基于对冲理论的城市多水源供水系统调度研究

Based on the theory of hedging, a joint dispatch model of multi-source water supply is established to solve the joint scheduling problem of multiple water sources in urban water supply systems. Firstly, the project researches the dispatch rules of multi-source combined water supply based on the characteristics of incoming water of transferred water, characteristics of urban water supply systems, and water consumption rules of unit user. Second, the project builds the multi-level water supply dispatch program about‘water source- water plant - zoning – users’in water supply system of cities. Third，taking Tianjin as an example, the deep learning about big data is used to analyze the water using law of urban-subdivision and user-level. Finally, the project proposes the water supply dispatching plan of Tianjin, which considers the characteristics of multiple water sources, such as the South-to-North water source, water source of Luan River, and the local water source. The project provides the technical support for the dispatching decisions and security of water supply, which also provides the reference of multi-water source supply and dispatch in other cities.

针对城市供水系统多水源联合调度问题，基于对冲理论，建立多水源供水联合调度模型，研究基于外调水来水特性、城市供水系统特性和用户单元用水规律的多水源联合供水调度规则，构建城市供水系统水源-水厂-分区-用户的多层级供水调度方案。以天津市为例，采用大数据分析的深度学习方法，开展城市分区-用户级的用水规律分析，提出考虑南水北调水源、引滦水源以及本地水源等多水源特性的天津市供水调度方案，为天津市的供水调度决策和供水安全保障提供技术支撑，同时为其他城市多水源供水调度提供借鉴。

城市供水系统呈现多水源的供水格局，多工程的调控，多层级用水的监控特性。实现精细化精 准化调度，提高供水保证率，降低供水风险，确保城市供水安全成为现代智慧城市建设首当其冲的 问题。本项目首先构建了构建城市供水系统水源-水厂-分区-用户的多层级供水调度方案，提出了新 型城市建设条件下水源-水厂-用户的多重供水安全保障方案。利用网络拓扑结构理论分析“水源- 水厂-用户单元”的关联关系，基于城市供水系统特征划分用户单元，以用水监测信息为基础，采用 大数据分析方法挖掘用户单元用水规律，分析了不同用户的用水行为特征，构建了城市用水户的行 为特征集。基于城市供水系统的“水源-水厂-用户单元-用户”的联通关系，提出了构建城市 供水系统调度的概念模型，通过点线面的数字化表达，实现了通用化的城市供水系统组 件表达，进而采用EPANET模型构建了典型城市的供水模拟与调度模型。项目研究成果初 步应用于广州市城市供水系统的安全调度和用水行为画像。