基于局部均值分解和极值差分的直流电机间接测速方法  

DC Motor Indirect Velocimetry Method Based on Local Mean Decomposition and Extreme Difference

作　　者：雷钢[1] 周铜[2] 黄传金[1] 谭联峰 

机构地区：[1]中州大学郑州市汽车人机关系与安全重点实验室,河南郑州450044 [2]中州大学科研处,河南郑州450044 [3]北京和利时电机技术有限公司,北京100085

出　　处：《电机与控制应用》2014年第2期57-63,共7页Electric machines & control application

基　　金：河南省科技厅科技攻关项目(132102310286);郑州市普通科技攻关项目(121PPTGG364-3);郑州市普通科技攻关项目(20130677);河南省教育厅科技攻关项目(14B470015)

摘　　要：针对直流电机无转速传感器间接测速问题,首先,根据局部均值分解(LMD)可将复杂信号按频率递减的顺序自适应地依次分离出各PF分量的特点,构建了基于LMD的自适应滤波器(AF)。其次,确定了AF中截止分量阶数以组成带通滤波器提取换向电流;最后结合直流电机换向特点,提出了基于极值差分(ED)的换向电流频率求取方法。试验对比了LMD和EMD提取换向电流特点,并对比ED方法和脊线算法、Hilbert变换及LMD中的反余弦法求取换向电流的频率特点,相关试验证明了所提方法的可行性和有效性。Focusing on the DC motor indirect velocimetry without senor measurement, according to the characteristic of the local mean decomposition（ LMD） , that PF components separated in decreasing order of frequency were arranged,adaptive filter （ AF） using LMD was built. Secondly, the order in AF cutoff components to form a band pass filter used to extract the commutation current was determined. Finally, with the characteristic of the DC motor commutation, extreme difference （ ED ） method for getting the reversing current frequency was proposed. The characteristics of reversing current respectively based on LMD and empirical mode decomposition （ EMD ） were compared by experiment, and the performance of getting reversing current frequency by ED method and ridge line algorithm, Hilbert transform and the inverse cosine of LMD were compared. Related experiments proved that the proposed method was feasible and effective.

关 键 词：局部均值分解 极值差分算法 换向电流 自适应滤波器 瞬时频率 

分 类 号：TM33[电气工程—电机] TN820.5[电子电信—信息与通信工程]