Research on temperature prediction of shearer cable based on bidirectional long short-term memory
Lijuan Zhao,Guocong Lin,Yadong Wang,Bo Xie,Chuanxu Wan,Hongqiang Zhang,Shuo Tian,Zhongjian Bai,Meichen Zhang,Xin Jin
DOI: https://doi.org/10.1016/j.ijthermalsci.2024.109597
IF: 4.779
2024-12-04
International Journal of Thermal Sciences
Abstract:This study investigates the temperature behavior of the MCP-0.66/1.14 (3∗95 + 1∗35 + 4∗10)shearer cable, focusing on the effects of ambient temperature and load current. A 3D simulation model was created using the Rhino-Grasshopper toolset, and a bidirectional coupled electromagnetic-thermal simulation model was developed based on electromagnetic and thermal field theories. The temperature changes in the cable were analyzed by varying the ambient temperature and load current. Results indicated that as ambient temperature increased from −20 °C to 50 °C, the cable temperature rose. With a constant environment, when load currents increased to 250A and 63A, the temperatures of the power wire core, sheath, control wire core and outer surface increased accordingly. A Bidirectional Long Short-Term Memory (Bi-LSTM) neural network was used to predict cable temperatures. Compared to other models, Bi-LSTM achieved superior performance in terms of Mean Absolute Error (0.1793 °C), Mean Squared Error (0.4104 °C), Coefficient of Determination (0.9907), and processing time (45.4386 ms). The model's predictions were validated against experimental data, showing a maximum error of 1.87 % for simulation predictions and 2.28 % for experimental comparisons. These results confirm the accuracy of the simulation model and temperature prediction algorithm, offering insights for enhancing shearer cable performance and developing mining cable temperature prediction systems.
engineering, mechanical,thermodynamics