A Gain-Scheduled Approach ToTheTranscient Stabilization OfPowerSystems
Kang-Zhi Liul,Shengwei Mei,Xiaoyang Gui
2006-01-01
Abstract:Inthis paper, anewapproach isproposed forthetransient stabilization control ofasingle-machine infinite-bus powersystem. Theproposed method isbased ontheconcept oflinear parameter varying (LPV)modelandlinear matrix inequality(LMI) basedgain-scheduling control. Itiswell-known that whenthedisturbances arelarge, thenonlinearity inherent inthepowersystem cannolonger beignored andtheperformance ofclosed-loop system designed based onlinear approximation degrades severely insuchcase.Here, wepropose acontrol methodwhichhandles thenonlinear model directly andisable toassign theconvergence rate. First, weshowthat thenonlinear modelcanbetransformed into an LPVsystem withtherotor angle asthetime-varying parameter. Then,again-scheduled controller isdesigned byusing thegain-scheduling control method. Inparticular, theinput saturation istaken into consideration inthedesign. Simulation results verify that theproposed method iseffective. Asiswell-known, a power systemisalarge scale sys- temandhasstrong nonlinearity. Thestability ofthepower systemhasattracted alotofattention andhasbeenstud- iedextensively. Thekeyistomakeuseoftheexcitation control technology andthefeedback control design boils downtothedesign ofsuitable excitation current/valtage. Various methods, ranging fromPIDexcitation control, power systemstabilizer tolinear optimal/suboptimal ex- citation control(8) havebeenproposed. However, theaforementioned designmethods are basedon an approximately linearized modelwhichis valid onlyinasmall neighborhood ofanequilibrium point ofthenonlinear system.Forthis reason,thefeedback controller wouldlosseffect whenthestatedeviates too faraway fromtheequilibrium point duetolarge distur- bancesuchasthelossofa majortransmission line or a fault atone ofthebusesinthesystem occurs.Thisisbe- causewhentransient stability ofthelarge disturbances is considered, thenonlinearities inherent inthepower sys- temcan no longer beignored anditseverely degrades the closed-loop systemperformance orevensometimes make thesystemunstable. Therefore, control methods arere- quired whichcanhandle thenonlinear modeldirectly and enhance thetransient stability ofpower system.Based on these motivations andwiththedevelopment ofthethe- oryofnonlinear control system, various typesofnonlinear stabilizing control havebeenproposed (8), (9), (3), (14), (5). However, itisverydifficult todevelop anoutput feed- backcontrol inthenonlinear control framework.