基于小波包分解和集合经验模态分解的列车转向架轴承智能故障诊断方法

摘要

Existing diagnosis methods for train bogie bearings have several shortcomings including unsatisfacto-ry fault feature extraction and low diagnosis rate.To address these deficiencies,this paper presented an intelli-gent fault diagnosis method for train bogie bearings.Combining the wavelet packet analysis and EEMD,this method fully extracted the fault characteristics from the signal,and identified the fault modes with the search algorithm for fault identification and the energy criterion,thereby further improving the diagnosis efficiency. To verify the validity of the proposed strategy,a test platform for bearing fault diagnosis was built ,where bo-gie bearings under three fault conditions,which were used in Guangzhou Metro,were tested and analyzed.The experimental results showed that the proposed fault diagnosis strategy for bogie bearings improved the rate and accuracy of train bogie bearing fault diagnosis by fully extracting fault features,quickly locating the search band,and accurately identifying bearing faults.%提取故障特征不理想、诊断速度慢等是目前现有列车转向架轴承故障诊断方法存在的主要不足。本文提出了一种列车转向架轴承故障的智能诊断方法。该方法将小波包分解和集合经验模态分解(Ensemble Empir-ical Mode Decomposition,EEMD)结合在一起,充分提取信号故障特征,并利用能量判别法和故障识别搜索算法进行故障模式识别,进一步提高了故障诊断速度。为了验证该方法的有效性,构建了轴承实验台,测试分析了广州地铁列车3种故障状态的转向架轴承。实验结果表明,该方法能够充分提取故障特征,迅速锁定搜索频段,准确识别轴承故障,提高了列车转向架轴承故障的诊断速度和准确性。