Analysis of the causes of minor injury accidents in coal mines based on improved HFACS model and GA−ANN optimized algorithm

To deeply analyze the manifestation forms and mutual coupling relationships of human−caused minor injury accidents in coal mines, this study takes 87 minor injury accidents that occurred in Huainan Mining Group from 2022 to 2023 as the research object. The CM−HFACS (Human Factor Analysis and Classification System of Coal Mine) accident causation model is improved from five levels: coal mine organizational influence, unsafe supervision, unsafe behavior premise, unsafe behavior, and accident emergency response. The GA−ANN (Genetic Algorithm−Artificial Neural Network) optimization algorithm is used to mathematically model the causation factors, and node degree, centrality, clustering coefficient, and importance analysis are conducted. The results show that: (1) Dense subnets in the high connection degree nodes are the key manifestation form of the accident cause. (2) Insufficient safety fund investment, unreasonable personnel allocation, and non−standard protective equipment are the "traffic bottlenecks" for the spread of the accident. (3) Insufficient safety fund investment, inadequate on−shift supervision and inspection, and unreasonable personnel allocation are the sources or positions in the middle of the accident chain. (4) Unreasonable self−disposal, delayed injury reporting, inadequate safety rewards and punishments, risk judgment based solely on experience, failure to formulate and implement relevant control measures, impatience, failure to report accidents according to procedures, and misoperation of equipment are the "powder kegs" that expose risks and have local transmission potential. (5) Failure to notice approaching danger, unfamiliarity with job operations, failure to report risks in a timely manner after identification, and failure to identify relevant job risks reveal the key factors of loss of control in the accident. (6) Inadequate skills for the job, unreasonable self−disposal, and violation of command are the "high−risk groups" that trigger accidents. (7) delayed injury reporting, fatigue work, maintenance without shutting down equipment, and low−quality pre−shift meetings are the key manifestation forms of causation with the highest comprehensive ranking. This research can provide support for the precise prevention and control of minor injury accidents in coal mines.