Abstract
To ensure the safe and stable operation of underground coal mining machines and accurately predict the structural mechanical behavior of their critical components, a structural performance monitoring method of the shearer cutting unit transmission system driven by digital twin is proposed, focusing on the output shaft gear as the research subject. First, a five-dimensional digital twin framework for the shearer is proposed, tailored to its operational characteristics. Subsequently, a shearer cutting test bench is constructed to collect roller load and shearer motion state data synchronously; the multi-body dynamics model and fatigue life model of the shearer cutting unit transmission system are established by using the discrete element method, coupled rigid-flexible dynamics, and fatigue analysis technology. Building upon this foundation, structural response data of gears under various operating conditions are collected, the dataset was constructed; multiple single surrogate models were established and trained, machine learning and neural network models were further integrated to develop hybrid surrogate model based on feature extraction, and the training results of the model are compared and analyzed. The results indicate that the convolutional neural network-gradient boosting regression hybrid surrogate model outperforms other models in prediction accuracy and prediction fitting degree. Finally, a digital twin monitoring system for the coal mining machine was developed. The system enables monitoring of the shearer’s motion state, as well as prediction of stress conditions and remaining useful life for critical gears, offering new insights for intelligent health monitoring and predictive maintenance of coal mining equipment.
Funding Information
  • The National Natural Science Foundation of China (52204149)
  • Fundamental Research Program of Shanxi Province (202203021221051)

This publication has 22 references indexed in Scilit: