Research On The Fault Diagnosis Method Of High-Speed Train Air Brake Pipe Based On Monitoring Data

The fault diagnosis of high-speed train is of great significance to ensure the safe operation of train,troubleshooting and diagnosis on site,and reduce maintenance and maintenance cost.In recent years,due to the rapid development of artificial intelligence,the theory of fault diagnosis based on intelligent methods has been widely concerned by many scholars.Therefore,this paper studies several methods of fault diagnosis of high-speed train based on monitoring data,and the main research work can be summarized as follows:(1)A fault diagnosis method based on Hidden Markov Model(HMM)is proposed.Due to the influence of operating environment and equipment,the monitoring data is often disturbed by noise and uncertainty,and because of network failure,transmission interruption,harmonic interference and other reasons,the monitoring data has low density data loss problem,at the same time the frequent changes of train working conditions and complex operating environment lead to the drowning of useful information and the complexity of fault analysis.Therefore,this method preprocessed the monitoring data through data interpolation,filtering denoising and feature extraction,then the HMM fault diagnosis model was established for different working conditions of air brake pipe,and then calculate likelihood probability to realize fault diagnosis.Experimental results show that the method can improve the accuracy of fault diagnosis for existing data.,but its generalization and robustness are insufficient.If an unknown fault is brought into this method,a diagnosis error will occur.In order to improve the accuracy of fault diagnosis,and to study the generalization and robustness of fault diagnosis,this paper proposes a method for fault diagnosis of high-speed train air brake pipe based on model space.(2)In view of the lack of generalization and robustness of the fault diagnosis method of high-speed train,this paper proposes a method of fault diagnosis of high-speed train air brake pipe based on model space.The pressure curve of air brake pipe was learned by using BP(Back Propagation)neural network.In the time series space,the fitting function of Fourier basis,Gauss basis,polynomial basis and sinusoidal basis and so on were used to approximate the air brake pipe pressure curve.In the model space,the air brake pipe pressure data was converted into model elements.The air brake pipe fault was diagnosed by using the topological relations of model elements in the model space.Compared with the HMM based fault diagnosis method,this method can effectively improve the accuracy of fault diagnosis,and also has a better diagnosis effect for the unknown fault.