應用人工智慧在汛期水庫防洪操作之研究

台灣四面環海、受季風氣候影響年降雨量雖高達2500mm，為世界各國平均之三倍，然由於地狹人稠、河川大多溪短流急、加上颱風降雨急促不易蓄存，這些都是在水資源管理上先天上的缺點。因此，台灣的水資源管理十分仰賴水庫的運作。然而現今水庫新壩址尋求不易且建新壩成本相當高的情況下，如何將現有的水庫在汛期高效能地操作，將會是影響水資源調配的重要關鍵。水庫在洪颱來臨前，若能掌握準確的降雨、入流資訊，如此才能有好的水資源調配決策。近年來，人工智慧技術快速發展，已成功地應用在各個領域，本文應用類神經網路(Artificial Neural Networks)、模糊理論(Fuzzy Theory)和適應性網路架構模糊推論系統(Adaptive Neuro-Fuzzy Inference Systems)，利用人工智慧技術的自我學習能力和邏輯推論等特性於水文資料的預測。本文共分為四個研究項目：颱風降雨量預測、水庫入流量、颱風洪水等級評判以及下游水位預測，各分別為第三章到第六章。石門水庫集水區和淡水河流域為本文之研究區域，下游水位預測的部分則以新海大橋水位為研究標的，期以準確的預測值提供水庫在洪汛時期，操作上的參考。

關鍵字

水庫操作；洪水預測；颱風降雨；類神經網路；模糊推論；水位預測

並列摘要

As a monsoon climate island, the annual average rainfall on Taiwan reaches 2500 mm, which is three times over world’s average. However, water resource in Taiwan count on typhoon rain due to its particular climate and geographic characteristics. It is hard for reservoir to consider both in reserve typhoon rain and flood control. Therefore, how to operate reservoir in a high-effect way is the key of water resource management in Taiwan. To make optimal operation decision, reservoir needs to forecast rainfall and inflow accurately before typhoon coming, and the purpose of this thesis is to build some models to meet reservoir’s demand. The models including four parts: typhoon rainfall forecasting, reservoir inflows forecasting, typhoon flood assessment, and downstream water level forecasting (in Chapter 3~6, respectively). All the proposed models are based on artificial intelligence (AI) technique. AI has been developed rapidly in recent years and applied extensively in many fields since it possesses advantages of self-learning and logical inference. Artificial neuron networks, fuzzy theory, and Adaptive Neuro-Fuzzy Inference Systems (ANFIS) are three AI methods applied here. Shihmen reservoir and Danshuei River basin are taken as study area; water level in Sin-Hai Bridge is prediction downstream water level. Some satisfactory results are showed in this thesis, and this thesis provides a well-forecast method for reservoir to refer to while facing operation problems.

並列關鍵字

Typhoon rainfall ； Flood assessment ； Reservoir operation ； Fuzzy inference ； water level prediction ； Artificial neural networks

參考文獻

[2] Holland, J. H., Adaptation in Natural and Artificial Systems, The University of Michigan Press, Ann Arbor, 1975.

[3] Krishnamoorthy, C. S., Artificial intelligence and expert systems for engineers. Boca Raton : CRC Press, 1996.

[4] Li, H. X., Fuzzy sets and fuzzy decision-making. Boca Raton: CRC Press, 1995.

[6] McCarthy, J., Recursive functions of symbolic expressions and their computation by machine, Communications of the ACM, 7, 184-195, 1960.

[9] Newell, A., Shaw, J. C., and Simon, H. A., Empirical explorations with the logic theory machine: a case study in heuristics. Computers and Thought, Feigenbaum, E. A. and Feldman, J. (Eds.), McGraw Hill, New York, 1963.

國際替代計量

應用人工智慧在汛期水庫防洪操作之研究

未授權

主題瀏覽