基于改进的EEMD-Hilbert包络解调轴箱轴承故障诊断

Fault Diagnosis of Axle Box Bearings Based on Improved EEMD-Hibert Envelope Demodulation

为解决在轴箱轴承故障诊断中,原始振动信号带通滤波器参数选择依靠主观工作经验,EEMD分解技术中多个IMF分量分析效率及准确性较低等问题,提出一种改进的EEMD-Hilbert包络解调方法。该方法采用EEMD对原始信号进行分解,利用双相关系数法及阈值法选取有效本征模态分量,通过对本征模态分量进行Hilbert包络解调,获取包络信号幅值谱作为故障特征量进行轴箱轴承故障诊断。并通过轴箱轴承实验台测试验证了该方法的有效性。

In order to solve the problems affecting application of traditional Hilbert envelope demodulation and EEMD technology,such as the parameter selection of the original vibration signal band-pass filter depending on experiences of individuals and the low accuracy and efficiency of analysis to many IMF components,a method for fault diagnosis of the axle box bearings based on improved EEMD and Hilbert envelope demodulation is proposed.Firstly,the original signal is decomposed into many components by EEMD,then the effective IMF components,chosen by using the double correlation coefficient method and the threshold value method,are analyzed with Hilbert envelope demodulation and the type of the axle box bearing fault is identified by the demodulation spectrum.An axle box bearing test results show the validity of the proposed method.