| Abstract:As the power plant for the electric locomotive, traction motor plays a vital role in whether the electric locomotive could run safely. Due to the harsh operating environment, the load changes greatly and other factors, traction motor is prone to failure, that affects production and life, and even leads to accidents. So it is necessary to carry out the condition monitoring and fault diagnosis for traction motor. Rotor bar breaking fault is one of the most common fault for traction motor. In this paper, it researched the rotor bar breaking fault deeply, and developed a fault diagnosis system.In this paper, it researched the mechanism of the rotor bar fault and the fault characteristic frequency of the stator current. Through extrapolation, it obtained the relationship among the fault characteristic frequency, power frequency and the actual motor speed. It can calculate the fault characteristic frequency of the stator current for the rotor bar breaking fault in various cases. The feature extraction is the most important part of the fault diagnosis. In this paper, it analysed and handle the stator current signal with the combination of wavelet packet transform and Fourier transform, and the feature extraction can well be done. In this paper, by using MATLAB, it did a large number of simulation, and verified the feasibility of this method. On this basis, it designed the best search algorithm. It is able to extract fault feature in the various speed conditions of the motor.By using the laboratory conditions, a test platform was set up. To create artificial rotor bar breaking fault, the algorithm was validated and improved. Relying on "electric traction test platform" this "985ã€211project", a diagnosis system for rotor bar breaking fault was designed. It mainly divided into four modules which is parameter setting, data acquisition, signal analysis and fault diagnosis. By collecting the stator current and speed signal of the traction motor, with the motor parameters, this system analysed and handle the data, and it is effective to identify the rotor bar breaking fault of traction motor in the various speed conditions. |