Abstract
In order to achieve the causal diagnosis of life health of high-speed spindle system, a digital twin system is designed and developed based on the combined programming of MATLAB, ANSYS, and LabVIEW. The domain and threshold models of heat generations of bearing and motor are established to diagnose the health status of the spindle system. The Long short-term memory recurrent neural network is used to predict the heat generations of the spindle system. A causal diagnosis model is proposed to judge the fault source according to the sudden change of thermal characteristics. The experimental results show that the digital twin accuracy of thermal characteristics reaches over 95%, and the proposed digital twin-driven causal diagnosis system for life health can successfully monitor the health status of the spindle system.

This publication has 14 references indexed in Scilit: