基于电气信号的异步电机故障识别  

Fault identification of asynchronous motor based on electrical signals

作　　者：陈长征[1] 王胤龙[1] 李明辉 白丽荣[3] 程锦生[1] 

机构地区：[1]沈阳工业大学振动噪声研究所,沈阳110178 [2]中国石油辽河宝石石油装备有限公司,辽宁盘锦124001 [3]沈阳电缆股份责任有限公司,沈阳110025

出　　处：《沈阳工业大学学报》2008年第3期241-244,265,共5页Journal of Shenyang University of Technology

基　　金：教育部春晖计划资助项目(Z2005-1-21007)

摘　　要：为了提高异步电机故障诊断的可靠性,防止故障错报与漏报,以定子电流分析方法为理论基础,采取同时采集异步电机定子电流与电压信号的电机电气信号分析方法,分析异步电机电气、机械故障,并通过神经网络对复杂的电流和电压信号进行故障模式识别.模拟异步电机绕组匝间短路和断条故障进行实验,实验结果与实际基本符合.利用电气信号分析电机故障,可以区分电机电气故障和机械故障,为诊断结论提供了更加可靠的依据,与神经网络结合,利于故障模式的智能识别.To improve the fault diagnosis reliability of an asynchronous motor and prevent misreporting and underreporting, the electrical signal analysis method was employed to analyze the electrical and mechanical faults based on stator current analysis theory and through simultaneously acquiri.ng the current and voltage signals of asynchronous motor. The fault model identification for complex current and voltage signals was conducted by neural networks. Both winding turn-to-turn short-circuit and broken rotor bar faults have been tested in laboratory, and the results are basically coincident with the real case. Electrical signals can be used to analyze and distinguish electrical and mechanical faults, which can provide reliable basis for diagnosis conclusion and is also benefidal to intelligent recognition of fault pattern with combining neural networks.

关 键 词：故障诊断 电气信号分析 神经网络 异步电机 频谱分析 

分 类 号：TM307.1[电气工程—电机]