Robust Risk-based Decision Analytics for Enhancing Seismic Resilience of Water Pipe Networks

美国的水务公司需要对供水基础设施系统做出艰难的修复决定。这些决定存在很大的不确定性，因为老化和通常埋在地下的水管网络，以及它们承受地震等灾害的能力。由于许多地区的基础设施修复资金严重缺乏，这些决定变得更加复杂。本研究将创建并实证检验一个强有力的决策框架，用于管理受这些不确定性影响的供水管道的维护和补救。预计这一框架将帮助水务公司从他们在水基础设施系统上花费的每一美元中获得最大的价值。该项目将吸引水务公司等行业利益相关者参与，并通过制作易于理解和实用的教育材料，广泛传播研究成果。德克萨斯大学阿灵顿分校西班牙裔学生的本科生和研究生，尤其是少数族裔学生，将参与该项目的每一步。这项研究还将通过实践活动吸引K-12学生参与研究，使他们熟悉在不确定性下做出工程决策的基本原理。本研究将通过(1)描述管网不确定性并将其与地震易损性评估相结合，以及(2)设计一种风险规避型随机组合优化算法并将其与地震易损性评估相结合，以识别管道中的易损性位置，从而解决现有模型在基于风险的管网鲁棒性决策方面的根本局限性。该项目涉及两个关键学科：决策科学和基础设施脆弱性评估。在高度不确定的环境下，迫切需要将管网不确定性整合到供水管网地震易损性评估中。该项目将通过整合借鉴定量金融的机制，改变现有的选择修复和维修政策的模式，以确保选择风险不会太高。将风险规避型随机组合优化模型引入土木工程界，为基础设施决策者提供了合适的可扩展模型，以便在高度不确定的环境中以可控的风险规避水平优化修复决策。该项目将创建计算上具有成本效益的算法，以解决大型城市水网制定的优化问题。该奖项反映了美国国家科学基金会的法定使命，并通过使用基金会的知识价值和更广泛的影响审查标准进行评估，被认为值得支持。

Water utilities in the US need to make difficult rehabilitation decisions about the water infrastructure system. There are deep uncertainties in these decisions because of aging and typically buried water pipe networks and their ability to sustain disasters, such as earthquakes. These decisions are further complicated with the enormous lack of funding in many areas for infrastructure rehabilitation. This research will create and empirically examine a robust decision-making framework for managing the maintenance and remediation of water pipelines subject to these uncertainties. This framework is expected to assist water utilities to achieve the most value out of every dollar that they spend on their water infrastructure system. This project will engage industry stakeholders, such as water utilities and disseminate research results widely through the creation of easy-to-understand and practical educational materials. Undergraduate and graduate students-especially minority students-at the Hispanic Serving University of Texas at Arlington (UTA) will participate in every step of this project. This research will also engage K-12 students in research through hands-on activities and make them familiar with the fundamentals of making engineering decisions under uncertainty. This research will address fundamental limitations of existing models for making robust risk-based decisions about water pipe networks by (1) characterizing the network uncertainties and integrating them with seismic vulnerability assessment, and (2) designing and integrating a risk-averse stochastic combinatorial optimization algorithm with the seismic vulnerability assessment to identify vulnerable locations in a pipeline. This project is at the nexus of two critical disciplines: decision science and infrastructure vulnerability assessment. The integration of network uncertainties in seismic vulnerability assessment of water pipe networks is critically needed in a highly uncertain environment. The project will transform existing models for selecting rehabilitation and maintenance policies by integrating mechanisms borrowed from quantitative finance to assure that selection risks are not too high. The novel introduction of risk-averse stochastic combinatorial optimization models to the civil engineering community equips infrastructure decision makers with proper scalable models to optimize rehabilitation decisions with a controllable risk aversion level in a highly uncertain environment. This project will create computationally cost-effective algorithms to solve the formulated optimization problems for large-scale city water networks.This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.