Integrated multimethod analysis of miners' safety behavior and risk interaction for practical applications.

Miners' safety behavior is a critical factor in mine safety management, and its implementation effectiveness significantly influences the establishment of safety ecosystems and sustainable industry development. This study constructs a multidimensional risk coupling analytical framework integrating Bayesian Network (BN) modeling, N-K theory, and Random Forest (RF) algorithms based on risk coupling theory. By systematically examining the three primary risk factors influencing miners' safety behaviors-individual perception, situational stress, and physical environment-the analysis reveals that the synergistic interaction between situational stress and physical environment is the principal pathway for systemic risk emergence. Furthermore, multidimensional factor coordination demonstrates significant potential in enhancing safety compliance rates. Heterogeneous behavioral patterns among miners were identified through cluster analysis, informing the development of customized intervention strategies. The study proposes three strategic recommendations: (1) Introducing differentiated risk classification protocols, (2) Optimizing intelligent early-warning response systems, and (3) Rebuilding safety culture ecosystems. These recommendations facilitate the transition from reactive regulatory compliance to proactive safety adaptation in mine operations. This research contributes novel perspectives to safety management research and establishes a systematic modeling foundation for optimizing miners' safety behavior interventions. The integrated methodology advances current analytical approaches in occupational risk assessment and provides actionable insights for industrial safety practices.