基于支持向量机的模拟电路故障诊断方法研究

Method of Analog Circuit Fault Diagnosis Based on Support Vector Machine

支持向量机是以有限样本下的统计学习理论为基础,它能较好地解决了高维数、非线性、局部极小点等问题,所以将它应用到模拟电路的故障诊断。首先介绍了支持向量机基本思想及分类算法,然后描述了支持向量机的模拟电路故障诊断方法,最后以一个带通滤波器电路作为诊断实例,诊断结果表明该方法可以准确有效地对模拟电路故障进行识别和诊断。

Support vector machine based on statistical learning theory for limited sample,can better solve the problem of high di- mension,nonlinear and local minima points,so it is applied to the analog circuit fault diagnosis.Firstly,the basic idea of sup- port vector machines and classification algorithm is introduced,then the method of analog circuit fault diagnosis based on support vector machine is described.Lastly,instance of a band-pass filter circuit as a diagnostic is supplied,diagnostic re- suits show that the method can ac.o.Hratp.lv and p.ffP.~：tiv^lv icl^ntifv ~nd cti^nnr^A ~n=lnn t＇-irz-~fit f^tdt