ABSTRACT
The traditional method for diagnosing the health status of shield tunneling machines mainly involves manual inspection and regular maintenance. This method is not only inefficient, but also difficult to accurately determine the real-time health status of shield tunneling machines. Therefore, the research of shield machine health state tracking and diagnosis method based on machine learning is proposed in this paper. Firstly, based on machine learning algorithms, design a health status tracking and diagnosis system for shield tunneling machines. Secondly, collect operational data of shield tunneling machines using sensor technology to provide a solid data foundation for subsequent status monitoring. Finally, based on the decision tree in recent learning algorithms, real-time tracking and diagnosis of the health status of the shield tunneling machine is carried out, and the tracking results of the healthy operation status of the shield tunneling machine are output, thereby completing the entire diagnostic process. The experimental results show that the machine learning based method for tracking and diagnosing the health status of shield tunneling machines can accurately determine the frequency of abnormal vibration states of shield machine bearings, while the diagnostic errors of the other two methods are relatively large. Therefore, the experiment can prove the effectiveness of this method.
