基于FCE-SPA的智能化煤矿人机交互作业人因失误风险评价

Risk Evaluation of Human Error in Intelligent Coal Mine Human-Machine Interaction Operations Based on FCE-SPA

摘要: 人因失误作为煤矿事故的主要诱因，其科学评价与有效防控尤为关键。随着智能化煤矿建设的深入推进，作业人员失误模式呈现新特征。为客观评价智能化煤矿人机交互作业人因失误的风险，本文提出了一种融合组合赋权法、模糊综合评价法和集对分析的安全预防评价模型。该模型首先构建智能化煤矿人因失误行为影响因素的层次化结构；再基于博弈论思想，将改进层次分析法确定的主观权重与熵权法确定的客观权重进行组合赋权，并结合模糊综合评价法进行风险评级，接着引入集对分析对模糊评价结果进行势态解析；最后利用该模型对湖南省某市某一智能化煤矿的作业人员进行风险评价。研究结果表明：1) 智能化煤矿人机交互作业的主要风险源为安全责任制度、风险隐患管理和网安数据韧性等，其中安全责任制度最为关键；2) 实例应用显示，该煤矿风险得分为84.35分，属于低风险等级且综合联系度为0.6706呈强同势特征，揭示了系统低风险状态的内在稳健性；3) 评价结果与现场实况高度吻合，验证了该模型在智能化煤矿人因风险评估中的准确性与适用性。

Abstract: Human error is a primary cause of coal mine accidents; thus, its scientific evaluation and effective prevention are critical. With the continuous advancement of intelligent coal mine construction, operator error patterns are presenting new characteristics. To objectively assess the human error risks in human-machine interaction (HMI) operations within intelligent coal mines, this paper proposes a safety evaluation model integrating the Combination Weighting Method, Fuzzy Comprehensive Evaluation (FCE), and Set Pair Analysis (SPA). First, a hierarchical structure of factors influencing human error behaviors is constructed. Based on game theory, the model combines subjective weights determined by the Improved Analytic Hierarchy Process (IAHP) with objective weights from the Entropy Weight Method (EWM). It then employs FCE for risk grading and introduces SPA to analyze the connection trends of the evaluation results. Finally, the model is applied to assess operator risks at an intelligent coal mine in Hunan Province. The results indicate that: 1) The primary risk sources in HMI operations include the safety responsibility system, hazard management, and cybersecurity and data resilience, with the safety responsibility system being the most critical; 2) The case application reveals a risk score of 84.35 (Low Risk), and the comprehensive connection degree is 0.6706, exhibiting a “strong identical trend”, which demonstrates the inherent robustness of the system’s low-risk state; 3) The evaluation results align closely with actual on-site conditions, verifying the accuracy and applicability of the proposed model for human factor risk assessment in intelligent coal mines.