Dual SOGI-Based Framework for IPMSM Sensorless Control and Fault Feature Extraction Under Early Interturn Short-Circuit Conditions

In the rail transit application, sensorless control of permanent magnet synchronous traction system is recognized as a research hotspot. When the motor operates at high speed, the inter-turn short circuit fault (ITSC) reduces the sensorless control performance, and the increased position estimation error simultaneously affects the fault feature extraction accuracy. In this work, the fault condition of 1% -5% short circuit turn ratio (SCTR) is selected to analyze its influence on position estimation performance, and then an improved position observer and fault feature extraction module based on dual SOGI framework is designed. Firstly, a second-order generalized integrator with adaptive speed estimation (ASE-SOGI) is introduced into the position observer to reduce the position estimation error. Secondly, a multi-resonant cascaded SOGI (MRC-SOGI) is developed to extract sensitive fault features. The simultaneous operation of the ASE-SOGI module and the MRC-SOGI module ensures the optimal evaluation of the early ITSC under sensorless control. The former ensures the robust position estimation under the ITSC fault, and the latter isolates the fault-related harmonics. Through the reconstruction of the relevant characteristic harmonics, the fault signal is enhanced to effectively detect the early ITSC fault. Finally, the effectiveness of the proposed dual SOGI framework is verified by experiments on a 2.2 k W built-in permanent magnet synchronous motor platform.