Novel Online Noninvasive Diagnosis and Classification Approach for Main Shaft Faults in Electric Drive System Based on Torsional Resonance Characteristics

The urgency of maintaining different types of main shaft faults varies due to differences in fault development speed and system fault tolerance. Shaft mechanical parameters contained in shaft torsional resonance characteristics can reflect different faults. However, traditional measurements of shaft resonance characteristics are offline and invasive, which is not conducive to online applications. Thus, a novel online noninvasive diagnosis and classification approach for main shaft faults based on sweep torque injection is proposed. First, a complex frequency domain-space model of shaft system torsional vibration considering motor control is established, interpreting resonance peak of each order, and further investigating the impact of different main shaft faults on resonance characteristics. Then, an online extraction scheme for main shaft resonance characteristics based on sweep torque active injection by drive motor is proposed, and the classification of main shaft faults is achieved by the K-nearest neighbor (KNN) algorithm. The feasibility and accuracy of the proposed model and method have been verified through experiments. The experimental results show that the classification accuracy of the method for different main shaft faults is around 99.0%. Compared to traditional shaft fault classification techniques based on vibration, the proposed method has higher accuracy.