安全与环境学报

中欧班列运输网络中关键节点风险影响班列线路规划和运营安全。通过分析中欧班列运输节点的物理网络拓扑特性，结合考虑服务网络中节点服务能力的重要性，提出熵-逼近理想解排序法(Technique for Order Preference by Similarity to an Ideal Solution, TOPSIS)节点重要度评估方法。考虑节点受外部宏观因素影响可能失效的情况，引入国家运输支撑力衡量节点失效可能，提出重要性-失效可能分析模型(Importance-Failure Possibility Analysis, IFPA)识别中欧班列关键运输节点风险，并采用K-Means++聚类算法对节点进行风险划分。通过构建中欧班列全域运输网络进行案例分析，结果表明该模型可有效识别节点风险且节点风险分类结果符合实际情况。高、中、低风险区和风险监测区分别对应5、10、7和6个关键运输节点。对运输节点进行风险分类有助于分级管理中欧班列运输风险，保障运输安全。

To identify the risk of key nodes in the network of China Railway Express(CRE) which affects the route planning and operation safety, this paper firstly analyzed the physical network topology characteristics of CRE transportation nodes and the importance of node serviceability in the service network. An entropy-TOPSIS node importance evaluation method was proposed. Secondly, considering the possibility of node failure caused by external macro factors, the national transportation supportability was introduced to measure the possibility of node failure. A novel Importance-Failure-Probability-Analysis(IFPA) model was proposed to identify the risk of key transportation nodes. Then, K-Means++ clustering algorithm was applied in risk analysis to divide the nodes into four risk types, that were high-risk zone, medium risk zone, low-risk zone, and risk monitoring zone. Finally, a global CRE transportation network was constructed as an empirical analysis. Through the case study, the results show that the model can effectively identify the node risk, and the node risk classification results are in accord with reality. High-risk zone, medium risk zone, low-risk zone, and risk monitoring area correspond to 5, 10, 7 and 6 key transport nodes respectively. The risk classification of transportation nodes is helpful to manage the transportation risk of CRE at different levels and ensure transportation safety. Different risk management strategies should be adopted for nodes in different risk zones. Besides, according to the clustering results, the possible value of node failure of 0.4 and the node importance score of 0.4 can be used as the dividing thresholds of risk areas respectively. The thresholds can provide practical guidance for relevant departments to quickly classify risk nodes and formulate targeted management and control strategies. In a word, the risk identification and classification of transportation nodes in this study is helpful to manage the transportation risk of CRE at different levels and ensure transportation safety.