基于改进 DFA 和 LDA 的永磁同步电机机械故障检测

In order to improve the detection accuracy, a mechanical faults detection method was studied for permanent magnet synchronous motors under variable speed conditions. Firstly, the vibration characteristics of the bearing, the eccentricity, and the compound faults were analyzed. Secondly, the components of fault characteristic were extracted with Vold-Kalman arithmetic. And the extracted signals were reconstructed to remove the influence of the speed change on the components of fault characteristic. And then, a mechanical fault detection method was proposed based on improved detrended fluctuation analysis (DFA) and linear discriminant analysis (LDA) to realize the reconstructed signal feature extraction and fault detection. Finally, a verification experiment was carried out for the proposed fault detection method. The results show that the detection accuracy of the proposed fault detection method can reach up to 88%.