Research On Fault Diagnosis Method Of Urban Rail Vehicle Passenger Compartment Door System

ABSTRACT:With the development of society and economy, urban rail transit which has the unique technical advantages shoulders the important mission in terms of alleviating traffic congestion. Passenger compartment door is one of the most frequently used parts in the operating process of Urban rail vehicles, therefore, it caused frequent faults of passenger compartment door system with the inconvenience of passenger travel, which impact the safety operation of urban rail transit seriously at the same time. How to solve the faults of the passenger compartment door system efficiently, so as to maintain and enhance the safety and operational efficiency of urban rail vehicles has been the urgent technical problems to study.The paper mainly studies on the fault diagnosis methods of the passenger compartment door system, which is as follows:1. Firstly, by analyzing the passenger compartment door system, its system structure, operating principle and system function can be understood fully; Secondly, according to arrange the fault statistical data which have been recorded in the operating process of Guangzhou Metro line2vehicles from2011to2012, the failure frequency and influence factors can be found to assess the reliability of vehicle system and research failure modes and mechanisms.2. Based on the fundamental component failure mode analysis and the theory of FTA, the fault tree of the sliding plug door of the subway vehicle is developed, minimal cut sets are obtained with qualitative analysis of fault tree, and probability of the system and minimal cut sets importance are calculated with the quantitative analysis of resulting minimum cut sets. The results points out the weaknesses in passenger compartment door system. In addition, using the weighted least-square method to improve the shortage of the traditional TOPSIS method, and in order to improve the computational efficiency of the method, MATLAB have been used to carry on the simulation, finally, the fault scheme of the passenger compartment door system is sorted.3. Aiming at the shortcomings of the fault tree analysis and the improved TOPSIS method, a fault diagnosis method based on Bayesian Network is proposed in this thesis. Then, the paper realizes the simulation of the model on the MATLAB software. The specific steps are as follows:first of all, based on the fault tree model of passenger compartment door system, the model which is based on Bayesian Network has been established and the nodes are assigned a value. In addition, according to different types of the data, this thesis uses the maximum likelihood estimation algorithm and EM algorithm for parameter learning. Finally, the joint tree algorithm is applied to network inference, and the assessment of fault information and the diagnosis of the fault cause are implemented.