Abstract: Expressway network connects all regions in China, and critical path identification is of great significance to ensure the reliable operation of expressway network. The traditional critical path analysis method is based on topology and does not consider the traffic volume characteristics of the traffic network. The existing analysis methods based on traffic volume data only consider the characteristics of some routes, which is difficult to reflect the actual operation of the expressway network. Using the route traffic volume data to build a random matrix model of traffic volume, according to the change characteristics of traffic volume after the expressway network is abnormal, the critical path evaluation index is defined to realize the quantitative evaluation of the degree of abnormal impact. On this basis, a data-driven critical path identification method for expressway network is proposed. Finally, the rationality and effectiveness of the proposed method are verified by analyzing the expressway network in Liaoning Province, and the method is applied to the urban road network case, which further proves the universality of the method.

Key words: transportation, critical path identification, data-driven, complex network, random matrix theory

摘要： 高速公路网络是我国各地区相互连接的重要纽带，高速公路网络关键路径辨识对确保高速网络的可靠运行具有重要意义。传统的关键路径分析方法基于拓扑结构，未考虑交通网络的运输量特性；而现有的基于运输量数据的分析方法只考虑部分路径的运输量特性，难以反映交通网络的实际运行情况。利用路径运输量数据，搭建运输量随机矩阵模型，针对高速公路网络异常后的运输量变化特性，定义关键路径评估指数，实现异常影响程度的量化评估，在此基础上提出一种基于数据驱动的高速公路网络关键路径辨识方法。最后，采用辽宁省高速公路网络进行分析，验证了所提方法的合理性和有效性，并将该方法应用于城市路网案例中，进一步证明该方法具有普适性。

关键词: 交通运输, 关键路径辨识, 数据驱动, 复杂网络, 随机矩阵理论