Wheel Tread Of Urban Rail Vehicles, Fault On-line Monitoring System For Research

Because wheel tread fault is one of significant risks in metro vehicle driving, its real time online monitoring is an important means to guarantee safety, which also is the key of providing higher passenger service. Thus developing a tread fault real-time on-line monitoring system is of great realistic significance.In this paper, various domestic and overseas tread fault detection methods are analyzed, and by comparing their advantages and disadvantages as well as considering the development status of domestic urban rail transit, we have chosen the ground type tread fault test method based on the vibration acceleration test. The main work is concluded as follows:Firstly, researches on main types and formation mechanism of the tread fault were done, consequently this paper established the theoretical model of flat wheel, a typical type of tread fault, and analyzed the association between the track vibration and tread fault with different severity levels. The fault is diagnosed by the energy of vibration and impact.Secondly, according to the vibration test theoretical model and the system design requirements, we designed the hardware scheme and the deployment and installation way of the detection system. The information acquisition schemes for vibration signals, wheel axle signals and vehicle numbers were given in detail. And then the hardware parameters for monitoring performance of the system were put forward.In the aspect of detection algorithms, we firstly analyzed the characteristics of the acquired vibration signal using Fourier transform and the wavelet theory. The urban rail tread fault detection algorithm was studied based on the vibration short-term energy detection method. Then we proposed using signal-detection threshold and energy compensation to improve the accuracy of the system. Finally, the detailed algorithm process of the tread fault detection system was brought forward, and the simulation results show it can well eliminate residual detection and error detection.In the last of the paper, the system software debugging platform based on Lab VIEW was introduced. The hardware was tested and the algorithms were debugged, and the results show that the system performance indicators can meet the design requirements, meanwhile the accuracy of the tread fault detection method developed in this paper has been proven by the long-term test.