Research On The Online Diagnosis For The Winding Turn-to-turn Fault Of Permanent Magnet Synchronous Motor

Automation and intelligence are the development tendency of modern industrial system and the stability of the system is the premise. Due to its high efficiency, high power density and robustness, permanent magnet synchronous motor (PMSM) has been widely used in modern industrial automation production, especially in the field of precision control. If some faults occur in motor and fail to be discovered in time, it has the potential risk to damage the motor itself and the entire system, which will cause huge economic losses. Therefore, it is significant to study the fault diagnosis of PMSM. In this dissertation, the current signal analysis and BP neural network are used to diagnose the winding turn-to-turn fault of the PMSM.Firstly, on the analysis of the mathematical models of PMSM under the normal state, a nonlinear fault model is proposed for the stator winding turn-to-turn fault of PMSM by introducing a short circuit branch in the faulty phase, focusing on the calculation of the faulty PMSM's stator winding resistance and inductance parameters. The model provides a good compromise between computational complexity and model accuracy and is versatile for both the healthy and the fault condition, avoiding to constantly switch PMSM model during the simulation analysis.Secondly, the third harmonic components of the motor stator current are extracted by using the Vold-Kalman Filtering order tracking algorithm (VKF-OT). According to that the amplitude of the third harmonic in faulty phase is much larger than that in the other two phases, it can locate the fault position. The fault size can be evaluated by the slope size of the curve of the current third harmonic amplitude versus the motor speed. A new method to calculate the negative sequence current based on the fundamental frequency is proposed, which is used as the other indicator of PMSM winding turn-to-turn fault. On the basis of the above, combined with BP neural network, a comprehensive fault diagnosis of PMSM winding turn-to-turn fault is implemented.Eventually, a fault diagnosis function module of PMSM is designed by using VC++and Matlab mixed programming. Online simulation and fault monitoring of motors can be realized through this function module. In the actual operation, it only needs to read the motor current and speed parameters from controller that online diagnosis of PMSM with winding turn-to-turn fault can be achieved.