因應都市化影響之智慧型水資源管理系統

Intelligent Water Management System Under the Influence of Urbanization

鄭仲廉(Chung-Lien Chen)

指導教授 ： 張斐章

國立臺灣大學/生物資源暨農學院/生物環境系統工程學研究所/碩士(2016年)

https://doi.org/10.6342/NTU201601834

摘要

面對台灣水資源時空間分佈不均及人口膨脹、工商業發展蓬勃所導致的缺水問題，北部地區石門水庫用水調度將是ㄧ個重大課題。本研究透過灌溉用水的調度彈性支援非灌溉用水標的，建置因應未來都市化用水衝擊的整合型水資源智慧管理系統，期能有效率地使用水資源並降低乾旱缺水所帶來的供水壓力及經濟損失，此研究並考慮桃園地區埤塘系統備援以增加本研究水資源系統之供應承載力。
本研究蒐集2005-2014年桃園地區農業、工業以及人口發展等長期歷史資料，應用系統動力模式，投射未來2015-2030年桃園地區需水情勢，並參考1977、1984、2002年等乾旱年之石門水庫旬入流量及水庫初始有效庫容50%、40%、30%等情境，共設定九種未來供水可能狀況，運用M-5操作規則模擬未來缺水嚴重程度;並透過非支配排序遺傳演算法-II(NSGA-II)搜尋最大平均有效蓄水率(RRS)及最小修正缺水指標(MSI)等兩個目標；透過NSGA-II搜尋於各水文情境的表現，結果顯示修正缺水指標(MSI)有著6.9%~24%的改善率，各旬平均有效蓄水率(RRS)最高可達9.6%的改善率;若在考量埤塘備援系統的情況下，修正缺水指標(MSI)改善率最高更可達35.5%，各旬平均有效蓄水率(RRS)最高達1.9%。
根據研究結果顯示，透過NSGA-II搜尋的最佳供水操作策略以及埤塘備援系統的支援，可有效因應未來都市化需水情勢，作為決策者於水資源管理上的參考依據。

關鍵字

水資源管理 ； 都市化 ； 非支配排序遺傳演算法-II(NSGA-II) ； 系統動力模式 ； 埤塘備援系統 ； 多目標水庫操作

並列摘要

Facing the uneven spatio-temporal distribution of water resources and the increasing water demands caused by population growth and industrial development, the water regulation of the Shimen Reservoir for Taoyuan has become a critical issue. This study aims to build an intelligent water allocation system in order to suitably make water regulation with flexible water transfer from irrigation sectors to industrial and municipal sectors for reducing water pressure in public sectors during drought periods. In addition, the farm ponds in Taoyuan are also considered as a back-up water resource in this study for enhancing the resilience of the intelligent water allocation system.
This study first simulates the future water demands of Taoyuan for the period of 2015 and 2030 by using the system dynamics theory based on agricultural and industrial data as well as population statistics collected during 2005 and 2014. We next design nine water supply scenarios in response to the possible drought conditions in the future based on the ten-day inflow data collected from the Shimen Reservoir in three drought years (1977, 1984, 2002) and three initial reservoir storage capacities (50%, 40%, 30%) of these drought years. According to the simulation results of future water demand and supply, the M-5 rule curves are used to simulate the water shortage conditions during 2015 and 2030 while the non-dominated sorting genetic algorithm-II (NSGA-II) is used to search the minimal modified shortage index (MSI) and the maximal ratio of effective reservoir storage capacity. The results of the NSGA-II for these nine designed scenarios indicate that the improvement rates of the MSI (as compared to the MSI obtained from M5 rule curves) ranges between 6.9% and 24% while the averaged effective reservoir storage capacity ratio for ten-day periods reaches as high as 9.6%. When the back-up water resource of farm ponds in the study area is incorporated into the proposed intelligent water allocation system, the new results of the NSGA-II indicate that the improvement rates of the MSI and the averaged effective reservoir storage capacity ratio for ten-day periods will further reach as high as 35.5% and 1.9%, respectively.
The results of this study demonstrate that the multi-objective reservoir operation strategy obtained from the NSGA-II with the back-up water resource of farm ponds can make effective water allocation in response to urban water demands and thus provide decision makers with reference guidelines in sustainable water resources management. We hope that the proposed intelligent water allocation system will pave the way to future research for integrated water resources management.

並列關鍵字

Water resources management ； Urbanization ； Non-dominated sorting genetic algorithm-II (NSGA-II) ； System dynamics ； farm ponds ； Multi-objective reservoir operation

參考文獻

23. 戴巧雯, 2014. 多目標遺傳演算法應用於滯洪池最佳化優選. 成功大學水利及海洋工程學系學位論文, 1-75.

17. 吳俊宏, 2013. 支持向量機與遺傳演算法應用於水庫即時操作. 中原大學土木工程研究所學位論文, 1-112.

14. 王國威, 2011. 智慧型區域水資源調配管理與休耕決策系統. 臺灣大學生物環境系統工程學研究所學位論文, 1-97.

22. 楊舜年, 2015. 建立颱洪時期抽水站智慧型最佳化操作規則. 臺灣大學生物環境系統工程學研究所學位論文, 1-126.

15. 王昱中, 2014. 智慧型水資源調配策略以因應用水需求成長. 臺灣大學生物環境系統工程學研究所學位論文, 1-92.