow network re-configurations or agents to leave, and then restart the negotiation again.
on_Inform ( COMMON_POLICIES ( input_NSN,
input_policies, input_agents ) ) {
if (input_NSN > myNSN)
negotiate ( input_NSN,
intersect (supportedPolicies, input_policies),
union (input_agents, set_of(myself)) );
else
if ((input_NSN == myNSN) AND
(negotiationState == NEGOTIATING)) {
commonPolicies =
intersect (commonPolicies, input_policies);
negotiatingAgents =
negotiate( increment(myNSN), supportedPolicies,
set_of (myself) ); }}
else
if (negotiationState == DECIDED)
INFORM (CURRENT_POLICY (myNSN,currentPolicy));
else
if
negotiatingAgents));}
A negotiation is normally started by the agent setting the common policies to those it supports, unless it joins a negotiation started by somebody else. No matter the initial parameters, during the first phase of a negotiation the agent informs the channel about the policies it knows as common, then waits for a period, during which it collects INFORMs from the other members of the organization, as described above. This process is repeated for (at most) max_repeats times, to allow recovery of any lost message by having agents repeating more than once what they know about the common policies. Of course, max_repeats depends on the reliability of the transport media in use for the channels: if the reliability is very high, the max_repeats value is low (two or three). The set of negotiating agents, which is not exploited by this algorithm, may be used in future for a more sophisticated recovery . procedure negotiate (negotiation_NSN: NSN, initial_policies: set of Policy, initial agents:
set of Agent_Identifier ) {
set negotiationState = NEGOTIATING;
INFORM (COMMON_POLICIES (myNSN, commonPolicies,
negotiatingAgents));
repeat max_repeats times {
suspend until timeout;
if (currentPolicy != nil)
break; /// out of the ’repeat’ block
INFORM (COMMON_POLICIES (myNSN, commonPolicies,
negotiatingAgents));}
suspend until (timeout OR currentPolicy != nil);
if ((currentPolicy == nil) AND
(myNSN == negotiation_NSN) AND
(LowestId (negotiatingAgents) == myself)) {
set currentPolicy = random_choice(commonPolicies);
ASSERT ( CURRENT_POLICY(myNSN,currentPolicy) );}
When an ASSERT of the current policy is received, the agent does a few checks to detect inconsistencies – for instance, that the NSN does not refer to a different negotiation, or that two independent oracles have not attempted to set the policy. If everything is fine, the Assertion is accepted, causing negotiate () to finish (see above). Otherwise, the assertion is either refused, or triggers a new negotiation.
on_Assert( CURRENT_POLICY ( input_NSN, input_policy ) ) { if (input_NSN == myNSN) {
if (((currentPolicy == nil) AND
commonPolicies.contains(input_policy)) OR
(currentPolicy == input_policy))
set currentPolicy = input_policy;
else
negotiate ( increment(myNSN), supportedPolicies,
set_of(myself));}}
on_PolicyReminder_Timeout() {
INFORM (CURRENT_POLICY (myNSN,currentPolicy));
set policy_reminder_timeout = getPolicyReminder_Timeout();}
on_Inform(CURRENT_POLICY (input_NSN, input_policy)) {
if (input_NSN > myNSN)
negotiate( increment(input_NSN),
supportedPolicies, set_of(myself));
else
if (input_NSN < myNSN) {
if (negotiationState == DECIDED)
INFORM (CURRENT_POLICY (myNSN, currentPolicy));}
else
if (input_policy != currentPolicy)
negotiate( increment(myNSN),
supportedPolicies, set_of(myself));}
Finally, when an agent leaves the channel, it has a social obligation to start a new negotiation process, to allow the others to adapt to the new situation. This is done by triggering the negotiation process with an INFORM about the common policies, with an incremented NSN and the policies set to a special value any which means “anything is acceptable”.
on_Leave() {
increment(myNSN);
INFORM(COMMON_POLICIES(myNSN,set_of("any"),{empty set}));
}
PRACTICAL EXAMPLE
We elaborate an example which have been chosen to show some practical implicit organizations and the usage of the policies discussed.
Collaborative Search Engines
A Citation Finder accepts requests to look for a text in its knowledge base and returns extracts as XML documents. For the sake of illustration, we model searching as an action (e.g., as in scanning a database) rather than a query on the internal beliefs of the agent.
An example of interaction is: REQUEST From: UserAssistant033 Receiver: Citation Finder
Content: find (Michelangelo) DONE To: UserAssistant033 Content:
done ( find (Michelangelo), results (Michelangelo born in Italy ,... ,
... ) )
Typically, different CitationFinders work on different databases. Any coordination policy of those presented above seems to be acceptable for this simple role. Particularly interesting is Simple Collaboration, where an agent, when done with searching, accepts to be the merger of the results; indeed, in this case, merging is just concatenating all results by all agents. Consider, for instance, the situation where CitationFinders are on board of PDAs or notebooks. A user entering a smart office causes its agent to tune into the local channel for its role; consequently, in a typical peer-to-peer fashion, a new user adds her knowledge base to those of the others in the same room. This could be easily exploited to develop a collaborative work (CSCW) system. In this case, collaboration may be enforced by the CSCW agent by acting as the oracle during policy negotiation.
CONCLUSIONS AND FUTURE WORKS
We proposed implicit organizations for the coordination of agents able to play the same role, possibly in different ways, exploiting group communication and overhearing in environments where messages may be occasionally lost and agents can come and go very frequently. We presented a protocol for negotiating a common coordination policy, outlined a general organizational coordination protocol, and discussed an example. In the near future, we will focus on practical experimentation and application to our domain, i.e. multi-media, multi-modal cultural information delivery in smart rooms (“active museums”).
REFERENCES
1. Foundation for Intelligent Physical Agents. FIPA Communicative Act Library Specification. http://www.fipa.org/repository/cas.html.
2. J. Y. Halpern and Y. O. Moses. Knowledge and common knowledge in a distributed environment. Journal of the Association for Computing Machinery, 37:549–587, 1990.
3. A. Kaminka, D. Pynadath, and M. Tambe. Monitoring Teams by Overhearing: A Multi-Agent Plan-Recognition Approach. Journal of Artificial Intelligence Research, 17:83–135, 2002.
4.Anand S. Rao. AgentSpeak(L): BDI Agents speak out in a logical computable language. In MAAMAW’96: 7th European Workshop on Modelling Autonomous Agen ts in a Multi-Agent World, LNAI 1038. Springer-Verlag, January 1996.
5. R. G. Smith. The contract net protocol: High level communication and control in a distributed problem solver. IEEE Transactions on Computers, C-29(12):1104–1113, 1980.
We propose group communication for agent coordination within” active rooms” and other pervasive computing scenarios featuring strict real-time requirements, inherently unreliable communication, and a large but continuously changing set of context-aware autonomous systems. Messages are exchanged over multicast channels, which may remind of chat rooms in which everybody hears everything being said. The issues that have to be faced (e.g., changing users’ preferences and locations; performance constraints; redundancies of sensors and actuators; agents on mobile devices continuously joining and leaving) require the ability of dynamically selecting the “best” agents for providing a service in a given context. Our approach is based on the idea of implicit organization, which refers to the set of all agents willing to play a given role on a given channel. An implicit organization is a special form of team with no explicit formation phase and a single role involved. No middle agent is required. Asset of protocols, designed for unreliable group communication, are used to negotiate a coordination policy, and for team coordination. We sketch a general computational model for an agent participating to an implicit organization.
INTRODUCTION
We use a form of group communication, called channeled multicast for coordinating agents. Channeled multicast often reduces the amount of communication needed when more than two agents are involved in a task, and allows overhearing of the activity of other agents. Overhearing, in turn, enables the collection of contextual information, pro-active assistance , monitoring , even partial recovery of message losses in specific situations, and is exploited by the protocols described in this paper. Our current implementation of channeled multicast is based on IP multicast, thus it features almost instantaneous message distribution on a local area network but suffers from occasional message losses. Objective of this paper is to describe our current work on an agent coordination technique based on channeled multicast. Rather than addressing the well-known problems of task decomposition or sub goal negotiation, we focus on achieving robustness and tolerance to failure in a setting where agents can be redundant, communication is unreliable, hardware can be switched off, and so on. Such an environment can evolve faster than the agents execute their task. We want to avoid centralized or static solutions like mediators, facilitators or brokers; rather, we aim at a fully distributed and flexible system, without necessarily looking for optimality.
IMPLICIT ORGANIZATIONS
In multi-agent systems, we define a role as a communication-based API, or abstract agent interface (AAI), i.e. one or more protocols aimed at obtaining a cohesive set of functions from an agent. An agent may play more than one role, simultaneously or at different times depending on its capabilities and the context. We call implicit organization a set of agents tuned on the same channel to play the same role but different ways and willing to coordinate their actions. This situation is commonly managed by putting a broker or some other form of middle agent supervising the organization. By contrast, our objective is to explore advantages and disadvantages of an approach based on unreliable group communication, in a situation where agents can come and go fairly quickly, their capabilities can change or evolve over time, and it is not necessarily known a-priori which agent can achieve a specific goal without first trying it out. An implicit organization is a special case of team. An implicit organization is in charge of defining its own control policy, which means how a sub-team is formed within the organization in order to a hive a specific goal and how the intentions of this sub-team are established.
ROLE BASED COMMUNICATION
In this initial work, we assume that any request – by which we mean any REQUEST and QUERY generates a commitment by an implicit organization to perform the necessary actions and answer appropriately.Thus, in principle the interactions between commanding agents and implicit organizations are straightforward, and can be summarized in the simple UML sequence diagram below. A generic Requester agent addresses its request to a role R on a channel; the corresponding implicit organization replies appropriately.
Plain Competition
This policy is nothing more than “everybody races against everybody else and the first to finish wins”. It is by far the easiest policy of all: no pre-work nor post-work coordination is required, while the on-going coordination consists in overhearing the reply sent by who finishes first. In summary, the policy works as follows: when a role receives a request, any agent able to react starts working on the job immediately. When an agent finishes, it sends back its results.The other agents overhear this answer and stop working on the same job. Coordinate enters do/notify(Work, ABORT). A race condition is possible: two or more agents finish at the same time and send their answers. This is not a problem since the requester must accept whatever answer comes first and ignore the others.
Simple Collaboration
This policy consists of collaboration among all participants for synthesizing a common answer from the results obtained by each agent independently. This policy does not require any pre-work coordination. The policy works as follows. As in Plain Competition, all agents able to react start working on the job as soon as the request is received. The first agent to finish advertises his results with an INFORM to the role, and moves to the post-work phase. Within a short timeout all other members of the sub team must react by sending, to the role again, either an INFORM with their own results, or an INFORM that says that they are still working followed by an INFORM with the results when finally done. The first agent collects all these messages and synthesizes the common result, in a goal-dependent way
Multicast Contract Net
This policy is a simplification of the well-known Contract Net protocol where the Manager is the agent sending a request to a role, and the award is determined by the bidder themselves, since everybody knows everybody else’s bid. Thus, effectively this policy contemplates coordination only in the pre-work phase, while neither on-going nor post-work are required. This policy has three parameters: the winning criteria (lowest or highest bid), a currency for the bid (which can be any string), and a timeout within which bids must be sent. The policy works as follows. As soon as a request arrives to the role, all participating agents send their bid to the role. Since everybody receives everybody else’s offer, each agent can easily compute which one is the winner. At the expiration of the timeout for the bid, the winning agent declares its victory to the role with an INFORM repeating its successful bid, and starts working. Some degraded cases must be handled. The first case happens when two or more agents send the same winning bid .The second case happens because of a race condition when the timeout for the bid expires, or because of the loss of messages; as a consequence, it may happen that two or more agents believe to be winners. This is solved by an additional, very short wait after declaring victory, during which each agent believing to be the winner listens for contradictory declarations from others. In spite of these precautions, it may happen that two or more agents believe to be the winners and attempt to achieve the goal independently. The winner declaration mechanism, however, reduces its probability to the square of the probability of losing single messages, since at least two consecutive messages (the bid from the real winner and its victory declaration) must be lost by the others.
Master Slave
This policy has many similarities with the Multicast Contract Net; the essential difference is that a master decides which agent is delegated to achieve a goal, rather than a bidding phase. The master is elected by the policy negotiation protocol. Typically, agents that support this policy either propose themselves as masters, or accept any other agent but refuse to be master themselves; this is because the master is necessarily part of any goal-specific sub-team, i.e. it must always be able to react to any request and must have an appropriate logic for the selection of the slave delegated to achieve a goal. The policy works as follows. On reception of a request, all agents of the sub team send an INFORM to the role declaring their availability. The master agent collects all declarations, and issues an INFORM to the role nominating the slave, which acknowledges by repeating the same INFORM. Message loss is recovered by the master handling a simple timeout between its declaration and the reply, and repeating the INFORM if necessary. Of course, there is no way that two agents end up believing to be slaves at the same time.
NEGOTIATING A POLICY
We discuss here how an implicit organization establishes its own coordination policy. The negotiation protocol works in two phases: first, the set of policy instances common to all agents is computed; then, one is selected. The protocol can be restarted at any time.In the normal case, i.e. when the protocol is not restarted mid-way, the total number of messages sent on a channel for a negotiation is equal to the number of agents in the implicit organization multiplied by a tunable parameter plus one. Some simple examples are the following:
name = MulticastContractNet
Param = Currency constraint = (one of Dollar, Euro)
Suggested = (Euro)
Param = Winning Criteria value = lowest
Param = Bid Timeout constraint = (in 100...2000 )
Suggested = (1000)
Name = Master Slave
Param = Master constraint = (not one of Agent1, Agent2)
Suggested = (Agent3, Agent4)
We show in the algorithm below how the negotiation is triggered. As said above, the negotiation is done in two straightforward steps. In the first, a mutual belief about the policy instances common to the entire organization is established. This can be easily achieved by having each agent sending INFORMs to the role on what it is able to support, and intersecting the contents of all received messages. The second steps consists in having one agent – called the oracle – selecting one policy among those common to everybody and notifying the organization of its decision, by means of a simple ASSERT sent to the role. The oracle can be any agent, either a member or external to the organization .
The algorithm presented below:
(1) allows for any external agent (such as a network monitor or an application agent interested in enforcing certain policies) to intervene just after the common policies have been established;
(2) provides a default oracle election mechanism if an external one is not present; and,
(3) handles conflicting oracles by forcing a renegotiation. For the negotiation protocol to work in an unreliable environment, one additional information and a few more messages to those already mentioned are needed.
All messages concerning policy negotiation are marked with a Negotiation Sequence Number (NSN). A NSN is the identifier of a negotiation process, and is unique during the lifetime of an organization. NSNs form an ordered set; an increment(nsn) function returns a NSN that is strictly greater than its input nsn. In our current implementation, a NSN is simply an integer; the first agent joining an organization sets the NSN of the first negotiation to zero. Goal of the NSN is to help in guaranteeing coherence of protocols
Implementing the protocol
This section describes, as pseudo-code, the negotiation algorithm performed by each agent of an implicit organization.A few primitives are used to send messages (INFORM, ASSERT, QUERY) to the role for which the policy is being negotiated; for simplicity, we assume that the beliefs being transmitted or queried are expressed in a Prolog-like language. For readability, the role is never explicitly mentioned in the following, since the algorithm works for a single role at the time. Each agent has three main beliefs related to the policies: the set of policies it supports, the set of policies believed to be common to all agents in the organization, and the current policy, which is the result of the negotiation.The three main beliefs described above and the states of the negotiation are represented by the following variables:
negotiationState: {UNKNOWN, NEGOTIATING, DECIDED};
commonPolicies: set of Policy; NegotiatingA ents: set of Agent_Identifier;
supportedPolicies: set of Policy; currentPolicy: Policy;
myNSN: Negotiation_Sequence_Number;
myself: Agent_Identifier;
When the agent starts playing a role, it needs to discover the current situation of the organization, in particular its current NSN. This is done by sending a query to the role about the current policy. If nothing happens, after a while the agent assumes to be alone, and forces a new negotiation to start; potential message losses are recovered during the rest of the algorithm.
On Start () {
Set negotiationState = UNKNOWN;
Set myNSN = MIN_VALUE;
Set myself = getOwnAgentIdentifier ();
REQUEST (CURRENT_POLICY(?,?));
suspend until timeout;
if ((currentPolicy == nil)
AND (negotiationState == UNKNOWN))
negotiate( increment(myNSN),
supportedPolicies,
set_of (myself) );
}on_Request ( CURRENT_POLICY (input_NSN, input_policy) ) {
if ((input_NSN == ? ) AND (input_policy == ? )
AND (currentPolicy != nil))
INFORM ( CURRENT_POLICY(myNSN,currentPolicy) );
else...... }
The negotiation process mainly consists of an iterative intersection of the policies supported by all agents, which any agent can start by sending an INFORM with its own supported policies and a NSN higher than the one of the last negotiation (see negotiate () Later on). Conversely, if the agent receives an INFORM on the common policies whose NSN is greater than the one known to the agent, it infers that a new negotiation has started, and joins it. The iterative step consists of intersecting the contents of all INFORMs on the common policies that are received during the negotiation. If the resulting common policies set is empty, i.e. no policy can be agreed upon, the agent notifies a failure condition, waits for some time to allow network re-configurations or agents to leave, and then restart the negotiation again.
on_Inform ( COMMON_POLICIES ( input_NSN,
input_policies, input_agents ) ) {
if (input_NSN > myNSN)
negotiate ( input_NSN,
intersect (supportedPolicies, input_policies),
union (input_agents, set_of(myself)) );
else
if ((input_NSN == myNSN) AND
(negotiationState == NEGOTIATING)) {
commonPolicies =
intersect (commonPolicies, input_policies);
negotiatingAgents =
negotiate( increment(myNSN), supportedPolicies,
set_of (myself) ); }}
else
if (negotiationState == DECIDED)
INFORM (CURRENT_POLICY (myNSN,currentPolicy));
else
if
negotiatingAgents));}
A negotiation is normally started by the agent setting the common policies to those it supports, unless it joins a negotiation started by somebody else. No matter the initial parameters, during the first phase of a negotiation the agent informs the channel about the policies it knows as common, then waits for a period, during which it collects INFORMs from the other members of the organization, as described above. This process is repeated for (at most) max_repeats times, to allow recovery of any lost message by having agents repeating more than once what they know about the common policies. Of course, max_repeats depends on the reliability of the transport media in use for the channels: if the reliability is very high, the max_repeats value is low (two or three). The set of negotiating agents, which is not exploited by this algorithm, may be used in future for a more sophisticated recovery . procedure negotiate (negotiation_NSN: NSN, initial_policies: set of Policy, initial agents:
set of Agent_Identifier ) {
set negotiationState = NEGOTIATING;
INFORM (COMMON_POLICIES (myNSN, commonPolicies,
negotiatingAgents));
repeat max_repeats times {
suspend until timeout;
if (currentPolicy != nil)
break; /// out of the ’repeat’ block
INFORM (COMMON_POLICIES (myNSN, commonPolicies,
negotiatingAgents));}
suspend until (timeout OR currentPolicy != nil);
if ((currentPolicy == nil) AND
(myNSN == negotiation_NSN) AND
(LowestId (negotiatingAgents) == myself)) {
set currentPolicy = random_choice(commonPolicies);
ASSERT ( CURRENT_POLICY(myNSN,currentPolicy) );}
When an ASSERT of the current policy is received, the agent does a few checks to detect inconsistencies – for instance, that the NSN does not refer to a different negotiation, or that two independent oracles have not attempted to set the policy. If everything is fine, the Assertion is accepted, causing negotiate () to finish (see above). Otherwise, the assertion is either refused, or triggers a new negotiation.
on_Assert( CURRENT_POLICY ( input_NSN, input_policy ) ) { if (input_NSN == myNSN) {
if (((currentPolicy == nil) AND
commonPolicies.contains(input_policy)) OR
(currentPolicy == input_policy))
set currentPolicy = input_policy;
else
negotiate ( increment(myNSN), supportedPolicies,
set_of(myself));}}
on_PolicyReminder_Timeout() {
INFORM (CURRENT_POLICY (myNSN,currentPolicy));
set policy_reminder_timeout = getPolicyReminder_Timeout();}
on_Inform(CURRENT_POLICY (input_NSN, input_policy)) {
if (input_NSN > myNSN)
negotiate( increment(input_NSN),
supportedPolicies, set_of(myself));
else
if (input_NSN < myNSN) {
if (negotiationState == DECIDED)
INFORM (CURRENT_POLICY (myNSN, currentPolicy));}
else
if (input_policy != currentPolicy)
negotiate( increment(myNSN),
supportedPolicies, set_of(myself));}
Finally, when an agent leaves the channel, it has a social obligation to start a new negotiation process, to allow the others to adapt to the new situation. This is done by triggering the negotiation process with an INFORM about the common policies, with an incremented NSN and the policies set to a special value any which means “anything is acceptable”.
on_Leave() {
increment(myNSN);
INFORM(COMMON_POLICIES(myNSN,set_of("any"),{empty set}));
}
PRACTICAL EXAMPLE
We elaborate an example which have been chosen to show some practical implicit organizations and the usage of the policies discussed.
Collaborative Search Engines
A Citation Finder accepts requests to look for a text in its knowledge base and returns extracts as XML documents. For the sake of illustration, we model searching as an action (e.g., as in scanning a database) rather than a query on the internal beliefs of the agent.
An example of interaction is: REQUEST From: UserAssistant033 Receiver: Citation Finder
Content: find (Michelangelo) DONE To: UserAssistant033 Content:
done ( find (Michelangelo), results (
Typically, different CitationFinders work on different databases. Any coordination policy of those presented above seems to be acceptable for this simple role. Particularly interesting is Simple Collaboration, where an agent, when done with searching, accepts to be the merger of the results; indeed, in this case, merging is just concatenating all results by all agents. Consider, for instance, the situation where CitationFinders are on board of PDAs or notebooks. A user entering a smart office causes its agent to tune into the local channel for its role; consequently, in a typical peer-to-peer fashion, a new user adds her knowledge base to those of the others in the same room. This could be easily exploited to develop a collaborative work (CSCW) system. In this case, collaboration may be enforced by the CSCW agent by acting as the oracle during policy negotiation.
CONCLUSIONS AND FUTURE WORKS
We proposed implicit organizations for the coordination of agents able to play the same role, possibly in different ways, exploiting group communication and overhearing in environments where messages may be occasionally lost and agents can come and go very frequently. We presented a protocol for negotiating a common coordination policy, outlined a general organizational coordination protocol, and discussed an example. In the near future, we will focus on practical experimentation and application to our domain, i.e. multi-media, multi-modal cultural information delivery in smart rooms (“active museums”).
REFERENCES
1. Foundation for Intelligent Physical Agents. FIPA Communicative Act Library Specification. http://www.fipa.org/repository/cas.html.
2. J. Y. Halpern and Y. O. Moses. Knowledge and common knowledge in a distributed environment. Journal of the Association for Computing Machinery, 37:549–587, 1990.
3. A. Kaminka, D. Pynadath, and M. Tambe. Monitoring Teams by Overhearing: A Multi-Agent Plan-Recognition Approach. Journal of Artificial Intelligence Research, 17:83–135, 2002.
4.Anand S. Rao. AgentSpeak(L): BDI Agents speak out in a logical computable language. In MAAMAW’96: 7th European Workshop on Modelling Autonomous Agen ts in a Multi-Agent World, LNAI 1038. Springer-Verlag, January 1996.
5. R. G. Smith. The contract net protocol: High level communication and control in a distributed problem solver. IEEE Transactions on Computers, C-29(12):1104–1113, 1980.
BIO WEAPON
In the progress of human civilization, development in any field may be questioned but development in one field had been prominent that is DESTRUCTION. We killed plants for food, trees for home, animals for cloth and totally destroyed the biosphere by pollution and nuclear weapons. The crown to all this comes the messenger of death, the BIO WEAPON.
“ if you kill one you’re a murderer , if you kill a hundred you’re a terrorist , if you kill a thousand you’re a Dictator, if you kill the whole world you’re a bio weapon”
Bio weapons are chauvinistic measures of malignancy. Vendetta and vengeance spelt voodoo for this malicious life demolishers. Simply Bio weapons mean bio –life and these are weapons against life. Defense for a country in this twentieth century requires to be highly uninterruptable and more technological and even more malicious. In this race of weapon raising the peak of this death race is the invention of bio weapons.
History:
After the world war II , from the Japan incident America learnt how destructive a nuclear bomb could be and how it could spoil its reputation among the so called peace loving nations. Still it did not wish to let loose its hold as a Big Brother, so it thought of a better so called defensive weapon. As the earth loving environmentalists of Americans would not accept the testing or research of bio weapons in their nation the Government shifted its Bio research to Africa where either awareness or the audacity against the Americans. In this political scenario , the American research on bio weapons were taking place peacefully. At this venture the American Scientists were ward off by the retard in their progress leaving their entities and tested living specimens. From their remains rose today’s global terror AIDS . after this incident America continued its research more seriously resulting in many more powerful bio weapons .One among them was Anthrax, which was robbed and spread by the Al-Quaeda.
NUCLEAR Vs BIO
The nuclear weapons restrains for centuries after its bombardment by its property of radiation preventing successful survival of life in its region where as bio weapons leaves no remains from its instant of explosion. Bio weapons make less noise but more destruction . They are better suitable for under cover wars.they simply nullifies our beloved earth.
CONCLUSION:
In this crucial world politics most probably the third world war would be fought with bio weapons then definitely the fourth world war will be fought with stones and sticks.
“ if you kill one you’re a murderer , if you kill a hundred you’re a terrorist , if you kill a thousand you’re a Dictator, if you kill the whole world you’re a bio weapon”
Bio weapons are chauvinistic measures of malignancy. Vendetta and vengeance spelt voodoo for this malicious life demolishers. Simply Bio weapons mean bio –life and these are weapons against life. Defense for a country in this twentieth century requires to be highly uninterruptable and more technological and even more malicious. In this race of weapon raising the peak of this death race is the invention of bio weapons.
History:
After the world war II , from the Japan incident America learnt how destructive a nuclear bomb could be and how it could spoil its reputation among the so called peace loving nations. Still it did not wish to let loose its hold as a Big Brother, so it thought of a better so called defensive weapon. As the earth loving environmentalists of Americans would not accept the testing or research of bio weapons in their nation the Government shifted its Bio research to Africa where either awareness or the audacity against the Americans. In this political scenario , the American research on bio weapons were taking place peacefully. At this venture the American Scientists were ward off by the retard in their progress leaving their entities and tested living specimens. From their remains rose today’s global terror AIDS . after this incident America continued its research more seriously resulting in many more powerful bio weapons .One among them was Anthrax, which was robbed and spread by the Al-Quaeda.
NUCLEAR Vs BIO
The nuclear weapons restrains for centuries after its bombardment by its property of radiation preventing successful survival of life in its region where as bio weapons leaves no remains from its instant of explosion. Bio weapons make less noise but more destruction . They are better suitable for under cover wars.they simply nullifies our beloved earth.
CONCLUSION:
In this crucial world politics most probably the third world war would be fought with bio weapons then definitely the fourth world war will be fought with stones and sticks.
Subscribe to:
Posts (Atom)
