基于RBF神经网络的永磁同步电机匝间短路故障诊断方法

打开文本图片集

中图分类号：TP277 文献标志码：A 文章编号：2095-2945(2025)14-0026-0

Abstract:Inter-tumshort circuitfaultsareoneof themostcommonandserious faultsthatoccrinpermanent magnet synchronousmotors (PMSM).Motorfaultdiagnosistechnologyisanimportantmeanstoimprovemotorreliabityandreducefault loses.Therefore,thispaperproposesamethodbasedonRBFneuralnetworktodiagnoseinter-turnshortcircuitfaultsin permanentmagnetsynchronousmotors.First,afiteelementmodelof inter-turnshortcircuitfaultof permanentmagnet synchronousmotorisestablished.Themotorwindingisdividedintomultiplesub-windings，andthetwoendsofthesubwindingsareconectedinparaleltosimulateinter-turnshortcircuitfaults.Secondly，theestablishedfiniteelementmodelis usedtosimulatethemotorperfomanceunderdiferentfaultdegres.Thepaperanalyzesandextractsfaultcharacteristicsfrom motortorque，phasevoltage,andphasecurent.Finally,afaultdiagnosissystem isestablishedusingRBFneuralnetwork.Ithas benverifiedthat theproposed faultdiagnosis methodcan diagnose diffrent degreesof inter-turn short circuits.

Keywords:permanentmagnetsynchronousmotor;inter-tunshortcircuitfault；faultdegree；faultfeatureextraction;RBF neural network

永磁同步电机(PMSM)具有结构简单、体积小、重量轻和可靠性高等优点-2，被广泛应用于电动汽车、风力发电等领域[3-4。（剩余9267字）