基于蝙蝠算法优化相关向量机的轴承故障诊断方法

Bearing fault diagnosis method based on relevancevector machine optimized by bat algorithm

由于轴承振动信号具有复杂性和非线性,难以有效提取故障特征,影响故障诊断的准确率。为了提高故障诊断准确率,提出一种蝙蝠算法(BA)优化相关向量机(RVM)的轴承故障诊断方法。首先结合变分模态分解和多尺度熵从轴承振动信号中提取出故障特征,作为相关向量机的输入向量;接着采用蝙蝠算法优化相关向量机的核函数参数;然后训练相关向量机模型;最后使用训练后的相关向量机进行故障诊断。通过仿真实验评估故障诊断方法的有效性,实验结果表明,该方法的故障诊断准确为100%,故障诊断准确率高于SVM方法、RVM方法,说明BA-RVM故障诊断方法是可行和有效的,满足一般轴承故障诊断的精度要求。

Because of the complexity and nonlinearity of bearing vibration signal,it is difficult to extract fault features effectively,which affects the accuracy of fault diagnosis.In order to improve the accuracy of fault diagnosis,a bearing fault diagnosis method based on relevance vector machine(RVM)optimized by bat algorithm(BA)is proposed.Firstly,the fault features are extracted from the bearing vibration signals by combining variational mode decomposition and multi-scale entropy.These features are used as the input vector of the relevance vector machine.Secondly,the kernel parameter of RVM is obtained by bat algorithm.Third,the relevance vector machine model is trained.Finally,the trained relevance vector machine is used for fault diagnosis.The validity of fault diagnosis method is evaluated by simulation experiment.The experimental results show that the fault diagnosis accuracy of BA-RVM method is 100%,which is higher than SVM method and RVM method.It shows that BA-RVM method is feasible and effective,and meets the accuracy requirements of general bearing fault diagnosis.