Inter-turn fault detection in PM synchronous motor by neuro-fuzzy technique

In this paper, a method for detecting the stator internal coil fault detection for a permanent magnet synchronous motor (PMSM) using the ANFIS algorithm is proposed and described. At first, the dynamic model of the synchronous motor along with its certain fault will be introduced. Since fault detection in these engines is very important and has a high value, different methods have been proposed for detecting stator deflection in electric machines. To determine the fault percentage in the permanent magnet synchronous motor, a neuro-fuzzy adaptive inference system is used to identify the fault. The advantages of the proposed algorithm are the ability to detect faults with different domains. It is flexible enough to be used for offline and online identification. For this reason, we have used neuro-comparative learning techniques in fuzzy logic in this paper. The inputs of the proposed algorithm are two PMSM current and torque signals in normal and faulty conditions. In the proposed algorithm, the membership function structure was created with the fuzzy C-means clustering method. The simulation results show that the proposed algorithm can accurately determine where and with what speed the fault occurs.