基于帶差分擾動的改進PSO的PMSM參數(shù)辨識

打開文本圖片集

引用格式：，.基于帶差分擾動的改進PSO的PMSM參數(shù)辨識[J].現(xiàn)代電子技術(shù)，2025，48（10）：25-30.

Abstract：Inordertoaddresstheissuesof slowspeedandlowaccuracyintheparameteridentificationof permanent magnetsynchronousmotors （PMSM），amethodofparticleswarmparameteridenticationbasedonthecombinationofrandom samplemeanlearngstrategyofdaptieinertialweightanddiferentialperturbationisproposed.Inisalgorithm，allparticles inascendingorderaresortedbasedontheirfitness，andtheparticleswithfitnessrankingbeforethecurrntparticle forma sample pool. The average behavior of k particles in the current particle sample pool can be randomly selected for learning，and anadaptiveinertiaweightbasedonlogisticfunctionisusedtoimprovetheconvergenceperformaneofthealgorithm.A diferentialperturbationmethodisadoptedandembeddedintoteimprovedalgorithmtoenhancetheglobalsearchabilityof the algorithmin theearlysearch stage.Thesimulationresultsshow thattheparameteridentificationcurvecanconvergewhen the number of iterations is around 5O，and the identification eror of four parameters including stator resistance Rs， inductance Lq and （20 Ld， and permanent magnet flux linkage ψf is only 1.58% at the highest and O.005 2% at the lowest.The proposed strategy can accurately identify motor parametersand has the characteristics of fast convergence speed and high accuracy.

Keywords：permanent magnetsynchronousmotor;parameter identification；diferential perturbation;adaptiveinertia weight; improved PSO; sample mean learning strategy; hybrid strategy

0 引言

永磁同步電機（PMSM）因具有結(jié)構(gòu)簡單和高功率密度的優(yōu)勢[-3]，被廣泛應(yīng)用于工業(yè)伺服系統(tǒng)和工業(yè)自動化等領(lǐng)域。（剩余6384字）