Research on data driven coal mine gas explosion risk assessment system
LI Shuang
LU Cheng
XU Ningke
ZHANG Yi
ZHANG Zhen'an
Abstract：Gas explosion is a destructive and frequent major disaster type in coal mines,and the ex-isting risk assessment system still has shortcomings in accuracy,objectivity,and real-time perfor-mance.In response to the practical needs and challenges in coal mine gas explosion risk assess-ment,this article constructs the data-driven coal mine gas explosion risk assessment system.First,a three-level indicator system covering multiple dimensions such as monitoring,manage-ment,hidden danger status,and environmental conditions was constructed by using the hybrid-driven method integrating text mining and disaster mechanisms,which enabled real-time collection and expression of multi-source heterogeneous data.Second,scoring rules for third-level indicators and threshold rules for second-level indicators were formulated based on domain knowl-edge to provide standardized support for model inputs.Third,a multi-level and multi-model fusion risk assessment method was developed by integrating FAHP,CRITIC,linear weighting models,end-point mixed triangular whitenization weight functions,and risk level-risk value transformation models,achieving a balance between interpretability and data processing capabilities.Finally,in-dustrial testing was conducted on coal working faces and return airways of three types of coal mines:low-gas,high-gas,and outburst mines.The results indicate that the proposed assessment system can integrate multi-source heterogeneous data in real time,enabling risk assessment across multiple regions and levels with an accuracy rate of 98.3%.Comparative analysis demonstrates that its as-sessment accuracy is improved by 1.6%and 8.7%respectively compared to single weighting method and traditional grey clustering model.The system achieves real-time,accurate,and adap-tive coal mine gas explosion risk assessment under various working conditions,significantly outper-forming traditional models and demonstrating strong industrial applicability and promising prospects for widespread adoption.
Keywords：coal mine gas explosion risk assessmentdata-drivenmulti source heterogeneous dataindustrial applicationintelligent control
Publication Date：2026-03-31
Online Publishing Date：2026-04-08(First online date of this platform, not the publication date of the document)
Pages：20( 350-369 )