運用賽局理論結合遺傳演算法於最佳化水庫操作之研究

A Study of Optimization of Reservoir Operation Using Game Theory and Genetic Algorithms

徐宗孚(Tsung-Fu Hsu)

指導教授 ： 張麗秋

淡江大學/工學院/水資源及環境工程學系碩士班/碩士(2017年)

https://doi.org/10.6846/TKU.2017.00348

摘要

近年來，因社會環境之進步與提升，對水資源需求標準提高，但在先天地形限制，符合新建水庫的位址已相當缺乏，加上全球暖化極端暴雨事件頻繁，造成水庫上游集水區土石崩落、水流常挾帶大量泥砂入庫、水庫泥砂淤積日益嚴重，水庫庫容驟減，面臨新的水源開發受限與現有水資源利用效益降低，因此如何操作水庫系統，使得有限的水資源能夠有效的調配與運用是台灣目前最需要重視的課題之一。
本研究以賽局理論結合遺傳演算法，探討石門水庫水資源如何有效地分配農業用水與公共用水，並維持水庫蓄水等三方面之最佳效益，期以建立一個水庫最佳化供水操作模式，使有限的水資源能獲得最適當的運用。從52年的模擬結果比較，賽局理論配水模式有別於傳統以加權各目標之權重法易發生極端配水情況，賽局理論配水不會發生完全不供水之情況，可有效且公平地解決水資源分配等衝突問題。顯示賽局理論結合遺傳演算法於水庫操作上確實可行，且在模擬結果中良好的呈現出優異的缺水指標與庫容狀態，並滿足水庫在實務操作上之特性與限制，藉以提供未來水資源規劃決策之參考。

關鍵字

賽局理論 ； Nash協議解 ； 遺傳演算法 ； 水庫操作

並列摘要

In recent years, the rapid change of global environment and socio-economic development has resulted in a tremendous and high standard demand for water resources. Besides, the high intensive and long-duration rainfalls have resulted in a number of debris and reduced the reservoir capacities. With the gradually decrement of the reservoir capacities, it is very important and urgent to improve the operational effectiveness and efficiency of existing reservoirs for maximizing the beneficial use of water storage.
Besides agricultural and domestic & industrial water supply, we should maintain the reservoir water storages for future use. In this study, game theory is integrated with genetic algorithms to investigate the optimal Shimen Reservoir operating policies and water allocation with the conflict problems among different water usages (objectives) due to limited water resources. For the purpose of comparisons, the weighted method is applied in Shihmen Reservoir water allocation policies. In 52 year’s simulation results, the proposed model can reasonably allocate water among the different water usages and objectives, and avoid severe water shortage for any water user or reservoir storage empty.

並列關鍵字

Game theory ； Nash bargaining solution ； Genetic Algorithms(GA) ； Reservoir operation

參考文獻

20. 王國威，2006，運用懲罰機制遺傳演算法於水庫颱洪操作之規劃，碩士論文，淡江大學水資源及環境工程學研究所。

1. Chang, L. C., Chang, F. J., Wang, K. W., & Dai, S. Y. (2010). Constrained genetic algorithms for optimizing multi-use reservoir operation. Journal of Hydrology, 390(1), 66-74.

2. Chen, L., McPhee, J., & Yeh, W. W. G. (2007). A diversified multiobjective GA for optimizing reservoir rule curves. Advances in Water Resources, 30(5), 1082-1093.

3. Holland, J. H. (1975). Adaptation in natural and artificial systems. An introductory analysis with application to biology, control, and artificial intelligence. Ann Arbor, MI: University of Michigan Press.

4. Hsu, S. K. (1995). Shortage indices for water-resources planning in Taiwan. Journal of Water Resources Planning and Management, 121(2), 119-131.