基于定子电流小波包分析的牵引电机轴承故障诊断  

Traction Motor Bearing Fault Detection Via Wavelet Packet Analysis of Stator Currents

作　　者：杨江天[1] 赵明元[2] 张志强[3] 李平康[1] 

机构地区：[1]北京交通大学机械与电子控制工程学院,北京100044 [2]中国北车集团大同电力机车有限责任公司,山西大同037038 [3]中国南车集团青岛四方机车车辆股份有限公司国家工程实验室,山东青岛266111

出　　处：《铁道学报》2013年第2期32-36,共5页Journal of the China Railway Society

基　　金：国家自然科学基金(61074104)

摘　　要：滚动轴承失效是机车牵引传动系统的主要故障源之一。轴承失效引起电机振动增加,导致定子电流发生调制。本文根据轴承失效对电机运行参数(振动、电流)的影响,将牵引电机轴承故障分为单点局部损伤和整体磨损。分析两种滚动轴承故障对振动和电机定子电流频谱的影响。定子电流分析可在不影响电机运行的情况下,检测电机的工作状况。小波包变换适应于处理瞬变、非平稳信号,用于电机定子分析能获得较高的频率分辨率,有效提取故障征兆。本文提出一种基于定子电流小波包分析的机车牵引电机轴承在线故障检测方法。HXD2型机车线路运行试验证明,该方法能有效诊断电机轴承早期故障。Rolling element bearing failures account for a large majority of mechanical faults in a locomotive trac- tion system. Bearing failures induce motor vibration, resulting in modulation of stator currents. In this paper, according to the effect of bearing failures on operating parameters of motor（e, g. , vibration and current）, bear- ing failures of traction motor were classified into single-point defects and generalized roughness. The influence of both types of rolling element bearing faults on the vibration and induction motor stator current spectrum was analyzed. Motor current signature analysis was introduced to provide a nonintrusive way to assess the health of a motor, being integrated with wavelet packet transform which offered an efficient decomposition of signals containing both transient and non-stationary components. Thus, fine frequency resolution was achieved and fault features were extracted via wavelet packet analysis of motor currents. The online locomotive traction mo- tor diagnostic method using motor current signature analysis was presented. The proposed method was effec- tively applied to the operation tests of Locomotive Type HXD2 , and the incipient bearing faults of traction mo- tor were diagnosed successfully.

关 键 词：牵引电机 轴承 故障诊断 定子电流分析 小波包变换 

分 类 号：U264.1[机械工程—车辆工程] TM343[交通运输工程—载运工具运用工程]