Abstract: Aiming at the problem of security risks caused by insufficient system life due to motor degradation in the 3D stage Wire system, a real-time system remaining useful life prediction method based on a multi-stage degradation model was proposed. Firstly, the multi-stage Wiener process is used to model the insulation aging of the servo motor, and the CUSUM is used to detect the change point to realize the stage division. Secondly, considering the influence of measurement noise, Kalman filtering, and RTS (Rauch-Tung-Striebel) smoothing algorithm are used to estimate the true degradation state, and the change point detection results are integrated into the parameter estimation process based on the expectation maximization algorithm to realize real-time online update and stage replacement of model parameters. Finally, using the real-time position error as the system performance evaluation index, the system-level online remaining useful life prediction of 3D stage Wire was realized. The simulation results show that the proposed method can not only accurately identify the change time of each stage of the degradation process, but also better align with the actual degradation process, which can effectively improve the prediction accuracy of the remaining life, so as to provide a reliable basis for the subsequent precise maintenance.

Key words: 3D stage Wire system, multi-stage degradation modeling, change point detection, system-level real-time life prediction