基于EEMD-RobustICA的车用永磁同步电机噪声源识别

针对电动汽车驱动用永磁同步电机噪声源问题,采用集合经验模态分解(EEMD)结合鲁棒性独立成分分析(Robust-ICA)方法对电机噪声信号进行分解与分离,利用傅里叶变换与小波变换相结合的方法对各噪声源进行频域与时频域分析。结果表明,EEMD-RobustICA的方法可以有效地识别出该永磁同步电机的主要独立噪声源,各噪声源的贡献量大小依次为开关频率噪声、径向电磁力噪声和机械噪声。开关频率噪声是一种高频瞬态噪声,在时频域上呈现周期性变化。径向电磁力噪声与机械噪声均为稳态低频噪声。基于EEMD-RobustICA的噪声源识别方法可以为降低永磁同步电机的噪声提供理论依据。

Noise Source Identification of Permanent Magnet Synchronous Motors for Vehicles Based on EEMD-RobustICA

Aiming at the problem of noise source of permanent magnet synchronous motor for electric vehicle,the method of ensemble empirical mode decomposition(EEMD)combined with robust independent component analysis(RobustICA)was introduced to decompose and separate the motor noise signal.The method of Fourier transform and wavelet transform was used to analyze the time frequency characteristics of each noise source.The results show that the EEMD-RobustICA method can effectively identify the electromagnetic noise and mechanical noise of the permanent magnet synchronous motor.Among them,electromagnetic noise caused by current harmonics is the most relevant,followed by electromagnetic noise and mechanical noise caused by the structure of the motor.The electromagnetic noise caused by current harmonics is a kind of high frequency transient noise,which has periodic changes in the time frequency domain.The other two kinds of noise are low frequency noise,and the signal energy in the time domain is stable.The EEMD-RobustICA method provide a theoretical basis for reducing the noise of the permanent magnet synchronous motor.