交通运输工程与信息学报

城市道路交通网络作为城市运行的重要基础设施，在交通系统中发挥着不可替代的作用。然而，随着全球极端天气频繁发生，城市道路交通网络往往最先受到影响，如暴雨引发的严重内涝、洪水、地震等自然灾害能够广泛地对城市内部道路交通系统施加严重的破坏，从而严重地影响城市内部正常运转，造成大量经济损失、负面社会影响等。本文以非常规事件下的城市道路交通网络为对象，针对事件后恢复阶段的路网构建韧性评估框架，综合考虑维修调度成本、路段重要度，以及路网修复过程中用户的出行行为，建立了非常规事件下路网最优修复策略双层规划模型，并设计了基于修复重要度的大邻域搜索启发式算法求解模型。最后运用Sioux Falls道路网络验证本文所提方法的可行性与有效性，通过分析各维修方案在不同资源环境下的路网性能和路段流量变化，得出在确定资源环境下的最优维修方案。结果表明，在一定数量的维修队约束下，该模型得到的路网修复策略能最大限度地提升路网韧性，在修复过程中路网性能呈阶梯式上升。调度费用权重系数影响修复策略的制定，维修队数量为3，调度费用权重系数取值为0.50时，路网的恢复效果最好，韧性最高。本文所提出的方法不仅为实际修复方案制定提供理论支撑，还为管理者在非常规事件下的决策提供建议。

Urban road-traffic networks are vital to transportation systems as they are critical infrastructures for city operations. Owing to the increased frequency of global extreme weather events,these networks are typically the first to be affected by natural disasters. Catastrophes, such as torrential rain-induced floods and earthquakes, can cause extensive damage to urban road systems, thus disrupting city functioning and exerting significant economic and social effects. This study focuses on urban road networks during recovery from post-unconventional events. A resilience-assessment framework is proposed that considers maintenance costs, road-segment importance, and traveler behavior during network restoration. A bilevel programming model for network-repair strategies under such events is established, in addition to a heuristic solution algorithm based on repair importance.The feasibility and effectiveness of the method are confirmed based on the Sioux Falls network case study, where the optimal repair strategy are determined within the specified resource constraints. The results indicate that network resilience can be enhanced significantly by adopting an appropriate repair plan, as indicated by the stepwise improvement in network performance during restoration. The number of maintenance teams and the weight of scheduling costs are key factors in strategy development. The optimal recovery and highest resilience are observed in three teams, in addition to a dispatch-cost weighting factor of 0.50. This study provides theoretical support for practical repair planning and decision-making guidance during emergencies..