改良式自組織映射線性輸出模式於水庫入流量預報之研究

林國峰臺灣大學:土木工程學研究所張家銓Chang, Chia-ChuangChia-ChuangChang2010-06-302018-07-092010-06-302018-07-092009U0001-1007200914554300http://ntur.lib.ntu.edu.tw//handle/246246/187761本研究以自組織映射線性輸出模式(Self-organizing Linear Output Map, SOLO)架構一個有效的水庫時入流量預報模式。在類神經網路領域裡，倒傳遞類神經網路模式(Back-propagation Neural Network, BPNN) 被廣泛使用。相對於BPNN而言，SOLO模式的優勢在於：(1)準確度高、(2)架構簡單、(3)訓練所需時間少、(4)有助於分析。為了展示上述的四個優勢，本研究將SOLO應用於翡翠水庫入流量預報，研究結果顯示SOLO的效能及效率比BPNN來得佳。又為了改善SOLO尖峰入流量的預報，本研究進一步加入資料前處理的步驟以改良SOLO，並命名為改良式自組織映射線性輸出模式(Improved Self-organizing Linear Output Map, ISOLO)。研究成果證實ISOLO可以明顯改善SOLO尖峰入流量的預報。因此，建議可以ISOLO作為現有模式的替代方案，其優異的預報能力對水庫操作也相當有幫助。Based on self-organizing linear output map (SOLO), effective hourly reservoir inflow forecasting models are proposed. As compared with back-propagation neural network (BPNN) which is the most frequently used conventional neural network (NN), SOLO has four advantages: (1) SOLO has better generalization ability; (2) the architecture of the SOLO is simpler; (3) SOLO is trained much more rapidly, and (4) SOLO could provide features that facilitate insight into underlying processes. An application is conducted to clearly demonstrate these four advantages. The results indicate that the SOLO model is more well-performed and efficient than the existing BPN-based models. To further improve the peak inflow forecasting, SOLO with data preprocessing named ISOLO is also proposed. The comparison between SOLO and ISOLO confirms the significant improvement in peak inflow forecasting. The proposed model is recommended as an alternative to the existing models. The proposed modeling technique is also expected to be useful to support reservoir operation systems.誌謝 I要 IIBSTRACT III錄 IV目錄 VI 1 章緒論 1 - 1 前言與目的 1 - 2 文獻回顧 3 2 章模式方法 5 - 1 倒傳遞類神經網路模式(BPNN) 5 - 2 自組織映射線性輸出模式(SOLO) 8 - 3 改良式自組織映射線性輸出模式(ISOLO) 13 3 章模式建立與應用 20 - 1 研究區域與資料 20-1-1 翡翠水庫 20-1-2 水文資料 21 - 2 交替驗證與評鑑指標 23-2-1 交替驗證 23-2-2 評鑑指標 23 - 3 模式參數及輸入項設定 27-3-1 輸入項設定 27-3-2 BPNN參數設定 27-3-3 SOLO及ISOLO參數設定 28 4 章結果與討論 37 - 1 異分佈分析 37 - 2 自組織映射拓撲分析 39 - 3 模式效能比較 41-3-1 預報準確度 41-3-2 訓練所需時間 43 5 章結論與建議 75 - 1 結論 75 - 2 建議 78考文獻 793727417 bytesapplication/pdfen-US自組織映射線性輸出模式水庫入流量預報自組織映射圖尖峰入流量預報資料前處理類神經網路self-organizing linear output mapreservoir inflow forecastingself-organizing mappeak inflow forecastingdata preprocessingneural network改良式自組織映射線性輸出模式於水庫入流量預報之研究Improved Self-organizing Linear Output Map for Reservoir Inflow Forecastingthesishttp://ntur.lib.ntu.edu.tw/bitstream/246246/187761/1/ntu-98-R96521321-1.pdf