基于小波变换及马氏距离的电路故障定位方法  

Method for Catastrophic Faults Location in Analog Circuits Based on Wavelet Transform and Mahalanobis Distance

作　　者：浦建开[1] 胡志忠[1] 竺琼[1] 邵元超[1] 

机构地区：[1]南京航空航天大学电子信息工程学院,江苏南京210016

出　　处：《电子科技》2012年第5期10-13,共4页Electronic Science and Technology

摘　　要：针对模拟电路中的硬故障,提出了一种基于马氏距离统计学原理的快速定位方法。利用Pspice电路仿真软件,对ITC'97基准电路集中的连续状态可变滤波器电路进行电路建模及故障注入仿真,用Pspice与Matlab软件的数据接口技术,将无故障及各故障状态的仿真数据导入Matlab中,进而在Matlab环境中进行小波变换处理及马氏距离的计算,并以马氏距离作为定位故障元件的依据。经检验,这种故障定位方法对模拟电路硬故障具有快速定位的能力,而且测点少、在线计算量小等优点。To solve the catastrophic fault of analog circuits,this paper proposes a fast location method based on Mahalanobis distance.Firstly,a simulation of circuit and fault modeling is made in PSPICE for the Continuous-Time Sate-Variable Filter in the ITC＇97 Analog and Mixed-Signal Benchmark Circuits.Secondly,simulation data of both fault-free and faulty circuit is imported into Matlab using PSPICE and Matlab software data interface technology.Then,wavelet transform and Mahalanobis distance calculation of the simulation data is done in Matlab.Finally,the faulty component would be detected based on the Mahalanobis distances of both fault-free and faulty circuits.Tests show that this method with quick positioning ability has the advantage of simplicity and small online computational amount.

关 键 词：模拟电路 故障定位 小波变换 PSPICE Matlab 

分 类 号：TN710.4[电子电信—电路与系统]