Study On Analog Circuit Fault Diagnosis Based On Fuzzy Neural Network

Researchers have been paying much attention to the test and fault diagnosis of analog circuit since 1960's. The development of fault diagnosis of analog circuit is very slow because of its variety and complexity. With the development of electrical technology, there is an urgent demand for reliability, servicing, automatic fault diagnosis etc. of analog circuit.Artificial neural network theories have developed rapidly in recent years and have been widely applied to many fields. Based on error back propagation (BP) arithmetic, many improvements are put forward because of its disadvantages. In this paper, resilient BP arithmetic is applied to radar power by the simulation of Matlab and PSpice software and correct fault diagnosis of tolerance analog circuit is realized.Fuzzy system is used easily to express knowledge and simulate human thinking, but it is short of the abilities of learning and adaptability. Artificial neural network possesses abilities of self-organization and self-learning, but doesn't have the ability of expressing structural knowledge and often gets into local extremum in the process of study. On the basis of the advantages of the two intelligence technologies, fuzzy neural network in form of series is put forward, which is applied to radar power by simulation of Matlab and PSpice software, and correct fault diagnosis of tolerance analog circuit is realized.