As a widely used industrial field equipment, the safety and reliability of permanent magnet synchronous motors are crucial for industrial production and life, so it is of extraordinary significance to quickly and accurately identify motor faults. Traditional fault diagnosis methods mostly rely on a single parameter or empirical judgment, which has certain limitations. In this paper, we propose an innovative method to construct a permanent magnet synchronous motor model based on the Bagging algorithm in integrated learning, realize the fault diagnosis of permanent magnet synchronous motors by decision fusion diagnosis of multiple fault features, and optimize the dataset to further improve the prediction accuracy of the model as well as the rate of rapid measurement. The experimental results show that the Bagging algorithm is more accurate than other integration algorithms, and the accuracy and efficiency of the data-optimized model are greatly improved. This verifies the feasibility of the proposed fault diagnosis method based on the Bagging algorithm, and the superiority of the data-optimized fault diagnosis method in terms of both accuracy and efficiency.