关联基础设施系统的弹性优化

In recent years, infrastructure systems, such as electric power, natural gas and transportation systems, are subjected to natural hazards, sharp increase of demand, climatic change and some other types of hazards, which make their failure probabilities and risks increase yearly. Also, these systems are interdependent (for example, gas systems need electricty to control gas pressure while power systems need gas to generate electricity), which makes failures in single systems easily spread to other systems, largely affecting regional economy and citizen life. Hence, many countries have taken resilient infrastructure systems as their strategic goals, which require these systems have strong resistant, absorptive and restorative capacities to different types of hazards. To realize these goals, how to optimize infrastructure resilience is the key. In the literature, the resilience optimization studies on intentional attacks did not model disaster restoration processes, and those studies on random events only optimized one or two strategies for resilience enhancement, while almost all these work only studied single infrastructure systems and did not consider the time-dependent evolution feature of system demand. This project intends to study resilience optimization of multiple infrastructure systems with interdependencies. By mainly taking interdependent electric power and gas systems as examples and considering many types of candidate resilience enhancment strategies, this project will first study the resilience optimization approach under constant system demand for intentional attacks and random events, respectively, then study the time-dependent resilience optimization approach under demand evolution and multiple hazards, and finally use several real systems to validate and update the proposed resilience optimization approaches, and develop a prototype decision support system for resilience optimization of interdependent electric power and gas systems.

近年来，自然灾害、需求激增、气候变化等因素使得电力、天然气与交通等基础设施系统的故障概率与风险逐年增加；同时，这些系统相互关联（比如天然气系统需要电能控制气压，电力系统需要天然气发电），使得单个系统故障极易在系统间扩散，严重影响区域经济和人民生活。因此，弹性的基础设施成为许多国家追求的目标，即要求系统对各类灾害有较强的抵抗性、适应性与恢复能力，其中如何优化基础设施弹性是关键。目前，文献中针对蓄意破坏的弹性优化研究未考虑灾后恢复，针对随机事件的研究仅考虑个别弹性优化策略，同时多数研究仅针对单个系统且未考虑系统需求演化。本项目拟针对关联的多个基础设施系统，重点以电力与天然气系统为例，考虑多类候选策略，首先研究需求不变时分别针对蓄意破坏与随机事件的弹性优化方法；然后研究需求演化时针对多重危害源的时变弹性优化方法；最后用实际系统验证与完善上述方法，并开发关联电力与天然气系统的弹性决策支持原型系统。

2020年我国城镇化率已超过60%，与之伴生的是城市安全危机的增加，韧性城市成为危机应对的新视角与新趋势（原申请书将resilience译为弹性，目前国内已统一将其译为韧性）。由于电力、天然气、交通等基础设施系统是城市的生命线，它们的韧性极大程度上决定城市应对各种灾害的韧性，因此提升基础设施韧性是实现韧性城市的关键。基于此，项目组在过去四年里开展了基础设施系统的韧性提升优化研究，主要研究内容包括：(1) 研究基础设施系统针对蓄意破坏的韧性提升方法；（2）研究基础设施系统针对自然灾害的韧性提升方法；（3）研究基础设施系统针对多重危害的韧性提升方法。根据上述研究内容，截止2021年1月13号，共发表SCI收录论文15篇(JCR一区14篇)，均标注基金编号，在学术会议上做报告20余次，主要成果包括：（1）针对蓄意破坏，建立了灾前加固策略的三阶防御者-攻击者-决策者韧性提升优化模型，提出了基于线性化处理技术与Benders分解思想的精确求解算法；（2）针对自然灾害，建立了灾前加固策略的三阶段随机优化模型，提出了综合灾害情形缩减技术、解空间缩减技术、L型算法以及采样平均近似方法的高效求解算法；（3）针对基础设施系统的灾后恢复过程，建立了韧性驱动下受损组件修复次序优化的决策模型，提出了在精度和计算效率上最优的启发式求解算法。（4）以日常运维风险与地震灾害为例，生成基础设施系统在生命周期内（50年）的灾害情形，结合不同灾害下的组件易损性模型、系统功能计算模型与灾后恢复过程模型，提出了全生命周期韧性评估方法，分析了不同韧性提升策略的效果。上述研究成果得到了较好的引用，Google Scholar引用273次，项目负责人在项目执行期间每年均入选Elsevier “中国高被引学者”。本项目的研究成果不仅能够在灾前针对各类危害提前做好预防与预警，还能够帮助决策者在灾后更好地调配各种恢复资源，缩短恢复时间、降低灾后损失，为韧性城市规划与管理提供理论指导。