High-Speed Fault Diagnosis Research Based On The Technologies Of Cluster Ensemble

With the development of High-Speed Railway, the train has become the indispensable traffic tools of people’s daily life, however, Security hidden danger of High-Speed Railway has become a problem which cannot be ignored. So accurate to solve the problem quickly is becoming more and more important.The traditional fault diagnosis methods are including analytical method of the model, signal processing method and the method of knowledge. Some methods need accurate mathematical model, some need large amount of calculation and some need a lot of knowledge of expert experience. But these conditions relative to the high-speed signal is difficult to obtain, at the same time it doesn’t satisfy the realtime performance for the high-speed signal fault diagnosis. Clustering ensemble is the combination of some independent cluster’s results, so as to get an optrnal clustering result to the original data. Clustering ensemble can reduce the influence of noise and outlier on the clustering result, at the same time it also can improve the robustness and stability of the clustering results. This paper apply the clustering ensemble technology in high-speed signal failure analysis.The main work of this paper is as follows:(1) Analysis of the high-speed signal fault diagnosis and the current research status of clustering ensemble, clustering ensemble technology is proposed for the high-speed signal fault diagnosis.(2) Feature selection for the high-speed signal data set with the different feature selection algorithms is the main content in the third chapter, each of these algorithms was introduced in detail, and its experiments and results.(3) The high-speed signal fault diagnosis method which based on the clustering ensemble was introduced in detail. In this paper we use matlab to write code, compare the result of single clustering algorithm to the results of clustering ensemble.This paper proposes the high-speed signal fault diagnosis method using clustering ensemble technology, the experiment data is from southwest jiaotong university tractive power laboratory of simulation data and the measured data. The experimental results show that this method has better performance than a single clustering algorithm, and can be more accurate and effective for fault diagnosis of High-Speed Rail.