支持向量机回归方法在故障诊断中的应用研究

The Application of SVM Regression Method in Fault Diagnosis

支持向量回归是支持向量机用于回归中的情况,首先介绍基于支持向量机的线性回归和非线性回归的基本原理,然后提出一种时间序列预测方法和误差评价的方法,最后在matlab中模拟简化同步的异常情况,通过实时采集的数据与多步预测值的残差来判断电机的异常,实验表明利用支持向量机回归能及时跟踪输出数据的变化,对设备进行在线故障检测是非常有效的。

support vector regression is the application of SVM in regression,this article firstly introduce the principle of SVM linear regression and SVM nonlinear regression,then bring forward a method of forecast and error estimation,finally,this article simulate simplified synchronization electric machine in matlab,through real time data and forecast to judge abnormal,test indicate SVM can trace change of data in time,Application of this method of online fault detection equipment is very effective.