Fault Detection and Identification in Railway Turn out System, using Fuzzy Algorithms

A turnout system in railway signaling based on its function is considered as a safety-critical system. Turnout system consists of point machines, switch rails and interfacing mechanisms. In this paper in addition to introducing the point machine and its functionality, most important faults and their causes and consequences have been described. As the main part of the paper, a general review on different fault detection and diagnosis methods and details of the developing and implementing fuzzy algorithms for fault detection and diagnosis of point machine has been done. The fuzzy rules are developed based on real data froma laboratory based system, installed in the signaling laboratory of the School of Railway Engineering. The algorithm was tested and evaluated for different fault modes of the system. In general, the algorithm is able to detect and locate most of the point machine systems studied in this research work.