基于FFT和EMD的直流电动机换向电流提取方法

A New Approach of DC Motor Commutation Current Extraction Based on FFT and EMD

针对运用经验模态分解(EMD)提取直流电动机换向电流过程中遇到的模态混叠问题,提出一种基于FFT和EMD的换向电流提取方法。首先,用Fourier分析起动电流得其频谱,结合电机参数初步确定稳态时换向电流频率;其次,根据EMD将起动电流分解为若干固有模态函数(IMF)并对相应的IMF做Fourier频谱分析,由此对特定IMF求和以提取换向电流;最后,将提取的换向电流通过指定的低通滤波器以滤除附加的换向电流。与IMF直接求和获取换向电流的方法相比,实验结果表明该方法获取的换向电流较平滑并且换向电流的幅值信息较完整。通过小波脊线求取的换向电流频率信息表明滤波后的频率曲线波动变化较小。

Focusing on the problem of modal aliasing,which using EMD to extract the DC motor commutation current,a new approach of extracting commutation based on FFT and EMD was proposed.The starting current of the Fourier spectrum was acquired and steady-state commutating current frequency,related with motor parameters,was determined.The starting current was decomposed into series of intrinsic mode functions （IMF） by EMD and related IMF Fourier spectrum was analyzed,thus the corresponding order of the IMF summed to extract the commutation current.The commutation current was extracted by using the specified low-pass filter to filter out additional commutation current.Compared with the methods of IMF summed directly,experimental results show that the commutation current curve obtained by the method proposed is smoother and more complete information on current amplitude.Commutation current frequency obtained by wavelet ridge shows that frequency curve changes small after filtering.