应用经验模态分解下的AR模型提取电动机故障特征  

AR Model from Motor Fault Feature Extraction Based on Empirical Mode Decomposition

作　　者：许允之[1] 牛曼[2] 仝年 Xu Yunzhi Niu Man Tong Nian(School of Electrical and Power Engineering, China University of Mining and Technology, Xuzhou 221116, China School of Energy and Electrical, Hohai University, Nanjing 211100, China School of Electrical ＆ Electronic Engineering, North China Electric Power University, Baoding 071003, China)

机构地区：[1]中国矿业大学电气与动力工程学院,江苏徐州221116 [2]河海大学能源与电气工程学院,江苏南京211100 [3]华北电力大学电气与电子工程学院,河北保定071003

出　　处：《煤矿机电》2017年第1期47-51,54,共6页Colliery Mechanical & Electrical Technology

摘　　要：为有效地诊断电动机断条故障,提出了一种基于EMD-AR模型的电动机断条诊断的信号分析新方法。该方法将时间序列的AR模型引入到电动机断条故障诊断中,采用了经验模态分解方法将电动机的电流信号分解成若干个平稳的IMF分量,对前三个分量建立AR模型,并对固有模态函数进行功率谱分析。通过对比正常电动机、一根断条满载电动机和一根断条空载电动机的A相电流信号的IMF1(V)~IMF3(V)分量,通过AR模型估计的功率谱图提取故障特征并分析。仿真和实验结果表明,此方法能有效识别电动机断条故障。In order to effectively diagnose the motor fault, a new kind of diagnosis approach for the motor of broken bar faults, which is based on EMD-AR model, is proposed. The method introduces time series AR models to motor fault diagnosis. And motor current is decomposed into the limited inherent mode function （IMF） by the EMD. Then the first three IMF components are used to set up AR model and the power spectrum density （PSD） of intrinsic mode functions are analyzed and compared. Fault features are extracted by comparing the PSD picture of good motor, full load motor of broken bar and the non-load motor of broken bar. The result of this experiment suggest that this method can be used effectively for motor fault diagnosis.

关 键 词：经验模态分解(EMD) 自回归(AR)模型 固有模态函数(IMF) 故障特征提取 转子断条 

分 类 号：TM343.2[电气工程—电机] TP206.3[自动化与计算机技术—检测技术与自动化装置]