Abstract: The reasonable post-disruption restoration plan of damaged components in interdependent critical infrastructure networks(ICINs)is the key issue of its safety management. Firstly, in this paper, the resilience metric of ICINs is defined and the post-disaster restoration strategy is analyzed. Then, aiming at maximizing the resilience of ICINs, a mixed integer programming model is formulated based on network flow theory for selection and sequencing of post-disaster restoration tasks of ICINs under the constraint of limited post-disaster restoration resources, and a genetic algorithm is developed to solve the problem. Finally, the proposed model and genetic algorithm are tested by applications with different sizes. The results show that: (1)the proposed model is feasible and effective; (2)the developed genetic algorithm can obtain high-quality solution, and the solution time and results are better than that of Cplex software for large size problems; (3)integrating the functional and spatial interdependencies between CINs into the proposed model can get higher resilience of ICINs. Our model can provide decision support for post-disaster restoration of ICINs.

Key words: critical infrastructure network, interdependences, post-disruption restoration, resilience, mixed integer programming

摘要： 科学合理制定相互依赖关键基础设施网络(Interdependent Critical Infrastructure Network, ICINs)遭灾后毁坏组件的修复计划是其安全管理的至关重要内容。本文首先明确了ICINs的韧性测度,分析了其灾后修复策略;然后基于网路流理论,以最大化ICINs的韧性为目标,构建了在有限灾后修复资源约束下,ICINs的灾后修复任务选择与调度的混合整数规划模型,并设计了遗传算法进行求解;最后通过不同规模的用例实验对模型和遗传算法进行了测试。研究表明:(1)该模型具有解决相关问题的可行性与有效性;(2)设计的遗传算法能获得质量较高的满意解,且对于大规模问题,遗传算法的求解时间与求解结果优于Cplex软件;(3)将网络之间的功能与空间相互依赖同时纳入模型中,能使ICINs的韧性达到更高。研究可为ICINs的灾后修复决策提供辅助。

关键词: 关键基础设施网络, 相互依赖, 灾后修复, 韧性, 混合整数规划