基于GA-SVM的永磁同步电机动态偏心故障诊断

Research on Dynamic Eccentricity Fault Diagnosis Method of Permanent Magnet Synchronous Motor Based on GA-SVM

针对如何提高永磁同步电机(PMSM)动态偏心故障的诊断准确率问题,提出了基于遗传算法参数寻优支持向量机的永磁同步电机动态偏心故障诊断方法。首先,对电机的定子气隙磁通密度进行理论分析,推导出了动态偏心状态特征频率表达式;其次,利用快速傅里叶变换(FFT)对定子电流信号进行分析,获取定子电流信号的动态偏心状态特征频率并提取出幅值;接着,通过分析动态偏心状态特征频率的幅值之间比值,提出了一种具有抗负载干扰能力的故障诊断指标;最后,利用遗传算法参数寻优的支持向量机对提取的特征进行故障的识别。结果表明,所提出的故障诊断指标受负载干扰的影响较小,能够实现对表贴式永磁同步电机动态偏心故障的准确诊断,准确率可以达到96.5%。

Aiming at the problem of how to accurately diagnose the dynamic eccentricity fault of permanent magnet synchronous motor(PMSM),a permanent magnet synchronous motor dynamic eccentricity fault diagnosis method based on genetic algorithm parameter optimization support vector machine is proposed.Firstly,theoretically analyze the magnetic flux density of the stator air gap of the motor,and derive the characteristic frequency expression of the dynamic eccentricity state.secondly,use the fast fourier transform(FFT)to analyze the stator current signal to obtain the dynamic eccentricity of the stator current signal State characteristic frequency and extract the amplitude.then,by analyzing the ratio between the amplitude of the dynamic eccentric state characteristic frequency,a fault diagnosis index with anti-load interference capability is proposed.finally,the support vector of genetic algorithm parameter optimization is used the machine performs fault identification on the extracted features.The research results show that the fault diagnosis index proposed in this research is less affected by load interference,and can accurately diagnose the dynamic eccentricity fault of surface-mounted permanent magnet synchronous motor,and the accuracy rate can reach 96.5%.