基于参数优化SVM的旋转机械故障诊断

Diagnosis of Rotating Machinery Fault Based on Optimized Parameter Support Vector Machine

介绍了支持向量机(SVM)分类算法,针对常用的C-支持向量分类机(C-SVM)中参数没有确切意义、选取困难的缺点,将ν-支持向量机(ν-SVM)用于对旋转机械振动故障的诊断,通过对实验台模拟数据进行训练、分类和测试,取得了良好的诊断效果。并通过现场故障实例证明了这种方法的准确性。

This paper recommend the classification algorithm of support vector machine.Traditionally,parameter C in C-SVM has no exact meaning and can not be selected easily.So ν-SVM is introduced to the diagnosis of rotating machinery fault.Through the training,classifying and testing of the stimulant data obtained in the lab,the proper result has been achieved.The practical example has confirmed the accuracy.