基于证据推理的改进加权CREAM模型

Improved Weighted CREAM Model based on Evidential Reasoning

为弥补传统认知可靠性与失误分析方法(CREAM)的缺陷,提高人因失误概率(HEP)的计算精度,提出一种基于证据推理(ER)的改进CREAM模型。该模型首先通过决策实验室分析法(DEMATEL)和模糊层次分析法来分别计算共同绩效条件(CPC)的相关性权重和重要性权重;然后利用组合赋权法将两种权重线性组合,获得CPC的综合权重,在此基础上利用ER算法对CPC的绩效效应进行加权数据融合,并利用融合后的数据获取情景影响指数(CII),通过CII实现了HEP点值的输出;最后将该改进的CREAM模型应用到高速铁路列车调度员的人因可靠性定量评估案例中,以验证模型的有效性。案例评估结果表明:调度员在“设置临时限速”时的认知行为控制模式为战术型,HEP值为3.522×10^(-3),并通过与其他方法的对比分析,验证了该模型的有效性和可行性。

In order to overcome the shortcomings of the traditional Cognitive Reliability and Error Analysis Method(CREAM)and improve the calculation accuracy of the Human Error Probability(HEP),this paper proposes an improved CREAM model based on Evidential Reasoning(ER).The model first calculates the relevance and importance weight of the Common Performance Condition(CPC)through the Decision Making Trial and Evaluation Laboratory(DEMATEL)and the fuzzy analytic hierarchy process,and then linearly combines the two weights to obtain the combination of CPC Weights by the combination weighting method.On this basis,the paper performs weighted data fusion on the performance effects of CPC through ER algorithm,obtains the contextual impact index(CII)based on the fused data,and realizes the point value output of HEP through CII.Finally,the improved CREAM model is applied to quantitative assessment of human reliability of high-speed railway train dispatchers.The results show that the cognitive behavior control mode of the dispatchers when“setting temporary speed limit”is tactical,and the HEP value is 3.522×10^(-3).Through comparison with other methods,the paper verifies the effectiveness and feasibility of the model.