In this paper, we present a methodological framework for resilience analysis of interdependent critical infrastructure systems and use artificial interdependent power and gas network as an example. We use deep learning to identify network topology attributes and analyze the vulnerability process of interdependent infrastructure systems to different failure scenarios and coupling modes under structural perspective. Then, functional model of the interdependent network is constructed, and the vulnerability process based on functional characteristics is analyzed. At last, we propose different recovery strategies and use a resilience triangle to study the restoration process, and the optimal resilience improvement strategy is acquired from both structural and functional perspectives. The method proposed in this paper can help decision makers develop mitigation techniques and optimal protection strategies.

在本文中，我们提出了一个对相互依赖的关键基础设施系统进行弹性分析的方法论框架，并以人工相互依赖的电力和天然气网络为例。我们使用深度学习来识别网络拓扑属性，并在结构的角度分析相互依赖的基础设施系统对不同故障场景和耦合模式的脆弱性过程。然后构建了相互依赖网络的功能模型，分析了基于功能特性的脆弱性过程。最后提出不同的恢复策略，并使用弹性三角形研究恢复过程，从结构和功能两个角度得到最优的弹性提升策略。