issn 1751-8687 an approach for coordinated automatic ...eng.uok.ac.ir/bevrani/bevrani/122.pdfnew...

Published in IET Generation, Transmission & DistributionReceived on 28th May 2011Revised on 19th August 2011doi: 10.1049/iet-gtd.2011.0411

ISSN 1751-8687

An approach for coordinated automatic voltageregulator–power system stabiliser designin large-scale interconnected power systemsconsidering wind power penetrationH. Golpı̂ra H. Bevrani A. Hesami NaghshbandyDepartment of Electrical Engineering, University of Kurdistan, Sanandaj, PO Box 416, Kurdistan, IranE-mail: [email protected]

Abstract: A new comprehensive criterion for the coordinated automatic voltage regulator–power system stabiliser (AVR–PSS)design in large-scale power systems is proposed. Then, a control strategy is introduced to make a trade-off between voltageregulation and small signal stability. The proposed control strategy combines switching technique and negative feedback toachieve a robust controller against load/generation disturbances. An adaptive angle-based switching strategy is employedinstead of fixed time-based switching and hence the proposed control methodology takes into account system size and statusmodes to improve the system performance. The control strategy is completely independent of the test case and fault type.Efficiency of the proposed method has been verified on several large-scale systems and is illustrated here on the New York/New England system with and without wind power penetration. The developed control strategy can be considered as a strongtool for the coordinated design of AVR, PSS and static var compensator (SVC) in the presence of wind turbines.

1 Introduction

Excitation control of generators is a fundamental controlaction to maintain power systems stability. A well-designedexcitation control could improve performance of a systemby supporting the voltage, enhancing transient stability anddamping oscillations [1]. It is well known that fast responseautomatic voltage regulators (AVRs) degrade power systemoscillation stability while improving voltage regulation [2].Power system stabilisers (PSSs), as auxiliary controllers,contribute in maintaining power system stability. PSSoutput signal is applied over the set point of excitationcontrol to modulate the field voltage, in such a way thatprovides positive damping for angle swings of generator inpower systems [3–5].

The conventional AVR–PSS synthesis approaches usuallyprovide a sequential design including two separate stages.Firstly, the AVR is designed to meet the specified voltageregulation performance and then the PSS is designed basedon the desired damping characteristic [6]. Sinceimprovement of both stability and voltage regulation areexecuted via a unique control signal, that is field voltage,the successful achievement of both goals is not possible [7].In other words, an enhancement in one may causedeterioration of the other. To overcome this challenge, anintegrated design approach for AVR–PSS design has beendeveloped.

Power systems continuously experience changes inoperating conditions because of variations in generation/

load and a wide range of disturbances. The AVR and PSSshould be coordinated in order to be robust againstperturbations which arise in the specified operating point[1]. Recently, several efforts have been placed to coordinatethe various requirements for stabilisation and voltageregulation within a single control structure [6–15]. Thecoordinated controller design using linear control theoryand model linearisation around operating point has beenpresented in [6, 8–10]. The employed H1 static outputfeedback in [6] reduces the coordination problem to find asimple fixed gain vector. The approach used in [8] involvesuse of desensitised controller for coordination of AVR andPSS. In another attempt, Quinot et al. [9] address practicaldesensitised four-loop regulator in the French power systemto achieve a coordinated AVR–PSS design. Design strategyis based on the linear quadratic Gaussian (LQG) whichprovides a methodology for robust control design. TheLQG-based controllers are dynamic control systems with ahigh-order structure (for large-scale power systems) andhence are not appropriate ones for large-scale powersystems. The coordinated AVR–PSS design is formulatedas an optimisation problem in [10]. The particle swarmoptimisation technique is employed to minimise a dampingindex which is obtained based on the system eigenvaluesfor linearised system. The control strategies based onoptimisation techniques usually provide a satisfactoryperformance for the pre-specified operating conditions andtherefore the robustness feature is in concern. In addition tothe above challenges associated with each algorithm, the

IET Gener. Transm. Distrib., pp. 1–11 1doi: 10.1049/iet-gtd.2011.0411 & The Institution of Engineering and Technology 2011

www.ietdl.org

Autho

r Per

sona

l Cop

y

Autho

r Per

sona

l Cop

y

major drawback of model linearisation-based controltechniques is in degrading of their efficiency in the presenceof large disturbances. Internal model control (IMC) isaddressed in [7] to affect a trade-off between AVR (voltageregulation, i.e. transient stability) and PSS (small signalstability). The IMC is a modern robust control theory, andrelies on the process (model) identification. Complexity ofmodel identification for large-scale power systems limits theapplication of this methodology. A model-predictive control(MPC) strategy is employed in [11] to address acoordinated AVR–PSS design. Time variant and non-linearinherent of power systems degrade performance of theMPC. Moreover, optimum performance of the MPC will beachieved for the process with great time constants. Instability studies with small time horizon phenomena, theMPC cannot be able to exhibit a satisfactory performance.A trade-off between voltage regulation and power systemstability is provided by switching strategy in [12, 13]. Theseresearches guarantee desired performance only for a specificfault on which optimum switching time is obtained. Guoet al. [14] and Yadaiah et al. [15] successfully overcomethe non-robustness feature of switching strategy byweighting the AVR and PSS output signal. Guo et al. [14]proposed a global control scheme to achieve a robustintegrated design approach for AVR and PSS. However, theperformance of the global controller highly depends on thedesigning of two positive constants which is notstraightforward. A fuzzy-based coordinated controller forpower system stability and voltage regulation is proposed in[15]. Designer knowledge about the controlled system is akey feature in all fuzzy-based control designs. The weightedcontrollers are only examined on a single machineconnected to the infinite bus. Since the infinite bus canconsidered as a machine with a huge inertia, hence itsdynamic cannot be affected by the faults. Thereforecoordination problem reduces to tune of AVR and PSSgains for a single machine related to the fixed reference.However, in large-scale interconnected power systems eachgenerator dynamic is severely affected by others and hence,by increasing the system size, the control law alsoincreases. In fact, implementation of weighted controllers inreal interconnected systems is complex and may beimpractical.

In general, lack of a comprehensive dynamic analysis forAVR–PSS (stability-voltage regulation) studies limits therecent research works to apply to a single machineconnected to the infinite bus or linearised model for multi-machine system. Hojo et al. [16] introduces graphicalfrequency deviation against phase difference (Df 2 Dd)plots as an analytical tool for power system dynamic studiesafter generators outage. However, the same inherent ofphase and frequency degrades the performance of thegraphs for power system studies.

Nowadays, increased needs for electrical energy as well asenvironmental concerns and growing attempts to reducedependency on fossil fuel resources have caused manypower system industries to set an ambitious target ofrenewable generation. Wind power is recognised as themost important renewable energy source (RES) because ofits economical and technical prospects. It is shown thatincreasing the penetration level could seriously affect powersystem dynamics [17]. Owing to the worldwide focusing onconnecting major volume of renewable generation sourcesto the networks in the future years, coordinated AVR–PSSdesign problem in the presence of RESs becomes moresignificant than in the past. However, simplified

assumptions in the recent proposed control methodologiesrestrict their applications in the presence of RESs dynamics.

In response to the above crudity, this paper tries to providea new analytical tool for small signal stability–voltageregulation studies. Thereafter, a new control methodology isaddressed for coordinated AVR–PSS design in large-scalepower systems concerning wind power penetration. Theproposed tool employs normalised phase difference againstvoltage deviation plane to design a control strategy that iscompletely independent from fault type and case study. Thecontroller is designed in such a way that all the inherentnon-linearities and uncertainties are considered in thesystem and hence, the introduced controller is reliable. Theproposed controller uses measurable signals and attempts tomake a bridge between the simplicity of control structureand robustness of stability/performance to satisfy thesimultaneous AVR and PSS tasks.

The rest of this paper is organised as follows. Section 2discusses the new criterion and control methodology. InSection 3, simulation results are explained in detail, andfinally Section 4 concludes the paper.

2 Proposed criterion and control strategy

This section describes the proposed criterion and controlstrategy in detail. Phase difference-voltage deviation(Dd 2 DV ) graphs are employed to provide a new usefulanalytical tool for coordination of AVR and PSS systems.The conflict behaviour of these indices makes the proposedgraph as a powerful tool for voltage regulation and smallsignal stability studies. After any load/generationdisturbance, a graph related to each connected generator isplotted in the plane of phase difference against voltagedeviation (Dd 2 DV ), which is defined as follows

Ddi(t) = di(t) − d0i (1)

DVi(t) = Vi(t) − V0i (2)

where d0i and V0i are the initial values of rotor angle andterminal voltage related to generator i, respectively.

In order to obtain a new criterion that is independent fromfault type, parameters normalisation is required. Thisprocedure is done via the following terms

Ddmax(t) = max{|Ddi(t)|} (3)

DVmax(t) = max{|DVi(t)|} (4)

Ddi =Ddi(t)

Ddmax(t)(5)

DVi =DVi(t)

DVmax(t)(6)

Economic reasons and environmental constraints have led thetransmission lines of systems to operate at the maximumtransfer power capacity. Therefore desired operation of apower system after disturbances could be achieved whensystem returns to the planned operating level characterisedby voltage profile, nominal frequency and power flowconfiguration. The mathematical representation of power

2 IET Gener. Transm. Distrib., pp. 1–11

& The Institution of Engineering and Technology 2011 doi: 10.1049/iet-gtd.2011.0411

www.ietdl.org

Autho

r Per

sona

l Cop

y

Autho

r Per

sona

l Cop

y

flow could be represented by

Pij =|V0i + DVi(t)||V0j + DVj(t)|

Xij

sin[(d0i + Ddi(t))

− (d0j + Ddj(t))] (7)

In order to achieve a stable condition and desiredperformance, the following conditions must besimultaneously met:

1. Voltage deviations for each generator converge to zero.2. All rotor angle deviations converge to the same value.

In ideal situation, all connected generators completelysatisfy the above conditions. Therefore each generator willbe fixed at (1, 1) point in the normalised plane of phasedifference against voltage deviation after a fault. However,after a fault in a practical system, the system returns to anew operating point and hence, the voltage and rotor angledeviations of generators differ from each other. Thereforecontrol actions have to be done in such a way that conducteach generator terminal voltage deviations to zero (0 in thenormalised voltage deviation axis) as well as the rotor angledeviations to the same value, that is 1 in the normalisedphase difference axis. In other words, for the stable, secureand reliable system operation, all the generators’ operatingpoints must be located in the minimum distance from (1, 0)point in the normalised (Dd 2 DV ) plane. This conditioncould be mathematically represented by the followingminimisation problem

min{||Ddi + DVi|| − ||(1, 0)||} (8)

or

min{��(Ddi)

2 + (DVi)2

√− 1} (9)

The optimum value can be achieved when derivation of theabove equation becomes zero

2Ddi + 2DVi(d(DVi)/d(Ddi))

2��(Ddi)

2 + (DVi)2

√ = 0 (10)

This equation leads to

2Ddi + 2DVi

d(DVi)

d(Ddi)= 0 (11)

which gives

Ddi = +DVi or |Ddi| = |DVi| = 0.707 (12)

As previously stated, control actions try to conduct all theconnected generators to the pre-fault condition, that is, (1,1) point in the normalised plane. Therefore

||Ddi|| ≥ 0.707 (13)

||DVi|| ≥ 0.707 (14)

In other hand, according to the parameters normalisation, one

can write

||Ddi|| ≤ 1 (15)

||DVi|| ≤ 1 (16)

It is noteworthy that (12) demonstrates the generic desiredborders for each probable condition. The occurred fault in asystem specifies the sign of phase and voltage variationsand thereafter the required borders, that is, sign of the (12),to control the system.

From (13) to (16), the stable region of power system in thenormalised plane of phase difference against voltagedeviation can be represented as shown in Fig. 1. It isassumed that the AVR and PSS have a dead band inresponse, which could be represented in the CDEFC area.For the generators located in the CDEFC area, the AVRand PSS are well tuned. However, for the other regions, theAVR and PSS parameters should be re-tuned to return theoperating point to the CDEFC area. However, afterinstalling of AVR and PSS in a power system, usually theappropriate tunable parameters are their gains. Therefore theintroduced method proposes an algorithm for the re-tuningof AVR and PSS gains. Based on (13) and (14), theoperation region of controlled systems on the proposedgraph could be mathematically explained by

||Ddi + DVi|| ≥ 0.707 (17)

In an interconnected power system, in order to provide therequired power in each area, deviations of voltages fromnominal values must be compensated by the phasedifferences of existing generators. However, increasinggenerators phase differences, that is, decreasing voltagesprofile may lead a system to an unstable mode. In this case,an option static var compensator (SVC) must be employedto avoid power system instability. The SVC is implementedto compensate the required reactive power to avoidundesirable phase differences increment. Therefore (17)reveals that in the presence of only AVR and PSS, agenerator must be located in the minimum distance from

Fig. 1 Stable region of the power system


www.ietdl.org

Autho

r Per

sona

l Cop

y

Autho

r Per

sona

l Cop

y

0.707. This statement could be mathematically explained by

min{||Ddi + DVi|| − 0.707} (18)

Following the same procedure as used for (8), the aboveminimisation problem results in |Ddi| ¼ |DVi| ¼ 0.5.

Therefore the operating region of AVR and PSS can bedescribed as follows

||DVi|| ≥ 0.5 (19)

||Ddi|| ≥ 0.5 (20)

For the remaining regions, in addition to the AVR and PSS,the SVC is also required to enhance the systemperformance. This problem could be clearly seen in thepower systems with wind power penetration. Windgenerators usually employ induction machines rather thansynchronous ones. Therefore the required reactive power forinitialising the induction machines is taken from the gridand hence, the buses voltages are decreased. Fig. 2demonstrates the generalised version of Fig. 1 whichdescribes the operating region of AVR, PSS and SVC.

After description of the required borders to design acoordinated controller, the mechanism on which thevariation of AVR, PSS and SVC gains could improvepower system performance is explained. As alreadymentioned, the controller gains are re-tuned to conduct eachconnected generators to (1, 1) point in the normalised(Dd 2 DV ) plane. Owing to the detrimental impact oftorques in phase with rotor angle deviations on oscillatorystability and also taking into account sub-synchronousresonance (SSR, will be explained in discussion), thecontrol strategy should be considered in such a way that theAVRs always encounter with gains reduction. For agenerator with less voltage deviation, variation of AVR gaincould enhance the system performance. In other words, ifvoltage deviation for a generator is smaller than 0.707,decreasing of AVR gain causes increment of voltagedeviation and hence, the generator is conducted to thedesired region. At below, the mechanism on which the PSSgain affects the power system dynamics is described.

Consider the swing equation as follows

2Hd2u

v0dt2= Pmech − Pelec (21)

where H, v0, u, Pmech, Pelec are inertia constant, synchronisedspeed, rotor angle, mechanical power input, and electricalpower output, respectively.

The reference for the rotor rotation is considered as a framewhich rotates at the synchronised speed. Therefore u could beexplained by

u = vt − v0t + u0 (22)

where u0 is the initial position of the rotor.By replacing (22) in (21), the following equation is

obtained based on the speed variations

2Hd(Dv)

dt= Pmech − Pelec (23)

It is well known that the PSS produces a torque in phase withpositive deviations of speed. For a negative speed deviation,that is Pmech , Pelec, decreasing of PSS gain accelerates therotor and increases the rotor angle deviations. Therefore for agenerator with a normalised phase difference smaller than0.707, the PSS gain should be reduced during the controlprocess. For positive speed deviations, that is, Pmech . Pelec,the PSS gain should be increased to conduct the generatorwith less phase deviations to the desired region. For agenerator with both variations smaller than 0.707, a minorvariation identifies the required control action to improvesystem performance. Regarding the above statements, threecontrol laws could be explained as:

1. every generator located in DCJHID area encounters withdecreasing of PSS gain in the proposed control strategy;2. every generator located in BCJFEB area encounters withdecreasing of AVR gain in the proposed control strategy;3. every generator located in ABCDA area encounters withincreasing of SVC gain in the proposed control strategy.

The proposed control methodology updates the controllergains by adding a correction term to the old ones toimprove the system performance. Later, the details on thecalculation of the controller gains are given.

For a located generator in the SVC reaction region,increasing of SVC gain forces the generator operation tothe reaction region of the AVR and PSS. Thereafter, aprocedure same as steps 1 and 2 enhances the systemperformance. Therefore the proposed criterion could beused for the coordinated design of AVR, PSS and SVC inmulti-machine power systems.

For the implementation of the proposed online controlstrategy, a combination of switching strategy and simplenegative feedback in a discrete manner is employed. Firstly,according to the position of each generator in the normalised(Dd 2 DV ) plane, an appropriate control action is selectedbased on the above control laws. Then, the feedback tries toconduct generators to the desired region. The implementedfeedback employs the angle between phase and voltagedeviations (a) as input signal. Note that the set-point for theemployed feedback is considered as 458, which is an ideal aat the (1, 1) point. Regarding the time horizon of voltageinstability (about 20 s) and transient instability (2 s [18]), theFig. 2 Generalised stable region of the power system



www.ietdl.org

Autho

r Per

sona

l Cop

y

Autho

r Per

sona

l Cop

y

sampling time interval could be selected equal to 2 s. However,the proposed control strategy employs the first sample at asmaller time (less than 2 s) to improve its reliability. Theproposed control mechanism could be explained as follows:

1. Setting the first sample interval (T ¼ Tsamp).2. Position of each generator is determined in the normalisedplane by |DV|, |Dd|, a.3. The desired control action is selected based on thegenerators position and control laws.4. New control parameters (Kpss, KAVR, Ksvc) are calculatedas follows

Knewpss = 1 − tan|(a− 45)|Kold

pss (24)

KnewAVR = 1 − tan|(a− 45)|Kold

AVR (25)

Knewsvc = 1 + tan|(a− 45)|Kold

svc (26)

5. Set T ¼ T + DT, and return to step 2.

where

DT ≤ 2 (27)

Note that the proposed method is applicable for any load/generation changes. However, as already mentioned for thepositive phase deviations, PSS gain must be increasedinstead of decreasing. In other words, for the positive phasedeviations, that is, Dd . 0, (24) will change to

Knewpss = 1 + tan|(a− 45)|Kold

pss (28)

The flowchart representation of the proposed algorithm isdepicted in Fig. 3. Here Kpss, KAVR and Ksvc are PSS, AVRand SVC gains, respectively. In fact, the problem ofcoordinated AVR–PSS design is reduced to design ofsimple switching and feedback methods which are easy forimplementation.

Fig. 3 Flowchart representation of control strategy

Fig. 4 Single line diagram of IEEE 68-bus test system [19]


www.ietdl.org

Autho

r Per

sona

l Cop

y

Autho

r Per

sona

l Cop

y

Fig. 5 Implemented controllers’ structures

a Excitation systemb Single-input PSSc SVC

Fig. 6 System response for the outage of generator 12 in the presence of proposed control strategy

a Time-domain resultsb Dd 2 DV graphsc Distribution of the normalised parameters in the Dd 2 DV plane at 1.5 sd At 20 s



www.ietdl.org

Autho

r Per

sona

l Cop

y

Autho

r Per

sona

l Cop

y

3 Simulation results

The capability of the proposed criterion and control strategyis investigated after application on several large-scaleinterconnected power systems. Here this issue is illustratedusing the New York/New England system, a 16-machine,five-area test system, as shown in Fig. 4. All the generatorsare represented by a two-axis model and equipped withPSS. Structures of the implemented controllers, that is,AVR, PSS and SVC, are also shown in Fig. 5. Efficiency ofthe proposed control strategy is also examined in thepresence of wind turbines (WTs). To investigate the impactof WTs on the coordination problem, wind farms areemployed at buses 9 and 15. The implemented wind farmsemploy simple squirrel cage induction generators withoutany control capability that operate at a substantiallyconstant speed, normally referred to as fixed speedinduction generators (FSIGs).

3.1 Case A

In the performed simulations without wind power penetration,the first sample is used at 1.5 s (Tsamp ¼ 1.5 s). Firstly,behaviour of the power system after outage of generatornumber 12 is investigated. Note that without applying theproposed control strategy the system encounters with

transient instability, that is, first swing instability. Theefficiency of the proposed criterion and control strategy isshown in Fig. 6. Fig. 6a demonstrates the performance ofthe power system after applying the proposed controlstrategy within 4 s in the sake of comparison with occurredfirst swing instability without applying the strategy. It couldbe seen that the proposed control strategy aids the system tokeep its transient stability. Fig. 6b explains the behaviour ofthe system within 30 s in the (Dd 2 DV ) plane. It is clearthat after a number of control reactions, the system isconducted to stable mode at steady state. Distribution of thevoltage–angle indexes in the normalised (Dd 2 DV ) planeis also shown in Fig. 6. It could be seen that generators14–16 are located in the PSS reaction region at the firstcontrol action. Therefore decreasing the PSS gains of thesegenerators besides decreasing the AVR gains for someothers make the system stable. It can be seen that thesystem response is quite improved using the proposedcoordination methodology (Fig. 6d).

As another example, Fig. 7 shows the dynamics of thesystem after outage of generator number 16. It could beseen that the system experiences disorders in the oscillatory,angle and transient stability behaviour during simulationtime. The applied control strategy eliminates the voltage/rotor angle oscillations and conducts the system to a stablemode in steady state. Note that without applying the

Fig. 7 System response for the outage of generator 16 by applying the proposed control strategy

a Time-domain resultsb Dd 2 DV graphsc Distribution of the normalised parameters in the Dd 2 DV plane at 1.5 sd At 95 s


www.ietdl.org

Autho

r Per

sona

l Cop

y

Autho

r Per

sona

l Cop

y

proposed control strategy the system encounters withinstability. Distribution of the voltage–angle indexes relatedto the outage of generator 16 in the normalised (Dd 2 DV )plane is also shown in Fig. 7. At t ¼ 1.5 s (Fig. 7c), exceptfor two generators, others are out of the desired region.According to the assigned control laws, by decreasing thePSS gains for this set of generators at the first controlaction, the system tries to keep its integrity. It could be seenthat all the generators have been located in the desiredregion at steady state (Fig. 7d).

The results related to the outage of loads prove satisfactoryperformance of the control strategy and successfulness of theproposed criterion against load changes as well as generationchanges.

3.2 Case B

Case B is performed to emphasise the efficiency of theproposed method for the power system in the presence ofWTs. In the performed simulations concerning wind powerpenetration Tsamp will be fixed at 1.8 s. Fig. 8 shows thedynamic response of the power system following outage ofG12, concerning 8% wind power penetration. It couldbe seen that by the penetration of wind power, generators14–16 try to be isolated from the system and operatein islanding mode. Therefore the system is going into anunstable mode. The non-similar phase characteristics of

generators beside diverged terminals voltage behaviour inFig. 8b also exhibit instability of the system. Distribution ofthe voltage–angle indexes in the normalised (Dd 2 DV )plane is shown in Fig. 8. It could be seen that generators14–16 are located in the operation region of SVC.Therefore SVC should be implemented in the areas relatedto these generators. However, because of fewer deviationsfor the generator 16 and taking into account the economicalreasons, it seems that SVC in this region could return thesystem to a stable mode. For this propose, at the beginningof simulation, SVC is considered in the model and afterthat, increasing of SVC gain improves the systemperformance. The dynamic behaviour of the power systemafter applying the proposed control strategy is shown inFig. 9. The proposed control strategy conducts the systemto a stable mode within several tuning steps. The initial andfinal values (controller gains at steady state) of controlparameters are given in Table 1. Distribution of thenormalised parameters in the proposed criterion for outageof G12 in the presence of WTs and SVC could be seen inFig. 9. From Fig. 9d, it is quite clear that all the generatorswill be moved to the desired region at steady state.

3.3 Discussion

In the performed simulations, it was seen that in transientperiod (based on the system inertia), interactions between

Fig. 8 System response for the outage of generator 12 in the presence of wind power

a Time-domain resultsb Dd 2 DV graphsc Distribution of the normalised parameters in the Dd 2 DV plane at 1.8 s



www.ietdl.org

Autho

r Per

sona

l Cop

y

Autho

r Per

sona

l Cop

y

AVRs and PSSs gains improve the system performance andthereafter, reduction of AVR gains conducts the system to astable mode. In other words, PSSs contribute to maintain

the transient stability. On the other hand, to avoidoscillatory instability, the torques in phase with rotor angledeviations, that is, AVRs gains, always decrease. Thecontribution of PSS for keeping transient stability is welldiscussed by Kundur [18]. A practical discontinuesexcitation control (DEC) which employs switching strategywas used in Ontario hydro power plant to achieve stabilityand voltage regulation. The DEC addresses a transientstability excitation control (TSEC) to improve stability. TheTSEC is switched off from the system after 2 s (transientperiod) to avoid oscillatory instability. In other words, theDEC employs fixed-time-based switching strategy.However, transient period could be varied from 2 s to about10 s based on the system size and the modes status.Therefore the fixed-time-based switching strategy couldaffect the system performance. The proposed algorithm inthis paper fulfils the above concern by using an adaptiveangle-based switching strategy. Based on the systemoperation, the synchronising and damping torques arechanged to achieve a satisfactory performance. Moreover,decreasing AVR gains in the proposed control methodologyeliminates probable SSR in the system. It is shown thatincreasing AVR gains could encounter system with SSR[20]. The SSR phenomenon, which results by increasingAVR gains, is shown in Fig. 10.

Owing to different kinds of uncertainty and parameterperturbation, load variations and modelling errors,

Fig. 9 System response for the outage of generator 12 following apply the proposed method in the presence of wind power

a Time-domain resultsb Dd 2 DV graphsc Distribution of the normalised parameters in the Dd 2 DV plane at 2 sd At 50 s

Table 1 Initial and final control gains for outage of G12

Generator

number

PSS gains

(Kpss × KWFa)

AVR gains

(KA)

SVC gains

(Ksvc)

Initial Final Initial Final Initial Final

1 100 100 100 100.000 – –

2 100 100 100 81.7805 – –

3 100 100 100 80.2319 – –

4 100 100 100 76.6451 – –

5 100 100 100 75.1585 – –

6 100 100 100 74.4325 – –

7 100 100 100 76.0875 – –

8 100 100 100 82.0945 – –

9 100 100 100 80.1594 – –

10 100 100 100 100.000 – –

11 50 50 100 82.2249 – –

13 110 110 100 81.7274 – –

14 100 75.8369 100 64.7924 – –

15 100 79.9705 100 64.8628 – –

16 100 82.3612 100 81.5471 50 58.3413

aKWF: washout filter gain


www.ietdl.org

Autho

r Per

sona

l Cop

y

Autho

r Per

sona

l Cop

y

considerable efforts have been devoted to the design of robustexcitation control [5, 21]. Jabr et al. [21] introduce a conicprogramming method that moves eigenvalues correspondingto the unstable and poorly damped modes to the left-handside of the s plane. In general, the success of all modellinearisation-based control methodologies severely dependson the operating points. Considering various operatingpoints, that is, robustness issue could degrade the controllerperformance by increasing size of evaluation, control laws,etc. However, the proposed graphical criteria and controlstrategy in the present paper is absolutely independent fromthe operating point, and hence, the robustness featurecompletely and successfully could be met. Moreover, theproposed control strategy relies on the basic power systemequations and therefore it is applicable to any power systemwithout any query. Simulation results prove satisfactoryperformance of the control strategy under contingencies, aswell as normal condition.

Pal and Mei [22] and Mei and Pal [23] show that theassociated controllers with doubly fed induction generators(DFIGs) separate the mechanical dynamics from electricalones, and therefore no electromechanical mode exists. Inother words, DFIGs do not have detrimental impact on thesmall signal stability. Moreover, the transient stability is notin concern considering DFIGs, but also fault performance orride-through of synchronous generator will be even better.Therefore DFIGs could not affect power system dynamicsseriously in such a way that would be an appropriate casestudy to demonstrate the ability of the proposed coordinationstrategy. However, as previously stated FSIGs directly affectpower system dynamics by interchanging reactive powerwith the connected system and hence significantly influencethe controller performance. Implementation of auxiliarycontroller such as SVC is mandatory to keep the powersystem integrity as addressed in this paper.

Simplicity of analysis using the developed graphical tooland model independency of the proposed coordinatedcontrol strategy can be considered as main advantages ofthe present work.

4 Conclusions

Power system stability and voltage regulation have beenconsidered as important control issues for secure system

operation over many years. Currently, because of expandingphysical set-ups, functionality, complexity of power systemsand penetration of RESs, the mentioned problems becomemore significant than in the past. Moreover, because ofconflict behaviour between stability and voltage regulationthe successful achievement of both goals using independentcontrol strategies is difficult.

In order to achieve stability and voltage regulationsimultaneously, a new criterion in the normalised phasedifference against voltage deviation plane is developed.Based on the introduced criterion, an adaptive angle-basedswitching strategy and negative feedback are combined toobtain a robust control methodology against load/generationdisturbances. The proposed method was applied to the well-known IEEE 68-bus test system with/without wind powerpenetration. The obtained results based on the proposedcontrol strategy exhibit satisfactory performance over awide range of operating conditions.

In all the investigations the long-term time-domainsimulation in the proposed graphical space is considered asa criterion for stability assessment. Considering all modeluncertainties, non-linearities and saturations besidesimplicity and well response of the proposed controlstrategy makes it reliable for real large-scale power systems.In addition to the robustness performance of the controller,it also allows direct effective trade-off between voltageregulation and damping performance.

5 References

1 Boules, H., Peres, S., Margotin, T., Houry, M.P.: ‘Analysis and design ofa robust coordinated AVR/PSS’, IEEE Trans. Power Syst., 1998, 13,pp. 568–575

2 Dudgeon, G.J.W., Leithead, W.E., Dysko, A., O’Reilly, J., McDonald,J.R.: ‘The effective role of AVR and PSS in power systems:frequency response analysis’, IEEE Trans. Power Syst., 2007, 22,pp. 1986–1994

3 Gupta, R., Bandyopadhyay, B., Kulkarni, A.M.: ‘Design of powersystem stabiliser for single-machine system using robust periodicoutput feedback controller’, IEE Proc. Gener. Transm. Distrib., 2003,150, pp. 211–216

4 Bayne, J.P., Lee, D.C., Watson, W.: ‘A power system stabilizers forthermal units based on derivation of accelerating power’, IEEE Trans.Power Appar. Syst., 1977, PAS-96, pp. 1177–1183

5 Jabr, R.A., Pal, B.C., Martins, N., Ferraz, J.C.R.: ‘Robust and coordinatedtuning of power system stabiliser gains using sequential linearprogramming’, IET Gener. Transm. Distrib., 2010, 4, pp. 893–904

Fig. 10 SSR in power system

a Time-domain resultsb Dd 2 DV graphs



www.ietdl.org

Autho

r Per

sona

l Cop

y

Autho

r Per

sona

l Cop

y

6 Bevrani, H., Hiyama, T., Mitani, Y.: ‘Power system dynamicstability and voltage regulation enhancement using an optimal gainvector’, Control Eng. Pract., 2008, 16, pp. 1109–1119

7 Law, K.T., Hill, D.J., Godfrey, N.R.: ‘Robust co-ordinated AVR–PSSdesign’, IEEE Trans. Power Syst., 1994, 9, pp. 1218–1225

8 Heniche, A., Bourics, H., Houry, M.P.: ‘A desensitized controller forvoltage regulation of power systems’, IEEE Trans. Power Syst., 1995,10, pp. 1461–1466

9 Quinot, H., Boules, H., Margotin, T.: ‘Robust coordinated AVR + PSSfor damping large scale power system’, IEEE Trans. Power Syst., 1999,14, pp. 1446–1451

10 El-Zonkoly, A.M.: ‘Optimal tuning of power systems stabilizers andAVR gains using particle swarm optimization’, Expert Syst. Appl.,2006, 31, pp. 551–557

11 Saidy, M.: ‘A unified approach to voltage regulator and power systemstabilizer design based on predictive control in analogue form’, Electr.Power Energy Syst., 1997, 19, pp. 103–109

12 Wang, Y., Hill, D.J.: ‘Robust nonlinear coordinated control of powersystems’, Automatica, 1996, 32, pp. 611–618

13 Wang, Y., Hill, D.J., Middleton, R.H., Gao, L.: ‘Transient stabilityenhancement and voltage regulation of power systems’, IEEE Trans.Power Syst., 1993, 8, pp. 620–627

14 Guo, Y., Hill, D.J., Wang, Y.: ‘Global transient stability and voltageregulation for power systems’, IEEE Trans. Power Syst., 2001, 16,pp. 678–688

15 Yadaiah, N., Kumar, A.G.D., Bhattacharya, J.L.: ‘Fuzzy basedcoordinated controller for power system stability and voltageregulation’, Electr. Power Syst. Res., 2004, 69, pp. 169–177

16 Hojo, M., Abe, K., Mitani, Y., Ukai, H., Saeki, O.: ‘Online powersystem monitoring by synchronized voltage phasor measurement atdemand sides’. Technical Meeting on Power Systems Engineering(PSE), January 2010, pp. 19–23

17 Gautam, D., Vittal, V., Harbour, T.: ‘Impact of increased penetration ofDFIG-based wind turbine generators on transient and small signal stabilityof power system’, IEEE Trans. Power Syst., 2009, 24, pp. 1426–1434

18 Kundur, P.: ‘Power system stability and control’ (McGraw-Hill,New York, 1994)

19 Rogers, G.: ‘Power system oscillations’ (Kluwer Academic Press, Boston,1999)

20 Widyan, M.S.: ‘On the effect of AVR gain on bifurcations of sub-synchronous resonance in power systems’, Electr. Power Energy Syst.,2010, 32, pp. 656–663

21 Jabr, R.A., Pal, B.C., Martins, N.: ‘A sequential conic programmingapproach for the coordinated and robust design of power systemstabilizers’, IEEE Trans. Power Syst., 2010, 25, pp. 1627–1637

22 Pal, B.C., Mei, F.: ‘Modelling adequacy of the doubly fed inductiongenerator for small-signal stability studies in power systems’, IETRenew. Power Gener., 2008, 2, pp. 181–190

23 Mei, F., Pal, B.C.: ‘Modelling of doubly-fed induction generator forpower system stability study’. General Meeting of the IEEE-Power-and-Energy-Society, 2008, pp. 287–294


www.ietdl.org

Autho

r Per

sona

l Cop

y

Autho

r Per

sona

l Cop

y

issn 1751-8687 an approach for coordinated automatic ...eng.uok.ac.ir/bevrani/bevrani/122.pdfnew...

Documents