基于贝叶斯网络的全自动运行系统弹性评估

结合列车全自动运行系统的弹性评估，提供一种基于贝叶斯网络模型的评估方法。根据列车故障前后的 性能变化曲线，总结系统弹性的 4 种能力，探究影响这些能力的相关因素，为影响因素的定量选取合适的评价指 标，作为贝叶斯网络的最底层节点。对指标制定打分表，获取专家的打分结果作为指标层的先验概率输入，同时 选择恰当的隶属度公式作为条件概率的计算方法，获得贝叶斯网络的条件概率表，建立列车全自动运行系统的贝 叶斯网络模型，定量评估系统弹性，同时对目标节点进行敏感度分析，提出提升弹性的改进建议。

Resilience Evaluation of a Fully Automatic Operation System Based on a Bayesian Network

To further improve the safety level of fully automated operation systems, an evaluation method based on a Bayesian network model is proposed for resilience evaluation of an automatic train operation system. Four types of system resilience capabilities were summarized based on the performance change curve before and after train failure. The factors influencing these four capabilities are explored, and an appropriate evaluation index is selected as the bottom node of the Bayesian network. Furthermore, an appropriate membership formula is selected for the calculation of conditional probability. In addition, the conditional probability table of the Bayesian network is obtained, and the Bayesian network model of the automatic train operation system is established to quantitatively evaluate system resilience. Concurrently, a sensitivity analysis of target nodes was conducted. Methods to improve resilience were determined based on the analysis results.