Open Access

ARTICLE

An Improved Support Vector Machine Method for Fault Diagnosis of Inter-Turn Short Circuit in PMSM with Enhanced Fault Representation

Yue Su¹, Shukuan Zhang^1,*, Jinghao Jiao¹, Jiankang Zhong², Qianxi Zhao¹

1 College of Marine Electrical Engineering, Dalian Maritime University, Dalian, China
2 Sichuan Key Technology Engineering Research Center for All-Electric Navigable Aircraft, Guanghan, Deyang, China

* Corresponding Author: Shukuan Zhang. Email:

Computer Modeling in Engineering & Sciences 2026, 147(1), 21 https://doi.org/10.32604/cmes.2026.079927

Received 30 January 2026; Accepted 08 April 2026; Issue published 27 April 2026

Abstract

This paper introduces a novel dual-layer optimization fault diagnosis framework for inter-turn short-circuit (ITSC) faults in permanent magnet synchronous motors (PMSMs). The synergistic of a SABO-optimized VMD for enhanced feature extraction and an MFO-optimized SVM for intelligent classification is proposed. Firstly, mathematical and simulation models of ITSC faults in PMSMs are established to obtain fault phase currents and motor electromagnetic torques as characteristic fault signals. Then, the SABO algorithm is used to optimize the VMD parameters, followed by VMD decomposition of the characteristic fault signals to obtain Intrinsic Mode Functions (IMFs), and the time-domain parameters of the optimal IMF are calculated to obtain feature vectors. Finally, the fault type is predicted using an SVM optimized by the Moth-Flame Optimizer (MFO). Simulation results show that the accuracy of fault diagnosis can reach 93.6%, indicating that the proposed method can achieve accurate diagnosis of ITSC faults and effectively improve the accuracy of fault diagnosis.

---

Keywords

---