CAREER: Predictive Estimation for Managing Water Distribution Systems and Applications of Distributed Sensing Technologies

这项教师早期职业发展（Career）资助将支持水系统管理者的决策过程，以推进和促进智能和弹性水系统。供水系统是我们城市环境的重要组成部分，为我们的家庭和企业提供安全的饮用水。这项研究将开发计算方法和工具，用于使用安装在整个供水系统分配网络中的新型传感设备收集的数据。这些工具将更好地估计当前系统在当前需求下的运行情况，以及在需求变化和不确定危险的情况下进行必要的修改。这项研究的主要影响是通过传播新的科学工具、培训未来的工程师和将研究与教育相结合，弥合智能水系统的知识差距。这项科学研究有助于美国国家科学基金会促进科学进步、促进国家福利和确保公共供水的使命。本项目有一个研究和教育计划，重点是通过参数识别、不确定性量化和模型简化，为大规模智能水系统在有限可观测性下的预测状态估计创建新的模型。研究活动包括：(1)创建新的状态估计模型，无缝集成来自分布式传感设备的异构测量；(2)通过结构化的节点消除、需求聚合和识别影响参数，对水力模型进行简化，为降低状态估计和不确定性传播问题的维数生成灵活的替代方案；(3)创建新的传感器放置策略，以最佳选择和定位异构计量仪器，提供最大的信息增益；(4)估计预测边界并量化预测对测量、需求和拓扑不确定性的敏感性。外联活动包括利益攸关方研讨会，由水务公司和地方机构、技术和咨询公司参与，确定并提出技术采用的实际解决方案；用于再现和传播研究成果的开源工具包和存储库，以及培训未来工程师使用先进建模工具和增加少数群体参与的技术讲习班；用于状态估计的传感器网络算法之战，作为挑战和培训研究人员，创造和分享新的科学工具的机制；水上比赛，透过有趣的实践活动，激发年轻人对工程的兴趣；改革工程课程，以适应传感和计算技术的进步，并将研究融入教育。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。

This Faculty Early Career Development (CAREER) grant will support decision-making processes by water system managers to advance and promote intelligent and resilient water systems. Water distribution systems are critical components of our urban environments providing safe drinking water to our homes and businesses. This research will develop computational methods and tools for using data collected by new sensing devices placed throughout a water system’s distribution network. These tools will better estimate current system’s operation given current demands and necessary modifications given changing demands and uncertain hazards. The primary impact of this research is to bridge knowledge gaps toward intelligent water systems through dissemination of new scientific tools, training future engineers, and integrating research into education. This scientific research contributes to NSF’s mission to promote the progress of science, to advance national welfare, and to secure public water provisions.This project has a research and education plan that focuses on creating new models for predictive state estimation under limited observability through parameter identification, uncertainty quantification, and model reduction for large-scale intelligent water systems. Research activities include: (1) creating new state estimation models that seamlessly integrate heterogeneous measurements from distributed sensing devices; (2) generating flexible alternatives for reducing the dimensionality of the state estimation and uncertainty propagation problems by reducing the hydraulic models through structured elimination of nodes, demand aggregation, and identification of influential parameters; (3) creating new sensor placement strategies for best selection and location of heterogeneous metering instruments providing maximal information gain; and (4) estimating prediction bounds and quantifying prediction sensitivity to measurement, demand, and topology uncertainties. Outreach activities include stakeholders’ workshop that will engage water utilities and local agencies, tech and consulting companies to identify and present practical solutions for technology adoption; open-source toolkit and repository for reproducibility and dissemination of research outcomes and technical workshop for training future engineers in advanced modeling tools and increasing minority participation; battle of algorithms for sensor networks for state estimation as a mechanism to challenge and train researchers, create and share new scientific tools; water competition to get young people excited about engineering through fun, hands-on activity; and revamping engineering curriculum to adapt to the advances in sensing and computational technologies and integrate research into education.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.