水資源系統串接之頻譜網路最佳化分析

Abstract

由於氣候變遷的影響，極端氣候出現頻率大幅增加，大型旱災水災頻傳。近年來水利署加強各方面的水利設施與政策，除了開源、節流外，更是積極規劃興建了許多輸水管道，這些輸水管道連接各個供水區，希望能藉由這些管道的串連讓水資源的調度更為方便迅速，在水資源受到天氣影響時，第一時間做到跨區支援，讓水資源的供應更加穩定。在這些管線興建的同時，該如何更有效率的分配這些水資源，便成了現階段需要重視的問題。 本研究針對主要幹管串連問題，試圖結合傅立葉級數的特性，將水文的時間序列資料使用傅立葉級數轉換成正交頻率域上的係數來進行最佳化模型分析，藉此探討串接對於供水能力的影響，並依據正常年、乾旱年的雨量數據，來模擬過去發生的限水事件時，若存在聯通管有沒有辦法改善當時的問題。本研究選擇2020年間發生的百年大旱作為主要研究對象，並以2019年的數據作為正常年對照組。經由模擬結果，發現在聯通管運作的情況下，搭配政策調整用水端需求，將能應對過去發生過的乾旱事件，解決大多數的缺水情況。

As the impact of climate change, the intensity and frequency of extreme drought events have increased substantially. In 2021, Taiwan faces the worst drought in a century. To overcome the drought, Water Resources Agency has planned to build water pipelines, the purpose of which is to connect the water resources systems between different region. In the future, it will be an important way to deal with droughts to achieve the stability of water supply scheduling through the series connection of pipelines in the water resources system. The research will provide methods and tools for systems analysis to assess regional water resource dispatch capabilities. The purpose is to analyze the influence of water resources connection on water supply and drought resilience. The case will analyze the supply and demand of water resources and scheduling water resource by building a network flow model. Hydrological data and agricultural water are all periodic time series data or functions, so the time series superposition or function superposition can be converted to optimal analysis on the frequency domain. The study will use Fourier spectrum analysis to deal with water resources data, and then combine with network flow analysis and hydrological time series data. The goal is to analyze optimization of water resources system piping series and uncertainty of water supply capacity.