Study On Design And Algorithm Of Flat Wheel Monitoring System For Urban Rail Transit Vehicles

The wheel’s status of metro vehicle is one of the key factors which directly impact on the metro’s operation safety. The detection of metro vehicle’s wheel flat is important. The traditional method of artificial static detection for wheel flat cannot catch up with rapid development of urban rail transit at present. Based on the research of predecessors, a detection system for wheel flat based on rail vibration is set up according to the metro corporation, and the analytical method for measured wheel-rail vibrationl isanalyzed in this paper.Firstly, the significance and current research status of metro vehicles wheel flat monitoring are introduced. Contrasting research of the wheel flat monitoring system domestic and overseas the research status of the wheel-rail vibration analysis are summarized.Secondly, the types and causes of metro vehicle wheel flat are presented, and the system overall design and all kinds of signals collection are given based on the needs of the tread flat failure monitoring system. The control scheme of various types of signal are detailedly designed, and the major hardware are selected combined with the actual system.Thirdly, the requirements of software design of the monitoring system are analyzed and LabVIEW is chosen to develop the software. The software design and division of functional modules of the monitoring system are completed. With the implementation of functional modules in LabVIEW, the detail programming designs are presented.Then, the application development and the basic theory of wavelet analysis are introduced. According to the noise characteristics of the wheel-rail vibration signal collected from the monitoring system, wavelet threshold de-noising method is applied to realize the de-noising of the vibration signal. The discrete wavelet decomposition and continuous wavelet transform are applied to deal with the denoised vibration signal, and the feasibility of the wavelet analysis is used to handle the wheel-rail vibration signal.Finally, The theory of Hilbert-Huang Transform is introduced. The validity of Hilbert-Huang Transform is verified by using the simulation signal. The empirical mode decomposition is used on the wheel-rail vibration to obtain each IMF component and extract energy feature vectors, meanwhile spectrum analysis of the vibration signal is conducted. The results shows that Hilbert-Huang transform has the higher time and frequency resolution and can describe the wheel-rail vibration signal time and frequency distribution in detail.