An Intelligent Fault Diagnosis Framework for Permanent Magnet Synchronous Motors Based on Dual-Source Time-Frequency Fusion and Multi-Scale Channel-Spatial Attention

Reliable fault diagnosis of permanent magnet synchronous motors (PMSMs) under complex industrial environments remains challenging due to weak fault signatures, strong background noise, and pronounced non-stationary characteristics. Conventional diagnostic methods based on single-domain features or fixed-scale analysis often exhibit limited robustness and adaptability under such conditions.