Demagnetization faults are a unique type of fault for PMSMs, and their occurrence can have a serious impact on the normal operation of PMSMs. In this paper, the Subtraction-Average-Based Optimizer is used to optimize the Variational Modal Decomposition algorithm and the SABO-VMD algorithm is proposed. In this paper, a PMSM demagnetization fault diagnosis method based on SABO-VMD and SVM is proposed. Firstly, the mathematical model of demagnetization fault of PMSM is established to simulate its demagnetization fault, and the characteristic quantity of demagnetization fault is analyzed. Secondly, the SABO-VMD algorithm is proposed to optimize the extracted fault features. Finally, the SVM algorithm and the optimized fault features are used to diagnose the demagnetization fault. In this paper, the effectiveness of the proposed fault diagnosis method is verified through simulation, and its diagnostic effectiveness is compared with that of other machine learning methods to verify the superiority of this method.