水資源頻譜網路最佳化模型分析

Abstract

在不同水庫之間分配水資源對政策制定者來說是一項困難的任務，因為他們必須確保水庫出水量能可以滿足各項用水。本文主要的目標是獲得最準確的多水庫水資源分配，並且探討曾文南化水庫聯通後的水資源分配狀況。因此需要一個水資源分配最佳化模型以提供參考。本文提出了一個考慮時間因素的網路流量最佳化模型，該演算法結合傅立葉頻譜分析，以解決水資源系統中的水資源分配問題。該方法應用於供應高雄與台南的四個水庫，曾文水庫、烏山頭水庫、南化水庫、高屏溪攔河堰。並且透過該模型分析未來曾文水庫與南化水庫聯通後的水資源分配狀況，並且與現行尚未聯通的水資源分配狀況進行比較。結果顯示，該演算法成功地有效最佳化水庫的出水量以滿足高雄台南地區對於這四個水庫的用水需求。並且利用缺水指標進行模型的評估，可以發現在聯通後各水庫各旬的缺水狀況都有所改善。

The allocation of water resources among different reservoirs poses a challenging task for policy makers, as they must ensure that the discharge from reservoirs can satisfy all water usage needs. The primary objective of this study is to obtain the most accurate multi-reservoir water allocation, and to explore the water resource allocation condition after the connection between the Zengwen and Nanhua reservoirs. Therefore, an optimization model for water resource allocation is required for reference. This paper proposes a network flow optimization model considering the temporal factor. This algorithm, combined with Fourier spectrum analysis, aims to solve the water resource allocation problem within the water resource system. This method is applied to four reservoirs that supply water to Kaohsiung and Tainan, namely the Zengwen Reservoir, Wushantou Reservoir, Nanhua Reservoir, and Gaoping River Weir. Moreover, through this model, we analyze the future water resource allocation situation after the connection between the Zengwen Reservoir and the Nanhua Reservoir, and compare it with the current situation where they are not yet connected. The results demonstrate that the algorithm effectively optimizes the outflow from the reservoirs to meet the water needs of the Kaohsiung and Tainan areas from these four reservoirs. Additionally, using a water shortage index to evaluate the model, it was found that the water shortage situation of each reservoir improved in every ten-day period after connection.