基于DBN的地铁隧道系统运行韧性评估研究

尹紫微

摘 要

近年来，随着我国地铁投资建设的迅速发展，地铁建设规模得到进一步扩大。地铁隧道系统属于城市重大地下系统，受到内外多种耦合因素作用，一旦在运营期出现受损情况，其维修难度较大，严重时会影响到地铁的正常运营以及隧道的安全性。传统的地铁安全性评估研究方法主要侧重于被动的研究，如可靠性、风险管理等，强调事故发生前的风险控制，较少考虑灾后恢复问题。在综合考虑经济性和安全性时，韧性已经成为描述突发事件下复杂系统的应对性能的重要概念，强调对突发事件的吸收、适应和恢复能力，因此，本文开展了对地铁隧道系统运行韧性的研究。

本文以韧性理论为基础，以地铁隧道系统为研究对象，分析了目前地铁运营安全评估面临的挑战，结合各界学者在韧性安全评估的理论研究，总结出地铁隧道系统运行韧性的概念，将地铁隧道系统运行韧性量化为对地铁隧道系统吸收能力、适应能力和恢复能力的研究，系统性分析了地铁隧道系统运营期间性能变化特点。

其次，本文基于地铁现场监测点的实测数据，选取管片收敛变形值作为性能评价指标，通过建立ARIMA时序模型，得到地铁隧道结构性能变化曲线，对地铁隧道结构性能进行预测分析，从而对地铁隧道系统运行韧性有初步的了解。

在性能曲线的基础上，采用概率分析的方法对韧性进行评估，讨论地铁隧道系统运行韧性的影响因素，确定节点变量，研究节点间的相互关系，构建动态贝叶斯网络，通过计算条件概率，完成对运营地铁隧道系统运行韧性的估计。

最后，本文对某段发生病害的地铁隧道进行实例分析，并对提升地铁隧道系统运行韧性提出建议。

关键词：运营地铁；韧性评估；隧道系统；ARIMA；DBN(动态贝叶斯网络)

Abstract

In recent years, with the rapid development of metro investment in China, the scale of metro construction has been further expanded.Subway tunnel system is a major urban underground system, which subject to internal and external coupling factors. Once the damage occurs in the operation period, it will be difficult to maintain and affect the normal operation of the subway and the safety of the tunnel seriously. The traditional research methods of subway safety assessment mainly focus on passive research, such as reliability and risk management, which emphasizes the risk control before the accident, and pay little attention to the recovery after the accident. When considering economy and safety comprehensively, resilience has become an important concept to describe the response performance of complex systems under emergencies, which emphasizes the ability to absorb, adapt and recover from emergencies. Therefore, this paper develops the research on operation resilience assessment of subway tunnel system.

Based on the theory of resilience, this paper chooses subway tunnel system as the research object and analyse the challenge of the current safety assessment of operational subway. Then combining with the resilience theory of other scholars, this paper concludes the concept of the operation resilience of subway tunnel system,and transforms it into the research of the ability of absorbing, adaptability and recovery, then analyzes the change characteristics of subway tunnel system during operation period.

Secondly, based on the data of subway on-site monitoring points, this paper selects the converged-deformation value of the tube pieces as the evaluation index. Then this paper obtains the structural performance change curve of subway tunnel through the establishment of ARIMA time series model. Therefore the structural performance of subway tunnel can be predicted and analyzed, thus to have a preliminary understanding of the operation toughness of subway tunnel system.

Based on the performance curve, this paper adopts the method of probability analysis to evaluate resilience, and discuss the influence factors of operation resilience of subway tunnel system. After determining the node variables and the relationship between nodes, this paper develops the Dynamic Bayesian Network. Then this paper completes the estimate of operation resilience assessment of subway tunnel system, through the computation of conditional probability.

Finally, this analyses a section of subway tunnel with structural disease, and gives some suggestions of improving the operation resilience of subway tunnel system.

Keywords: Operating subway; Resilience assessment; Tunnel system; ARIMA; DBN(Dynamic Bayesian Networks)