Method Of Turnout Fault Diagnosis Based On Grey Correlation Analysis

With the rapid development of railway transportation toward high speed, high density direction, it puts forward higher requirements on the safety and reliability of railway signal equipment. The turnout as an important part of railway station signal interlocking system is the key equipment which arranging train route and implementing train route conversion, so the chronergy and efficient implementation of the turnout fault diagnosis is an important part of the protection of railway transport safety. The present turnout monitoring method is observing the information of actuation time, working current and power of the turnout through mocrocomputer monitor system. But the manual analysis diagnosis method for the turnout fault has been unable to meet the railway development needs on diagnosis efficiency, chronergy and cost. Meanwhile as the complexity of the equipment structure and surrounding environment, the turnout fault modes have diversity and uncertainty. So the study on the corresponding turnout fault diagnosis has the important practical significance and practical value.This paper analyzes the basic operation principle and turnout machine power information from the mocrocomputer monitor system, puts forward the turnout fault diagnosis method based on the gray relational analysis, and focuses on the analysis of the grey correlation model building in order to achieve the best perfomance diagnosis. The main research work in the paper is included as follow.1. Divide turnout into startup, unlock, conversion, locking, location expression based on the basic structure and working principle of the turnout equipment, and divide the power curve into five time segments based on turnout working process to analyze the turnout fault mode.2. The turnout fault diagnosis method is proposed based on grey relational analysis. Through the pretreatment of the switch machine power information, eliminate the disparate impact between the different switch machines, extract each time zone of power curve based on the time domain feature parameters, and then select the best feature from each time zone to constitute the feature sequence sets based on the Fisher criteria. Establish the feature sequence model of turnout fault modes though analysising the sequence changes of grey relational system, select Deng correlation algorithm to achieve the calculation of grey correlation degree after comparing all kinds of gray correlation degree model, finally realize the turnout fault diagnosis though the fault recognition strategy.3. According to the importance of the distinguishing coefficient on Deng’s correlation algorithm this paper achieve the best coefficient value through the test sample set experimental verification, and verify the functionality and performance of turnout fault diagnosis system. The final results show that the turn fault diagnosis method based on grey relational analysis can realize turnout fault diagnosis effectively, and have good robustness that solve the diversity and uncertainty of turnout equipment and environment factors.