Author(s)

Yilong Xiao ¹, Jidong Yao ¹, Ruiyuan Su ¹, Zecheng Zhu ¹, Linxin Du ¹, Xu Xia ¹

Affiliation(s)

¹ College of Artificial Intelligence, North China University of Science and Technology, Tangshan, China

Corresponding Author

Yilong Xiao

ABSTRACT

This study proposes a stacking-SHAP method for predicting spontaneous combustion temperature, aiming to improve prediction accuracy and explain the decision-making process of the black-box model in order to develop targeted solutions for different spontaneous combustion fire scenarios. The method first performs data preprocessing and the construction of composite indices, then uses the Grey Wolf Optimization (GWO) algorithm for hyperparameter optimization of the base learners in the stacking fusion model, and finally predicts the test set. Experimental results show that the stacking model achieves a coefficient of determination (R²) of 0.989, an average absolute error (EMA) of 6.003, a mean squared error (EMS) of 56.708, and an average logarithmic error (EMAP) of 6.34%. Comparison with the three base learners (GBDT, RF, and XGBoost) indicates that the stacking model outperforms them in terms of prediction accuracy and generalization ability. SHAP is used to interpret the stacking model, revealing the five most influential features on the prediction of spontaneous combustion temperature in the order of their impact: CO > O2/CO > CO2/O2 > CO2 > CO2/CO. Finally, ablation experiments confirm the accuracy of the SHAP interpretation method.

KEYWORDS

Coal Spontaneous Combustion Temperature Prediction, Interpretable Fusion Model, Machine Learning, SHAP Interpretation Method, Grey Wolf Optimisation Algorithm

CITE THIS PAPER

Yilong Xiao, Jidong Yao, Ruiyuan Su, Zecheng Zhu, Linxin Du, Xu Xia, Coal Spontaneous Combustion Temperature Prediction Based on an Interpretable Fusion Model. Modern Physical Chemistry Research (2025) Vol. 5: 1-12. DOI: http://dx.doi.org/10.23977/mpcr.2025.050101.

REFERENCES

[1] Jinhu Li, Zenghua Li, Yongliang Yang, Chaojie Wang, Liutao Sun. Experimental study on the effect of mechanochemistry on coal spontaneous com bustion[J]. Powder Technology, 2018, 339: 102-110. 
[2] Deng Jun, Zhang Qi, Chen Weile, et al. Monitoring and early warning technology and development trend of spontaneous coal combustion disaster in mines [J]. Coal Mine Safety, 2024, 55(03):99-110. DOI:10. 13347/j. cnki. mkaq. 20231578. 
[3] Cao F. R., Wu X. S., Li J., et al. Prediction of spontaneous combustion temperature of coal with multi-gas indicator based on machine learning[J]. Coal Mine Safety, 2024, 55(04):106-113. DOI:10. 13347/j. cnki. mkaq. 20231426. 
[4] Jun Deng, Zujin Bai, Yang Xiao, et al. Current status and challenges of coal spontaneous combustion disaster prevention technology[J]. Coal Mine Safety, 2020, 51(10):118-125. DOI: 10. 13347/j. cnki. mkaq. 2020. 10. 018. 
[5] Zhai Xiaowei, Cheng Zhuo, Experimental study on the preference of spontaneous combustion marker gas in 4~(-2) coal seam of Chaijiagou Mine[J]. Coal Mine Safety, 2019, 50(11):18-23. 
[6] AN Jingyu, ZHANG Jian, JI Dege et al. Optimisation of coal spontaneous combustion indicator gases based on segmental fitting [J]. China Safety Production Science and Technology, 2021, 17(10):25-31. 
[7] Jiang H. Research on multi-parameter gas main and auxiliary prediction index system of coal spontaneous combustion [J]. China Mining Industry, 2023, 32(04):147-153. 
[8] Hao Tianxuan, Zhang Longlong, Li Fan et al. Preferential selection of coal spontaneous combustion marker gases based on principal component analysis [J]. Coal Technology, 2023, 42(10):176-181. 
[9] Deng Jun, Lei Changkui, Cao Kai, et al. Support vector regression method for coal spontaneous combustion prediction [J]. Journal of Xi'an University of Science and Technology, 2018, 38(02):175-180. 
[10] ZHOU Xu, WANG Zhizhuo, Dai Yaxun, et al. Research on graded warning of coal spontaneous combustion based on BO-XGBoost[J]. Coal Engineering, 2022, 54(08):108-114. 
[11] Lei C, Deng J, Cao K, et al. A random forest approach for predicting coal spontaneous combustion[J]. Fuel, 2018, 223. 
[12] Kong B, Niu S, Cao H, et al. Study on the application of coal spontaneous combustion positive pressure beam tube classification monitoring and early warning[J]. Environmental Science and Pollution Research, 2023: 1-17. 
[13] Kong Biao, Zhu Si, Hu Xiangming, et al. Research on coal spontaneous combustion prediction based on optimised BP neural network with improved whale algorithm[J]. Mining Safety and Environmental Protection, 2023, 50(05):30-36. 
[14] Gao Yunwei, Luo Limin, Xue Fenglong, Liu Yang, Yan Hao, Zheng Shuangjin. A mechanical drilling speed prediction method based on Stacking integrated learning [J/OL]. Petroleum Machinery. 2024, 52(05)17-24 
[15] Wei Yixi, Hu Dan, Xu Yanhong, et al. Low sidelobe and wide-zero beam synthesis based on grey wolf algorithm[J/OL]. Radio Engineering, 1-10[2024-11-03]. http://kns. cnki. net/kcms/detail/13. 1097. tn. 20240925. 1424. 002. html. 
[16] Gao Rui, Zhang Zhenbi, Li Lichen, et al. An interpretable machine learning prediction method for landslide stability based on BO-SVM[J]. Resource Environment and Engineering, 2024, 38(03):274-284+307. DOI:10. 16536/j. cnki. issn. 1671-1211. 2024. 03. 005. 
[17] Zan Ziyi, Yue Weifeng, Zhao Hangzheng, et al. Construction of agricultural water consumption prediction model based on machine learning and SHAP algorithm[J/OL]. Water Saving Irrigation, 1-16[2024-11-03]. http://kns. cnki. net/kcms/detail/42. 1420. TV. 20241017. 1204. 006. html. 
[18] Hou S K, Liu Y R. Early warning of tunnel collapse based on Adamoptimised long short-term memory network and TBM operation parameters[J]. Engineering Applications of Artificial Intelligence, 2022, 112: 104842.

Sponsors, Associates, and Links

---

• Forging and Forming
  
  
• Composites and Nano Engineering
  
  
• Journal of Materials, Processing and Design
  
  
• Metallic foams
  
  
• Smart Structures, Materials and Systems
  
  
• Chemistry and Physics of Polymers
  
  
• Analytical Chemistry: A Journal
  
  
• Inorganic Chemistry: A Journal
  
  
• Organic Chemistry: A Journal
  
  
• Progress in Materials Chemistry and Physics
  
  
• Transactions on Industrial Catalysis
  
  
• Fuels and Combustion
  
  
• Casting, Welding and Solidification
  
  
• Journal of Membrane Technology
  
  
• Journal of Heat Treatment and Surface Engineering
  
  
• Trends in Biochemical Engineering
  
  
• Ceramic and Glass Technology
  
  
• Transactions on Metals and Alloys
  
  
• High Performance Structures and Materials
  
  
• Rheology Letters
  
  
• Plasticity Frontiers
  
  
• Corrosion and Wear of Materials
  
  
• Fluids, Heat and Mass Transfer
  
  
• International Journal of Geochemistry
  
  
• Diamond and Carbon Materials
  
  
• Advances in Magnetism and Magnetic Materials
  
  
• Advances in Fuel Cell
  
  
• Journal of Biomaterials and Biomechanics