Demagnetization Fault Detection of Permanent Magnet Synchronous Motor Based on Continuous Variational Mode Decomposition and Random Forest

Aiming at the problem of slow speed and low diagnostic accuracy of uniform demagnetization fault diagnosis of permanent magnet synchronous motor, a uniform demagnetization fault diagnosis method based on continuous variational mode decomposition and random forest is proposed. Firstly, the uniform de magnetization fault model of the motor is established to obtain the stator current fault signal of the motor. Secondly, the Successive variational mode decomposition (SVMD) is used to decompose the current signal. Compared with the traditional variational mode decomposition method, SVMD improves the decomposition speed and can obtain the modal component faster. The random forest algorithm is introduced to train and test the algorithm based on the decomposed data, which reduces the correlation between the vectors, ensures that the random forest model can achieve high accuracy prediction in the multi-feature classification problem, and realizes the diagnosis of uniform demagnetization fault. Compared with the traditional support vector machine and other classification methods, the diagnosis accuracy is improved. Finally, the simulation results verify the effectiveness and accuracy of the proposed uniform demagnetization fault diagnosis method.