Real-time precision reliability prediction for the worm drive system supported by digital twins

Worm drive systems are widely used in machine tools, metallurgical machinery, and ship engines due to the high precision, large transmission ratio and unique structure. Currently, neither theoretical simulation methods nor condition monitoring methods can achieve satisfactory results for precision reliability prediction, which will lead to unreasonable maintenance services. Therefore, a hybrid approach supported by digital twins is proposed for real-time precision reliability prediction of the worm drive system. Firstly, a description of the five-dimension DT architecture is provided to illustrate the process of predicting precision reliability. Secondly, virtual mirror is established by combining multiple physical models, including the initial precision evaluation model of the worm drive system, the numerical model for the thermal strain, the collection of test and historical data. Physical entity is designed to obtain the real-time perception data. Finally, the Wiener process-based degradation model is employed to describe the precision evolution process and handle the hybrid-approach data. Service system is built to develop the prediction steps for the precision reliability of the worm drive system. The comparison between the predicted results driven by different methods is proposed to verify the accuracy and effectiveness of proposed method.