基于VMD与GWO优化SVM的轴承故障诊断

针对传统振动信号特征提取方法与支持向量机(SVM)分类方法的缺陷,提出一种基于变分模态分解(VMD)故障特征提取方法与灰狼优化器(GWO)优化SVM的诊断模型。首先,将滚动轴承的原始振动信号采用VMD得到若干本征模态分量(IMF);其次,将IMF的多尺度加权排列熵作为特征向量并使用t-sne方法做降维处理;最后,使用GWO对SVM进行优化并对样本数据进行分类判别。实验结果表明,该方法相比于其他传统算法能够有效提高故障分类精度。

Fault Diagnosis of Bearing Based on VMD and SVM Optimized by GWO

Aiming at the defects of the traditional way to extract the feature of vibration signal and support vector machine(SVM) classification method, a way of fault diagnosis based on variational mode decomposition(VMD) and SVM optimized by gray wolf optimizer(GWO) was proposed. Firstly by using VMD method,the original vibration signals of bearing were decomposed into a finite number of intrinsic mode functions(IMF). Secondly the multi-scale weighted permutation entropy of every IMF was calculated as the characteristic vector and reduced the dimensions by t-sne method. Finally by using SVM optimized by GWO,classified and discriminate fault types. The experimental results show that it is more accurate compared to other traditional algorithms.