Research On The Detection And Diagnosis Scheme Of Induction Motor Fault

This paper focuses on the detection and diagnosis of rotor bar breaking (RBB), stator winding inter-turn short circuit (SWITSC) and bearing fault in induction motors. Many contributions to improve the sensibility, reliability and practicability of the detection and diagnosis methods have been made, listed below.1. By combining the Multi-Loop model and Finite Element Method together, the ML-FEM mathematical model of induction motors is established. And thus, the numerical simulation of RBB, SWITSC fault can be performed, as paves the way to investigate the corresponding detection and diagnosis methods.2. The simulation of RBB fault in induction motors has been completed successfully, as well as the corresponding experiments. Based on thorough analysis of the simulation and experiment results, a novel method to detect RBB, which perfectly blends subdivision Fourier transform, self-adaptive filter, rotor slot harmonics based slip estimation and detection threshold self-tuning techniques, is proposed. Fault detection instances in laboratory demonstrate that the presented method is effective, even taking into account the intrinsic asymmetry, air-gap eccentricity and rotor disalignment of induction motors. At the same time, a novel criterion to estimate the number of broken rotor bars is deduced, which possesses better performance than those presented previously.3. The simulation of SWITSC fault has been completed successfully. Also, the corresponding experiments have been carried out. According to that, a novel detection method of SWITSC, which blends perfectly the spectrum correction and self-adaptive filter techniques, is proposed. Simulation and experiment results demonstrate that the novel method is immune to RBB fault and thus more reliable than those presented previously. Furthermore, by taking the stator winding apparent impedance angle as the feature, a method to locate the stator winding faulty with inter-turn shorts is proposed.4. The joint detection necessity for RBB and SWITSC fault in squirrel cage induction motors is clarified, and the corresponding joint detection method, valid and practical, is put forward successfully. The simulation of RBB and SWITSC double fault in squirrel cage induction motors has been completed successfully, as well as the corresponding experiments. The double fault features have been generalized, and moreover, the appropriate detection strategy is recommended.5. By thoroughly analyzing the mechanism and features of bearing fault in induction motors, the vibration signal based and stator current signal based methods to detect and diagnose bearing fault have been proposed respectively.6. The Motor Incipient Fault Online Detector (MIFODor), which possesses the capability of detecting and diagnosing all the RBB, SWITSC and bearing fault in induction motors, has been developed successfully and utilized widely on-site.