Fault Diagnosis of Railway Turnout Based on Fuzzy Cognitive Map

Abstract

Turnout is an important part of the railway signal system and also an important equipment to ensure the safety of traffic. In view of the current situation of low efficiency of using computer monitoring system to locate faults, this paper combines the theory of fuzzy cognitive map with the fault diagnosis of the turnout, and uses real-coded genetic algorithm which tells the initial weight of the network to construct a classifier model for fuzzy cognitive map to classify the faults of turnouts. The simulation experiments show that fuzzy cognitive map classifier model based on real-coded genetic algorithm can effectively classify the faults of the turnout. Compared with the commonly used classifiers such as BP and SVM, it can achieve better classification performance.