Fuzzy Logic Controld Sssc(Basic Sssc )

8/13/2019 Fuzzy Logic Controld Sssc(Basic Sssc )

1/6

2004 Intemationaf Conference onPower System Technology - POWERCON2006Slngapore,21-24 November 2004

An Intelligent Fuzzy Controlled SSSCto Enhance Power System StabilityK.M. Sze, L.A. Snider,T.S. hung, K.W. Chan

Abstrart-This pap er proposes and experimentallydemons t ra tes two analyses of various novel control strategies forSSSC controllers, including intell igent rulebased controllers(where the power f l ow oscillation stabilizer action can beadjusted on-line using automatic gain scheduling criteria) a n dfuzzy logic controllers. Modern control strategies are found tosignificantly improve the p erform ance of the SSSC. In particular,the proposed fuzzy logic controller uses the steady s tate relationso f the machine rotors angle and electrical power as the inputsignals and does no t requ i re a dynam ic model of the system for asatisfactory control design. It is not sensitive to the variation ofsys tem s t ruc tu re , pa ramete rs and opera t ion po in ts and can beeasily implemented in a large scale nonlinear system. In addition,ano the r novel p roposa l is intell igent error driven integrator forSSSC. The intell igent error driven integrator is based on theconcept of the e rr or excursion plane where the stabilizing actionis scaled by the magn i tude of the power error signal, voltageer ro r s igna l and the reac tance e r r o r s ignal in o rder to ensureand tolerates system parameter variations a5 well as modelinginaccuracies, s ince the control level (Gain of the PI controller) isonly scaled by the inpu t e r ro r s igna l

adequate compensation. The proposed rule based design is robust

Index Terms-- Power system stabili ty and control, FACTS,fuzzy logic control, a daptive contr ol

I. bTRODUCTIONlectric utilities are. operating in an increasinglyE ompetitive environment due to the recent trend of

deregulation. At the same time, economic and environmentalpressures limit their possibilities to M e r expand theirtransmission facilities. It thus becomes increasingly importantto be able to control power f low on individual transmissionlines. In this context, FACTS ( flexible a.c. transmissionsystem ) devices are becoming increasingly used in highvoltage transmission systems. The improvement of voltage

.- .- .Tbis work was supported the Bong Kong Polytechnic University and theHong Kong ResearchGrants Committee.K M. Sze is with the Department of Electrical Engineering, i e Hong

Kong Polytechnic University, Hung Hom, Kowloon , Hong Kong(e-mail: kmsze.ea&polyu.edu.bk).L.A. Snider is with the Departmentof Electrical Engine&g, The HongKong Polyte&uic University, Hung Hom, Kowlmn , Hong Kong(e-mail: [email protected] +T.S.C h u g is with the Department of Electrical Engineering, The HongKong Polytechnic University, Hung Hom Kowloon , Hong Kong(e-mail:eetschun~inet.polyu.edu.hk).K. W. Chan is with the Department of Electrical Engineering, he HongKong Polytechnic University, Hung Hom, Kowloon , Hong Kong(e-mail: [email protected] .

and current limits of power electronics devices has led to fastdevelopment of FACTS devices in the last decade; forexample, the development of high power GTOs led to thedevelopment of STATCOMs which offer significantadvantages over the thyristor-controIIed SVCs in commonuse. In the not too distant future it is likely that high powerIGBTs will make PWM controlled FACTS devices feasiblefor transmission applications. These devices will offerimproved speed of response and cause less waveformdistortion on the network.

The Static Synchronous Series Compensator (SSSC) is animportant FACTS device which can allow rapid andc o n t i n u o u s changes in the transmission line impedance so thatthe active power flow along the compensated transmissionline can be maintained within a specified range under a rangeof operating conditions.

Applications of SSSC t o modem power s y s t e m are lead t omore flexible, secure and economic operation. Previousresearch has shown that while power system stabilizers (PSS)have been able to provide damping of system oscillations insome cases, they have not been universally successful, andFACTS devices are being increasingly used for this purpose.However, FACTS devices equipped with conventionalcontroilers which were designed based on a linear systemmodels cannot provide satisfactory performance over a widerange of operating conditionsand under large disturbances.This paper presents an analysis of novel control strategiesfor SSSC controllers, including intelligent rule-basedcontrollers (where the power flow oscillation stabilizer actioncan be adjusted on-line using automatic gain schedulingcriteria) and uzzy logic controllers. Modem control strategiesare found to significantly improve the performance of theSSSC. In particular, the proposed uzzy logic controller usesthe steady s t a t e relations o f the machine rotors angle andelectrical power as the input signals and does not require adynamic model of he system for a satisfactory control design.It is not sensitive to the variations of system structures,parameters and operating points and can be easilyimplemented in a large scale nonlinear system. Another novelproposal is intelligent error driven integrator for SSSC. Theintelligent error driven integrator is based on the concept ofthe error excursion plane where the stabilizingaction is scaledby the magnitude of the power error signal, voltage errorsignal and the reactance error signal in order to ensureadequate compensation. The proposed rule based design isrobus t and tolerates system parameter variations as well as

0-7803-8610-8/04/ 20.00 2004 IEEE 1183


2/6

modeling inaccuracies, since the control level (Gain of the PIcontroller) s only scaled by the input error signal.

Comparative studies of the proposed control schemes werecarried out in the SimPowerSystems MATLAB) environment.The simulation comprises a detailed (device-level) model ofPWM-based SSSC and the associated control systems,installed in a tie line of an interconnected two-area EHVsystem. The SSSC is mainly applied to control the powerthrough the tie-line.Case studies show that, when equipped with modemcontrol systems, SSSC c n be very effective in maintainingthe transient and OscilIatory stability of a power system byproviding extra damping to power flow oscillations andmachine angle oscillation.

U. BASIC PIUNCIPLEOF THE ssscA schematic diagram of an SSSC is shown in Fig. 1.TheSSSC is connected in series with a simple transmission line

between Bus 1 and Bus 2 whereV, nd V are the sending endvoltage source and receiving end voltage source respectively.The SSSC consists of a three phase 24-pulse full-wave bridgeinverter which is controlled by the SSSC controller to inject athree-phase synchronous voltage in series with the line so asto control the active and reactive power flow through thetransmission line.

vs k '

Bus3

SSSC Cd rr I lFig. 1 A simplemumission system with SSSC

The real and reactive power (PSLQ) flow at the receiving-end c n be expressed as:

and

The SSSC introduces a virtual compensating reactance, ,(both inductive and capacitive) in series with the transmissionline inductive reactance, XL The expression for the powerflow will become:

sins. V 2=- s m 6=7

where, , is the effective reactance of the transmissionline, including the emulated variable reactance insertedthrough the injected voltage source suppliedby the SSSC. Thecompensating reactance X, is defined to be negative when theSSSC is operated in an inductive mode and positive when theSSSC is operative in a capacitive mode. Therefore, bycontrolling the value of q,he power flow of the line can becontrolled [1] 2].

IU. SSSC CONTROLLERv * n

Fig. Simplifiedcontrolblock d i a g of the SSSCFigure 2shows the simplified control block diagram of theSSSC. An instantaneous 3-phase set of line voltages, V1, at

the transmission line is used to calculate the reference angle, 8,which is phase-locked to the red phase of the line voltage,V1R. An instantaneous 3-phase set of line currents, I, ismeasured and the amplitude and the relative angle, Ob of theline current with respect to the phase-locked-loop angle, 0 mecalculated by using Park's Transformation.

In the Constant Impedance Emulation Mode, the calculatedmagnitude of the line current, Iw, i s multiplied by thecompensating reactance reference, X and the result is theinjection voltage magnitude, V*ref. The phase angle of theline current, Bi is then added to the phase angle of the linevoltage, Qv,and the result, is used to calculate 99. Thephase angle, Bq of the injected voltage is either8, + 90 if therequired compensating reactance is inductive or Bir - 90 if therequired compensating reactance is capacitive. Thecompensating reactance demand,X is either negative if theSSSC is emulating an inductive reactance or positive if it isemulating a capacitive reactance.In Constant Voltage Injection Mode, the insertion voltageamplitude demand, Vd, may be specified and the SSSC willinject the desired voltage almost in quadrature with the linecurrent.

1184


3/6

In Constant Power Control Mode, the calculatedinstantaneous power P is compared to the active powerreference, P,f, so as to control the injected voltage magnitudeand phase angle [l] [2]. P x)\hp x1IV. FUZZYLOGIC AMPINGONTROLLERFLDC)

Pig. 3 shows the block diagram of the proposed FuzzyLogic Damping Controller(FLDC). The input signal is the tie-line real power, and the output is the damping signal which isused as the series inverter input controller si al. [3][4]

NL NM NS ZE PS PM PL

W: Washout block, I: Integration

ZEPS

Fig. 3 Fuzzy Logic DampingControtler or the SSSC

~~ ~~ ~~-NL NM N S ZE PS PM PLNM NS ZE PS PS PM PL

The proposed uzzy supplementary controller block isbasically a non-linear PI-type fuzzy logic controller with twoinputs and one output. The real power though the tie-line andits integral are used as the input signals of the FLDC and theoutput of the FLDC is sent to the main controller for voltageinjection into the tie-line. The two gains K, and K aremultiplied by the two inputs individually to make PI K I Pand PI = K2P. The g in K3 or the output signal is alsoapplied to scale the output signal of the defuzzifier as thedamping signal.

Each input, for the uzzy rule base, is divided into sevenlinguistic values or uzzy subsets, which are PL (positivelarge), PM (positive medium), PS (positive small), ZE (zero),NS (negative small), NM (negative medium) and NL(negative large). Fig. 4shows the membership function of theinput signals namely magnitude of the error signal and rate ofchange of the emor signal.

l s1 . 1 a . 1 1 :

centroid de k f ic a t io n is used to compute the output values.TABLE IFUUY ULE-BASE TABLE

NM N L N M N M N M N S Z E P SNS NL NM NS NS ZE PS PM

PM I NS I ZE I PS I PM PM PM PLPL I ZE I PS I PS I PM PL I PL PLI I I I I I

The differentiable Gaussian membership function is used tofuzzify the two input signals. The triangular function is usedto def kif y he output uzzy sets. Five funy sets are defmedfor each of the input signals and seven uzzy sets are deiinedfor the output signal.

v. AN ERRORRTVEN INTEGRATOR FOR STATICSYNCHRONOUS ERIES COMPENSATOR

SSSCs are utilized in applications including current control,damping oscillations, improving transient and dynamicstability, as well as voltage stability. The effectiveness of theSSSC is determined largelyby the control strategy. In order toimprove the static and dynamic performances of the SSSC, anintelligent error driven integrator for Static SynchronousSeries Compensator was developed and is presented in thispaper. The intelligent error driven iategrator is based on theconcept of the error excursion plane where the stabilizingaction is scaled by the magnitude of the power error signal,voltage error signal and the reactance error signal in order toensure adequate compensation. [5] The proposed rule baseddesign is robust and tolerates system parameter variations aswell as modeling inaccuracies, since the control evel (Gain ofthe PI conixoller) is only scaled by the input error signal.

The PI regulator inside the SSSC controller is used tocontrol the output magnitude of the compensated voltage ofthe Static Synchronous Series Compensator. In order toshorten the response time of the SSSC, prevent undesiredoscillation of the output voltage and improve the powersystem stability, an intelligent error driven integrator wasdeveloped and applied to the SSSC controller.

Fig. 4 M e m b d p functionsfor input and output signalsTable 1 iflustrates the fuzzy control rules in the form of a

decision table which can be express as the following equation:IF KI Xl is Ai and KzXf is Bi TH EN output K3Y is Ci

VI. INTELLIGENT ERROR DI IVEN INTEGRATORThe intelligent error driven integrator is based on theconcept of the on-line gain scheduling with adjustment basedon the magnitude of the input error signal [5]. Figure 5showsthe structure and the detailed block diagram o f the intelligenterror driven integrator. The control equations are shown asfollowing:

where Ai and Bi are the input fizzy sets with triangularmembership functions, while Ci is the output fuzzy set withtriangular membership function. These linguistic values aredetermined ftom the results of the simulation. Finally, the

1185


4/6

A symmetrical 3-phase phase-to-ground fault was appliedStage 1 Normal operating state, t = Os to 1sStage 2: During the three-phase fauk, t = Is to 1.12sStage 3: After the three-phase fault and system recoveringThe results of the simulation with (i) no SSSC, (ii) SSSCWith FLDC and iii) SSSC with intelligent error driven

integrator are presented in Fig. 7- 10. Fig. 7and 8 show thecomparisonsof the active and reactive power flowthrough hetie-line with and without SSSCs. The comparisonsof the rotorangle difference and the machine speeds of the generatorswith and without SSSCs are shown in Fig. 9and 10.

Fig. 7and 8 show that the active and reactive power f lowoscillations through the tie-line were successfully damped byinstalling an SSSC into the system. From Fig. 7, we can seethat the SSSC is very effective in damping active power flowoscillations, and both of the FLDC and adaptive controllerprovide substantial improvementof he dynamic performanceo f the SSSC,damping he system oscillations in less than fourseconds. %ere are fewer swings and no non-linearity of theactive power flow through tie-line with FLDC-SSSC whencompared wt the performance of adaptive controller.Furthermore, as c n be seen in Fig. 9, the adaptive controlleris effective in eliminating steady state errors. The SSSC withthe adaptive control system is also effective in improvingangle stability, as can be seen in Figure 9, where after the firstswing; the maximum rotor angle difference between the twoateas is reduced by more than 50%. The FLDC-SSSC is evenmore effective in improving angle stability compared to thatof the Adaptive SSSC. There is less angle oscillationwt theFDLC-SSSC than with the Adaptive SSSC, which is able todamp down the angle and power flow oscillation faster.I 0

atBus 2 as follows:

state, t = 1.2s to 10s

1 :r s p m n a l h u w a w a d a m w

1-l s f4.=P, p,r = (F(4 'z

K , =K J ~where Pe is the input error signal Pref is the reference

power, KI , K2 and K3 are the selected moduIation gain theintelligent integrator, and F(s) is a transfer function which isused to calculate the rate-of-change of the error signal andperform the filter effect of the error si@ by a second orderfilter to avoid undesired hunting, oscillation and instabikity ofthe system due to the unsuitable output of the intelligentintegrator.

At any sampling time instant t:e&) K , ( t )P , , ( t )er

Pur

where Pin (t) i:s the input signal of the integrator at anysampling time (t),

P e : nver l t i

PC n t e g i r t o ie m 1ertcnnionP b

Fig 5 Structureof he intelligent e r m driven lutegator

w. APPLICATIONOF ssscTO E'ROVTN0TkANSIENT STABILITYThe dynamic performances of the SSSC with fuzzy logic

damping controller (FLDC) and intelligent error drivenintegrator were evaluated in a representative two area systemshown in Figure 6.The system consists o f two individualareas interconnected by a 300 km tie-line. During normaloperating conditions, the interconnecting tie-line carries 680w from area 1 o area 2. [3] [4] 161 [7]

Gen Gen

Fig. 6 A represeutaliveinterconnected reas networkAn SSSC i s installed between bus 3 and bus 4 one of the

ends of the interconnected tie-line). In normal operation, theSSSC will inject constant voltage in to the system in order tomaintain constant power flow through the tie-line.

Fig. 7 Simulation results of the active power flow with adaptwe and FLDCSSSC

1186


5/6

Fig. 8 Simulation resufts of the reactivepower flow wt adaptive and FLDCSSSC

Fig. 9 Simulation results of Delta-theta of the two generatorswith adaptiveand FLDC SSSC

.-i c : > , : ; : . : :--;-----iFig. 10 Simulation results of machine speeds of the two generatorswithadaptive and FLLDCSSSC

m CONCLUSIONThis paper presents the application of a Static SynchronousSeries Compensator, SSSC, o improve the transient stability

of power system and the use of the Fuzzy Logic DampingController and adaptive controller to improve the dynamicperformanceof the SSSC.

Different control methods of the SSSC such as constantvoltage injection control, constant impedance emulationcontrol and constant power control modes were described inthe paper. Fuzzy Logic Damping Controller in whichproportional and integral gains are tuned online based onfuzzy inference rules and an adaptive PI controller

incOrporating on-line gain scheduling and adjustment basedon the magnitude of the input error signal were deveIoped soas to improve the dynamic performance of the SSSC,including better damping ability to power flow oscillation andinter-area power flow oscillations.

IX ACKNOWLEDGEMENTThe authors gratefully acknowledge he support of theHongKong Polytechnic Universityand the Hong Kong ResearchGrantsCommittee for their research grant to c rry out thisresearch

X. REFERENCES[l] Schauder, C D ; yugyi, L.; Lund,M.R.: amai, D.M.; ietman T.R.:Torgerson, D.R.; dris, A.; Operation of the Unified Power FlowCOntmUer (UPFC) Under Practical Constraints, IEEE TransactionsonPower Delivery, Vol. 13,No. 2, April 1998, Pages 630 - 639Schauda, C.D.: Seq KK; Static Synchronous Series Compensator: A

Solid-State Approach to the Series Compensat ion of Transmissionlines, Power Engineering Review, IEEE ,Volume: 17 ,Issue:I , January 1997, Pages62 - 62Yanan Zhao and Emm anuel G. Collins, Jr. Fuzy PI Control Designfor nn Industrial Weigh Belt Feeder, IEEE Transactions on FupySystems,Vol. 11, No. 3, June 2003, pages 311 - 3 19T.K Mok Yixin Ni and Felix F.Wu, Design of F u p y DampingController of VPFC through Genetic Mgodm Power EhgineahgSociety SummerMeeting, 2000. IEEE, Volume 3, 16- 20 July 2000,Page1889-1894~01.3Snider L.A. and Dr. A.M. haraf, A IntelligentVoltage COntroUer ForStat ic Var Compensstot-s, Proceedings of the 26th SoutheasternSymposium, 0-22Mar 1994 Electric Power Systems ResearchJournal, Pages 239 249Klein,M.; ogers, G.I.;Kundur, P.; A Fundamental Study of Inter-areaOscillations in Power Systems, EEE T m a c t i o n s on Power System,Vol. 6,No. 3, August 1991, Pages 914 -921Schode, K.; Hasaoovic, A.; Feliachi A , Enhancing Transient Stabilityusing a Fuzzy Conmd Scheme for the Unif ied Power Flow Controller(UPFC),; ircuits and Systems, 2000. Proceedings of the 43rd IEEEMidwest Symposium on,Volume: 3.8-1 1 Aug. 2000, Pages: 1382 - 1385vo1.3

[2]

[3]

[4]

[5]

[6]

171

kI

XI. BIOGRAPHIESRaymond K.M. Sze received his B. Eug HonsinElectrical Engineering fkom The Hong KongPolytechnic University, Hong Kong in 2002.He is curreutly Mphil research student in theDepartment of Electrical Engineering, TheHong Kong Polytechnic University. Hisresearch interestsare in the areas of flexible a.c.Transmission system power system stabilityand secur i ty assessments, real-time powersystem simulation and advanced control system.T.S. Chung is Professor in EE Dept, HongKong Polytechnic University. He obtained hisPh D, M Sc and B Sc degrees in StrathclydeUniversity, London Universify ImperialCollege and Hong Kong Universityrespectively.His research interests are in power systemanalysis, stability, control and AI applications.

1187


6/6

distribution systems, powercoordination.

Laurence A. Snider received his B.Eng degreein Electrical Engineering from the McGillUniversity, Canada in 1966, M. Sc and PbDdegrees in Electrical Engineering from thcUniversity of Birmingham, UK in 1969 and1971 respectively. He is currently a seniorlecturer in the department of the ElectricalEngineering, The Hong Kong PolytechnicUniversity. His general research interests are inthe areas of power system real-time powersystem simulation, application of powerelectronics in power transmission andquality, high voltage engineering and insulation

Kevin K. . han received his B . Sc Hons andPhD degrees in Electronic and ElectricalEngineeringfrom he Univcrsity of Bath, UK in1988 and 1992 respectively. He is currentlylecturer in the Department of ElectricalEngineering, The Hang Kong PolytechnicUniversity. His research interests are in the areasof power system stability and securityassessments, real-time power system simulation,distributionand paral lcl processing.

1188

Fuzzy Logic Controld Sssc(Basic Sssc )

Documents