To address the issues that permanent magnet synchronous motors (PMSMs) are prone to faults under complex operating conditions and traditional diagnostic methods have deficiencies in multi-sensor information fusion and robustness in noisy environments

a fault diagnosis method for PMSMs based on multi-signal Gramian angular difference field (MGADF) fusion is proposed.This method first acquires vibration acceleration signals from multiple locations of the motor

then uses a newly designed MGADF technique to fuse the multi-channel sensor signals into a single image

extracts multi-texture features from the fused image.Through fault feature learning and classification comparisons on different machine models

the results show that the proposed method achieves an average diagnostic accuracy of 99.54% and a standard deviation of 0.19

maintains excellent performance and high robustness even in high-noise environments

and significantly reduces the risk of overfitting.The conclusion indicates that the proposed non-invasive method has high diagnostic accuracy and strong anti-interference capability and can be effectively extended to condition monitoring and intelligent diagnosis of industrial motors.