A new approach for fault section diagnosis in distribution network based on the combination of rough set theory and ANN
Zhiwei Liao · 2003
This paper presents a rough set theory (RST) based artificial neural network (ANN) method to deal with distorted information caused by missing and incomplete information in fault section diagnosis (FSD) system of distribution network (DN) in most practical application. In this approach, RST is used to analyze knowledge region data set and NN is built on the basis of diagnosis knowledge. So the fault tolerance performance of diagnosis system is assured with the combination of the qualitative analysis ability of RST and the generalization ability of NN. The high fault tolerance performance of proposed approach is proved through comparison with that of RST-model and NN-model based fault diagnosis system.