基于HFACS-BN的炼化企业动火作业人因风险分析OA

In order to explore the core human factors influencing the safety risks of hot work in refining and chemical enterprises,a Human Factors Analysis and Classification System(HFACS)index system adapted to these specific risks was constructed.A novel approach integrating the fuzzy evaluation method with a Bayesian Network(BN)model was employed.Specifically,fuzzy numbers and natural language fuzzy evaluations were u-tilized to calculate the prior probabilities of the root nodes and the conditional probability tables(CPTs)within the BN model.Subsequently,GeNIe software was used to conduct bidirectional inference simulations and root node sensitivity analyses,thereby achieving the quantitative tracing of human factors risks in hot work and the ranking of root node factors.The results indicated that the inadequate establishment and execution of work per-mit regulations,insufficient identification and control of operational safety risks,lax auditing of construction per-sonnel and equipment,and inadequate safety investment were the primary human factors exacerbating hot work safety risks.Consequently,prioritized risk control and management measures based on the sensitivity ranking are proposed to enhance operational safety.