基于GA-SVM的DR-PMSM匝间短路故障诊断方法研究  

GA-SVM Based DR-PMSM Interturn Short Circuit Research on Fault Diagnosis Methods

作　　者：潘景宜 张旭[1] 晏佳胜 胡旭辉 PAN Jingyi;ZHANG Xu;YAN Jiasheng;HU Xuhui(China Electronics Technology Group 20th Research Institute,Xi’an 710068,China)

机构地区：[1]中国电子科技集团第二十研究所,西安710068

出　　处：《微电机》2025年第2期14-20,32,共8页Micromotors

摘　　要：针对双余度永磁同步电机(DR-PMSM)常见的绕组匝间短路故障问题,首先建立故障余度绕组数学模型,并分析故障后绕组的阻抗变化,在此基础上建立匝间短路故障有限元模型。然后利用DR-PMSM的结构特点,分析电机发生匝间短路故障后非故障余度绕组电流,将电流信号经过变分模态分解(VMD)得到各固有模态分量(IMF)的能量信息作为不同故障类型的特征量。最后使用支持向量机(SVM)作为故障诊断方法,并采用遗传算法(GA)优化SVM参数。实验结果表明基于GA-SVM的故障诊断方案能准确识别DR-PMSM的匝间短路故障,在空载和负载情况下故障诊断准确率分别达到85.2%和88.9%,表明所提方法具有较高的故障识别率。Aiming at the common problem of interturn short circuit fault in double redundancy permanent magnet synchronous motor(DR-PMSM),the mathematical model of the fault redundancy winding is established first,and the impedance change of the faulty winding was analyzed.Then,based on the structure characteristics of DR-PMSM,the non-fault redundancy winding current was analyzed,and the energy information of each intrinsic mode component(IMF)was obtained through variational mode decomposition(VMD)of the current signal as characteristic quantity of different fault types.Finally,support vector machine(SVM)was used as fault diagnosis method,and genetic algorithm(GA)was used to optimize SVM parameters.The experimental results show that the fault diagnosis scheme based on GA-SVM can accurately identify the inter-turn short circuit fault of DR-PMSM,and the fault diagnosis accuracy reaches 85.2%and 88.9%respectively under no-load and load conditions,indicating that the proposed method has a high fault recognition rate.

关 键 词：双余度永磁同步电机 匝间短路故障 变分模态分解 遗传算法 支持向量机 

分 类 号：TM351[电气工程—电机] TM341