基于改进认知可靠性与失误分析方法的隧道驾驶人因可靠性分析模型

为有效预测驾驶人通过隧道时的人因失误概率，减少人为差错，提出一种基于改进认知可靠性与失误分析方法（cognitive reliability and error analysis, CREAM）的隧道驾驶人因可靠性分析模型。结合隧道驾驶任务的特点，调整共用绩效条件（common performance condition,CPC），给出评估细则。使用模糊数将CPC因子的绩效效应值从确定值转换为概率值，以降低主观因素对评价结果的影响。考虑CPC因子间的相互影响关系，给出调整规则。最后建立基于贝叶斯网络的改进CREAM模型，确定控制模式。对隧道驾驶实例进行分析，计算驾驶人完成隧道驾驶任务的人因失误概率。结果表明，该模型可以准确计算隧道驾驶任务中的人误概率，且灵敏性比CREAM基本法更高，可为隧道交通人因可靠性分析提供支持。

Human Reliability Analysis Model for Tunnel Driving Based on Improved CREAM Method

In order to effectively predict the probability of human error when the driver passes through the tunnel and reduce human errors.A human reliability analysis model for tunnel driving based on improved CREAM (cognitive reliability and error analysis) was proposed.Firstly, combined with the characteristics of driving tasks, improve the CPC (common performance condition) factors and give the evaluation rules.Then the fuzzy scoring was used to change the performance effect value of CPC factor from the determined value to the probability value, so as to reduce the influence of subjective factors on the evaluation results. Considering the interaction between CPC factors, the adjustment rules were given. Finally, an improved CREAM model based on Bayesian network is established to determine the control model.According to the analysis of the driver passing through the tunnel, the human error probability of the driver when completing the task was calculated.The results show that the model can accurately calculate the human error probability in tunnel driving task, and its sensitivity is higher than CREAM basic method, it can provide support for human reliability analysis of tunnel traffic.