Performance diagnosis of the drive unit for high speed motorized spindle based on hidden state mapping model

Abstract As the key component of the motorized spindle system, the performance of drive unit will directly affect the processing quality of CNC machine tools. This paper proposed a performance diagnosis method of hidden state mapping model to prevent the motorized spindle from failure. The hidden state model was obtained by feature extraction of the parameter vectors identified from the dynamic responses to test input signals. The thresholds of performance degradation were determined by the hidden state to evaluate the performance of the drive unit in the motorized spindle system. The feasibility and effectiveness of the model can be verified by taking IGBT degradation failure as an example. The results showed that the model parameters extracted from different degradation degrees of IGBT devices can be mapped to hidden state characteristics and used for performance evaluation of the drive unit. The proposed hidden state mapping model can be used in the framework of monitoring the safety and reliability of the drive unit, and lay the foundation for the prognostic and health management of the motorized spindle system.