随着城市的快速发展，城市面临的各类自然和人为灾害的风险也在快速聚集，给城市居民的生命财产安全带来严峻威胁。基础设施系统作为城市功能的重要支撑，其灾害应对能力是城市灾害管理水平的重要体现。韧性为理解基础设施系统在灾害全过程中的反应，评估并提升基础设施系统灾害应对能力提供了重要的新的视角。基础设施系统的韧性建模方法研究对于基础设施系统韧性的评估、诊断及提升具有重要意义。然而，现有研究在韧性建模时多采用高度抽象整体的模型，在考虑系统自身异质性与跨系统关联性方面存在不足，无法精细化模拟系统真实的运行状态与跨系统的复杂交互关系，导致关联基础设施系统韧性建模与分析方法存在不足。为了解决上述问题，本研究在评估并阐明异质性对级联失效影响的基础上，引入各系统领域知识以考虑系统自身异质性，利用联合仿真技术集成并贯通各领域知识或模型以考虑跨系统关联性，进而提出关联基础设施系统韧性建模与分析方法。首先，本研究在识别关联基础设施间系统层面异质性的基础上，基于现有研究中模拟级联失效的典型方法，建立考虑异质性因素的级联失效改进模型，定量评估异质性对级联失效过程的影响，并展开案例研究验证所提方法的有效性及在级联失效过程中考虑异质性的必要性。其次，基于联合仿真技术，考虑系统异质性、跨系统关联性及灾害动态影响，提出关联基础设施系统级联失效过程联合仿真方法；在此基础上，本研究进一步提出不同破坏类型下关联性对级联失效过程影响效应的评估方法，并基于案例研究验证级联失效过程仿真方法及关联效应评估方法的有效性。最后，在级联失效分析结果的基础上，本研究从精细化模拟修复序列对基础设施系统状态影响的角度出发，提出关联基础设施系统恢复过程联合仿真方法；基于恢复过程的精细化模拟数据，本研究提出修复序列决策模型求解方法的改进准则，建立基于遗传算法的修复序列决策模型的改进求解方法，并基于案例研究验证恢复过程模拟和改进求解方法的有效性。本研究提出的韧性建模与分析方法可服务于灾害影响评估、恢复决策、关联系统界面优化设计及恢复方案比选等，为关联基础设施系统韧性评估、诊断及提升提供了方法基础。

With the rapid growth in urbanization around the globe, modern cities increasingly face various risks of natural and man-made disasters which pose serious threats to the safety of people and infrastructures. Infrastructure systems play a major role in sustaining urban functions, and thus, their disaster response capacity reflects the level of disaster preparedness of urban cities. Resilience provides an important new perspective for understanding the response of infrastructure systems throughout the lifecycle of disasters, and for assessing and enhancing the disaster response capacity of infrastructure systems. Studies on the resilience modeling of infrastructure systems are important for the assessment, diagnosis and enhancement of the resilience of infrastructure systems. However, most prior studies adopt highly abstracted monolithic models, which are inadequate for analyzing systemic heterogeneities and cross-system interdependencies, and are unable to simulate with sufficient granularity the intrinsic complexities of system operational mechanisms and interactions, demonstrating the inadequacy of existing resilience modeling methods.To address the above issues, this study assesses and clarifies the impact of systemic heterogeneities on the cascading failure of interdependent infrastructure systems, and proposes a co-simulation approach for modeling and assessing the resilience of interdependent infrastructure systems. The said approach captures various systemic heterogeneities by leveraging and integrating the domain-specific knowledge or models of different infrastructure systems. Specifically, this study begins by identifying various dimensions of systemic heterogeneity among infrastructure systems, and proceeds to develop an improved cascading failure model by incorporating the identified heterogeneities in a typical cascading failure simulation method. A case study is then conducted to evaluate the impact of heterogeneities on cascading failure, verify the need to consider heterogeneities in the modeling of cascading failure processes, and validate the proposed method. Next, to account for, and assess the impacts of, systemic heterogeneities, interdependencies and the dynamic impact of disasters on cascading failures, a modeling approach based on the co-simulation of domain-specific infrastructure system models is proposed. On this basis, this study further proposes a method to assess the interdependency effect on the cascading failure of infrastructure systems under different types of disruptions, and the proposed method is validated by means of a case study. Finally, based on the results of the cascading failure analysis, this study proposes a co-simulation method for the recovery process of interdependent infrastructure systems, that allows for detailed simulation of the impact of repair sequences on the state of the infrastructure systems. The simulation data is used to suggest an improvement criterion for the repair sequence decision model, and develop a genetic algorithm -based solution for optimizing the recovery of interdependent infrastructure systems. A case study is then conducted to verify the effectiveness of the recovery simulation and validate the proposed method.The resilience modeling and analysis methods proposed in this study can be used for disaster impact assessment, recovery decision making, optimization of interdependent system design, and comparison of recovery strategies. Moreover, the findings and deliverables of this study provide a strong foundation for the assessment, diagnosis and enhancement of the resilience of infrastructure systems.