针对冲偏出跑道事件调查未全面揭示人为因素之间的复杂因果关系问题，提出了一种航空器冲偏出跑道事件关键人为因素分析方法。首先，改进人因分析和分类系统模型，用以识别冲偏出跑道的根本原因，基于事故树分析方法识别冲偏出跑道的直接原因。其次，构建人为因素调查决策支持模型，得到航空器冲偏出跑道事件人为因素致因链条。然后，构建冲偏出跑道事件人为因素的贝叶斯网络，通过预测推理、诊断推理及灵敏度分析量化致因因素的重要程度，得到冲偏出跑道事件的关键人为因素。结果表明：基于人为因素调查决策支持模型得到32条冲偏出跑道事件致因链条，全面揭示该事件的人为因素及其耦合关系；基于贝叶斯推理得到该事件的关键人因致因链条为“动作执行不当／信息预处理不当→不安全行为→冲偏出跑道”。以上结论与航空器冲偏出跑道事件调查结论基本一致，对航空器冲偏出跑道事件预警和防范能力的提升有积极意义。

A critical human factors analysis method for aircraft runway overrun and veer-off event was proposed to address the complex causal relationship of human factors that have not been fully revealed in the investigation of the event. Firstly, improve the human factors analysis and classification system model to identify the root cause of overrun and excursion event, and use the fault tree analysis method to identify the direct cause of runway overrun and veer-off event. Secondly, a human factors investigation decision support model was constructed, and obtained the causal chain of aircraft runway overrun and veer-off event. Then, a Bayesian network was constructed to quantify the importance of the causal factors for the runway overrun and veer-off event caused by human factors. Through predictive reasoning, diagnostic reasoning, and sensitivity analysis, the key causal chain of the runway overrun and veer-off event was obtained. The results indicate that 32 causal chains of overrun and veer-off events is obtained based on the proposed human factors investigation decision support model, comprehensively revealing the human factors and their coupling relationships of the event. Bayesian inference can obtain the key causal chain of "improper action execution/improper information preprocessing → unsafe behavior → runway overrun and veer-off". The above conclusions are basically consistent with the findings of investigations into aircraft runway overrun and veer-off events, and they hold positive significance for improving the early warning and prevention capabilities of such incidents.