基于小波包变换和BP网络的铁道车辆滚动轴承故障诊断方法

Fault Diagnosis Method for the Rolling Bearing of Railway Vehicle Based on Wavelet Packet Transform and BP Neural Network

针对铁道车辆滚动轴承故障诊断,提出1种改进的小波包与BP神经网络相结合的故障诊断方法,并开发出基于该方法的铁道车辆滚动轴承故障诊断系统。用压电加速度传感器采集轴承试验台的模拟故障轴承振动信号,对采集到的信号先进行小波降噪,再通过小波包分解,构造特征向量,以此作为故障样本对改进的BP网络进行训练,实现智能化故障诊断。实验结果表明,基于该方法的故障诊断系统能够很好地诊断出铁道车辆滚动轴承内圈、外圈及滚动体表面出现的疲劳、剥落、磨损和裂纹等故障,具有实际工程应用价值。

In order to diagnose the fault of railway rolling bearing,a new method of fault diagnosis based on the improved wavelet packet and BP（Back Propagation）neural network was put forward,and a fault diagnosis system for railway rolling bearing was developed.The vibration signal of the analog fault bearing of the bearing test rig was collected by Piezoelectric Accelerometer.At first,the signal was de-noised by wavelet.Then it was decomposed by the improved wavelet packet to construct the eigenvector that can be taken as the fault samples to train the improved BP neural network.Finally,the intelligent fault diagnosis was realized.The test results show that the fault diagnosis system based on this method can well diagnose such faults as the fatigue,spalling,wear,crack,and so on,occurred in the inner ring,the outer ring and the rolling body surface of railway rolling bearing.The fault diagnosis system has high application value in practical engineering.