Study On Fault Diagnosis Of Rolling Bearing In The Key Parts Of EMU Walking Unit

With the increase of the passage flow and mileage of the high-speed railway, the amount of the EMU in our country is rising rapidly. Security has become the most important part in the operation process. As an essential component of the walking unit, situation of rolling bearing always influences safety of the operation directly. While operated in high speed, early fault of rolling bearing is easily to be expanded to cause safety accidents such as train overturn. The real-time and effective fault diagnosis system for rolling bearing of EMU is able to avoid the accident and help to decide the maintenance scheme by judging the type of the bearing fault. This is conductive to reduce the operation and maintenance cost.In this essay, basing on the mechanism of the rolling bearing fault, the diagnostic method is determined. Moreover, the fault diagnosis system for rolling bearing of EMU based on vibration signal is designed. The system could conduct the real time monitoring of the bearing vibration signal, which focuses on the domain characteristic parameters in order to monitor the bearing running state timely. If any fault takes place, it will be saved automatically. Furthermore, the fault characteristic is extracted by the method of non-stationary signal analysis and processing. Then, the fault type is judged. Eventually, the system can achieve the online monitoring and off-line diagnosis function for the rolling bearing of the walking unit. In this essay, the following work has been done:1. Complete the type selection of sensors and the selection and design of the hardware circuit of embedded data pre-processor. It involves signal conditioning circuit, AD conversion circuit, network port circuit, etc. Complete the design of core algorithm of the software. Utilize wavelet transform method to de-noise the original signal in order to enhance the system diagnostic accuracy. Use domain characteristic parameters of the vibration signal as evidence to achieve the real time monitoring for rolling bearing operating state.2. Regarding the CRH3 type EMU as the model, build up the analog signal of the bearing outer ring. Use EMD method to extract the fault characteristic. Through simulation test, the method is proved to be adjusted to the system.3. Complete the vibration test and result analysis. The vibration signal of the outer ring, the inner ring, the rolling element fault and the normal bearing are collected. Firstly, compare the domain characteristic parameters of the vibration signal to prove that these parameters could be used as the evidence of judging the bearing operating state. Then, utilize EMD algorithm to extract the fault characteristic. The result is tested the same with the predict fault type. Finally, the vibration test is done to prove the feasibility of the system.