Data-based Automatic Health State Analysis For The Brake Pipe And Axle Of High Speed Train

In the last ten years, the high speed train has developed rapidly in china. The key technology of high speed train has been basically mastered. And the high speed train goes out of the country gradually. At the same time, the train running safety has become the top priority of high speed train development. Relevant researches have been carried out both at home and abroad. But the safety of high speed train is crucial. Only by developing the core technology with independent intellectual property rights, can we guarantee the healthy development of high speed train intelligence and green in our country. In particular, the axle and brake pipe are the key equipment for high speed train, and its operating condition is directly related to running safety of the train. The existing research focuses on the axle performance though the vibration signal of axle. It is lack of real-time. Most of the researches on the signal data of high speed train belong to signal fault diagnosis. In addition, researches on the performance of train axle are focused on the axle vibration signal. The performance state of brake pipe that is another key device relies on manual testing and relies heavily on manual experience. The direct impact of these problems leads to increased risk and cost of driving. This study mainly focuses on the safety assurance technology of high speed train axle and brake pipe, which mainly reflects the following three aspects.Firstly, due to the sensor problem, there are the isolated zero points and missing data problem in the complex high speed train monitoring data which is large and varied. Automatic preprocessing of monitoring data is realized. This method for monitoring data will not cause too much data deviation. And it may affect the result of future analysis and research. In the face of a wide range of missing data, this thesis adopts manual interpolation, and targeted interpolation according to expert analysis to minimize error.Secondly, for analysis of high speed train axle state, an effective indicator which can judge high speed train axle aging state is proposed through the analysis of monitoring data.The effectiveness and practicability is demonstrated based on the actual monitoring data of a high-speed train. Although common theory and method tools such as linear interpolation and discrete wavelet transformation were employed, the indicator and processing framework for measuring the ageing state of train axle are useful for early warning and precise localization of impending damaged axles, which can reduce the maintenance cost, and improve the system safety.Lastly, through analyzing the character of the pressure data, a practical criterion for the brake pipe of high speed train is proposed for health state analysis of high speed train brake pipe in this thesis. Then the actual fault characteristics of brake pipe are analyzed by the criteria, and the fault diagnosis for the brake pipe is realized by calculating the relative characteristics of brake pipe pressure data. The results of practical application prove the effectiveness and practicability of the proposed method.