Fault Diagnosis Of Metro Traction Control System Based On SaCE-ELM

With the rapid development of urban rail transit, the safety operation of the urban rail transit has been widely concerned. As the main carrier of urban rail transit, subway has made great contribution to solving the increasing serious traffic congestion. However, due to the failure of the train, not only the normal operation of the train but also the safety of passengers are affected. Therefore, it is important to diagnose the faults timely and effectively to ensure the normal operation of the rail transit.In view of shortcomings of the traditional fault diagnosis method, such as the need to understanding of equipment operation mechanism, difficulty in obtaining relevant knowledge, slow learning speed, falling into local optimum easily and poor generalization performance, an adaptive differential evolution algorithm called SaCE is proposed based on the differential evolution algorithm to optimize the input weights, the implicit layer parameters and the output weights of the extreme learning machine in this study. Different from artificial selection mutation strategy and parameter of standard differential evolution algorithm, the adaptive mechanism based on chaotic sequence enable the selecting of the differential mutation strategies more effective, and make coding process simpler. Experimental results on SaCE and SaDE in ten standard functions show that SaCE can effectively reduce the training time and has better optimization effect. Then SaCE-ELM is compared with other algorithms on the UCI data set. The results show that the method has better classification performance and performance.Finally, this paper introduces the metro train control system and traction control subsystem briefly. After that, we summarizes the related fault traction control system and input information associated with the traction unit by analyzing the function of each module of the traction system. And the method of extracting relevant sample data of faults from off-line data is introduced. After normalization of the raw data, the fault samples are obtained. The SaCE-ELM fault diagnosis algorithm proposed in this study is applied to the fault diagnosis of metro traction system, experiment result shows that it can diagnose the fault of the system effectively.