The Fault Diagnosis Of Power Electronic Circuits Based On Rough Set Theory And Artificial Neural NetWorks

The fault diagnosis of power electronic circuits is one of the most important task in power system. Compared to other circuits, the fault diagnosis of power electronic circuits depends on the out-voltage wave of the circuits. This paper introduces the fault diagnosis of power electronic circuits based on Rough-set theory and artificial neural networks, and presents a new fault diagnose method of those circuits. Rough-set theory can analyses and deals with the incorrect , in consistent and intact data effectively ,and simplifies them. It can determines the fault type correctly by classifying the fault sign information directly. So this is a new fault diagnosis method in power electronic circuits. Artificial neural networks is a effective fault diagnose method. In this paper, neural networks is used to identify the location of fault elements. Because of the enormous data that are used to train neural networks, the speed of training neural networks and diagnosing is very slow. So it is very effective to combinating neural network and Rough-set. One method is that it takes two steps in diagnosing. First, the fault type is identified by Rough-set, then the neural networks determines the fault elements by the sampling voltage values. With regarding to the other method, the Rough-set is looked as the pre-system of neural network.. The fault information as well as the sampling voltage values are simplified by Rough-set, and the outcome of this processing is the input of neural network.. So the scale of neural network is reduced accordingly. The combinating of these two theory begins new trying in power electronic circuits fault diagnose.