Modeling and analysis of open-pit coal mine accident causation based on directed weighted network

Abstract

Open-pit coal mining is a complex process with a broad operational scope, increasing the risk of accidents and posing management challenges. This study presents a directed weighted network modeling approach based on accident cases. The approach integrates grounded theory, event chain analysis, and complex network theory to construct the open-pit coal mine accident causation network (OPCMACN). The OPCMACN encompasses causal nodes across five dimensions: human, equipment, environment, management, and technology, along with accident nodes, illustrating their complex interconnections. A topological analysis framework suitable for directed weighted networks is proposed to analyze the structure of the OPCMACN. By considering four dimensions: node neighbors, path hubs, random walks, and positional information, topological metrics suitable for directed weighted networks are used to identify key causal factors, enabling the recommendation of targeted preventive measures. Furthermore, a comprehensive accident causation governance approach (CACGA) is introduced, integrating the advantages of various topological metrics across different stages of causal factor governance. The robustness analysis reveals significant vulnerabilities in the OPCMACN when key nodes are governed while confirming the superiority of CACGA throughout the entire governance process. The research findings provide essential theoretical support for decision-making in managing the safety of open-pit coal mines and offer a comprehensive, novel perspective for accident analysis in other system safety fields.