地铁网络服务韧性评估与最优恢复策略

针对已有基于拓扑效率的地铁网络韧性指标无法反映地铁运营实际的不足，构建考虑线 路流量影响的路网服务效率指标和基于服务效率的路网服务韧性指标，以及基于路网服务效率 的节点重要度指标；提出以路网服务韧性最大化为目标的优化模型，并基于遗传算法求解模型获 得最优恢复策略。算例结果表明：分别以服务效率和拓扑效率作为路网性能指标，获得的失效节 点恢复次序明显不同；蓄意攻击下，最优恢复策略获得的路网服务韧性分别比基于重要度的优先 恢复策略、基于节点度的优先恢复策略和随机恢复策略高16.76%、72.11%和86.21%。上述结果 表明，必须根据地铁运营实际合理选择路网性能指标和恢复策略，否则可能得到次优甚至明显偏 离实际的方案，无法实现预期目标。

Evaluation and Optimal Recovery Strategy of Metro Network Service Resilience

Since the existing network resilience indexes based on topological efficiency cannot reflect the actual metro operation situations, a novel index of metro network service efficiency considering the influence of line flows, an index of metro network service resilience based on service efficiency, and an index of node importance based on metro network service efficiency were constructed. An optimization model was proposed to maximize the metro network service resilience, and then an optimal recovery strategy was obtained by solving the model using a genetic algorithm. The results show that the recovery orders of failure nodes are different by taking service efficiency and topological efficiency as metro network performance index respectively. Under the deliberate attack, the network service resilience obtained by the optimal recovery strategy is 16.76%, 72.11%, and 86.21% higher than the priority recovery strategy based on node importance, the priority recovery strategy based on node degree, and the random recovery strategy, respectively. The above results indicate that the network performance indexes and the recovery strategies should be selected reasonably according to the actual metro operation situations. Otherwise, the suboptimal or even worse schemes which significantly deviate from the actual situations may be obtained, and thus the expected goals cannot be achieved.