电路故障诊断的支路代替法和模糊聚类法

Branch Substitution and Fuzzy Clustering for Circuit Fault Diagnosis

本文提出了一种利用非口器和零口器进行模拟电路故障诊断的新方法,不受拓扑条件限制,不需要求解方程。同时讨论了存在容差时故障支路的分类问题,并提出了一种新的分类方法——模糊聚类法,克服了传统方法的漏判和误判现象,使算法更加可靠。

A new branch substitution approach using norator and nullator is described for analog circuits fault diagnosis, this method doesn't require any limitation of the topology conditions and any solution of equations. Simutaneously, a new fault clustering approach——Fuzzy Clustering Algorithm is described, which can prevent the escape or wrong judge of fault when the network consists of torerance elements.