全矢融合的二元PELCD样本熵列车故障诊断

Train Fault Diagnosis Based on Binary PELCD Sample Entropy with Full Vector Fusion

长期高速运行的服役状态会造成高速列车转向架关键部件性能蜕化甚至发生故障等情况,所导致的安全事件将造成严重的经济损失甚至人员伤亡。考虑到高速列车振动信号的特性,将部分集成的局部特征尺度分解方法拓展至二元信号处理领域,同时结合全矢谱理论对同阶分量信号进行信息融合,得到更加完备的数据特征,并对融合后的数据进行样本熵特征提取,得到列车的故障特征;采用灰狼优化算法对支持向量机进行参数寻优,通过实验对比单一故障工况、复合故障工况以及部件性能退化下的故障识别率,验证所提方法的有效性、优越性。

The service state of high-speed train for a long time operation will cause the deterioration of the performance of the key components of its bogie,and the breakdown of the safety events will cause serious economic losses and even casualties.In this paper,considering the characteristics of high-speed train vibration signals,the partial integrated local feature scale decomposition method is extended to the field of binary signal processing.At the same time,based on the theory of full vector spectrum,the information fusion of the same order component signals is carried out to obtain more complete data features,and the sample entropy features of the fused data are extracted to obtain the train fault features.The Grey Wolf optimization algorithm is used to optimize the parameters of support vector machine.Finally,the fault recognition rates under single fault condition,compound fault condition and component performance degradation are compared by experiments to verify the effectiveness and superiority of the proposed method.