查詢結果分析
相關文獻
- 高解析降雨資料應用於河川洪水位預報
- 結合群聚分析與人工智慧於颱風時雨量預測
- 結合類神經網路之動力波模式--蘭陽溪洪水預報為例
- 結合河川洪流演算與類神經網路模式於不同前導時間之洪水位預報分析
- Neural Network Procedures for Taguchi's Dynamic Problems
- A Fast and Efficient Competitive Learning Design Algorithm Based on Weight Vector Training in Transform Domain
- 專家系統振動訊號圖型判別之研究
- 反傳遞模糊類神經網路於流量推估之應用
- 類神經網路(Neural Networks)的種類及其在影像處理上的應用
- C++Fuzzy類神經網路物件導向發展系統之建立
頁籤選單縮合
題名 | 高解析降雨資料應用於河川洪水位預報=Application of High-resolution Rainfall Data to River Flood Stage Forecasting |
---|---|
作者 | 許銘熙; 傅金城; 黃鵬豪; 林李耀; 蔡孟原; Hsu, Ming-hsi; Fu, Jin-cheng; Huang, Peng-hao; Lin, Lee-yaw; Tsai, Meng-yuan; |
期刊 | 臺灣水利 |
出版日期 | 20100900 |
卷期 | 58:3=231 2010.09[民99.09] |
頁次 | 頁1-16 |
分類號 | 443.42 |
語文 | chi |
關鍵詞 | QPESUMS系統; 類神經網路; 雨量預測; 洪水位預報; QPESUMS; Artificial neural network; Rainfall forecasting; Flood stage forecasting; |
中文摘要 | 台灣位於副熱帶季風區,每年因梅雨及颱風侵襲常帶來豪雨,加上地形陡峻,河川短促,經常引發嚴重水災,年損失約新台幣170億元。若經由即時的河川洪水位預報模式,可適時提出預警資訊以降低洪災損失,然而洪水預報中最主要的關鍵在於降雨量預報。因此本研究主要目的是試著應用高解析降雨的QPESUMS資料於洪水位預報模式以提昇其預報精度。本研究首先利用徐昇多邊形法劃分出地面雨量站對應於地表空間雷達網格的範圍,然後藉由面積權重法嘗試將QPESUMS網格式雷達雨量銜接至單點式的地面雨量站。其次透過等比法進而推測地面雨量站之未來1小時雨量資訊,最後使用類神經網路建立雨量-水位模式,探討增加推測雨量對於模式預報未來3小時水位之影響。由數場颱洪事件的模擬結果顯示,加入1小時推測雨量資訊時,在雨量-水位模式與洪水演算模式不僅能將水位站之計算水位逼近觀測水位,也可使河道各斷面之預報水位剖面線更接近實際的觀測值。 |
英文摘要 | Taiwan is threatened by disastrous storms that result from subtropical cyclones ortyphoons every year. The torrential rainfall of storms which combined the geographicalfeatures such as steep land slope and short rivers course often results in disastrous damageswith losses around 17 billions NT dollars per year. The river stage forecasting model is a usefultool to simulate the river stages in advance for the flood damage mitigation. However, theprediction of rainfall plays an important role in the flood forecasting model, and the predicted accuracy concerns the consequences of flood stage forecast. The main purpose of the study isto apply the high-resolution rainfall data of QPESUMS (Quantitative precipitation estimationand segregation using multiple sensors) to improve the accuracy of river stage forecast. Firstof all, the Thiessen polygons method was used to divide the basin into several control areasbased on the location of rainfall gauge station. Then, the area weighting method was attemptedto convert the QPESUMS radar rainfall grid to a rainfall gauge stations and the 1-hr aheadleading rainfall can be predicted for rainfall gauge stations based on the equal proportionmethod. The rainfall-stage forecasting model with ANNs (Artificial neural networks) wasemployed for the stage prediction at river gage stations. The results from the simulationsof several typhoon events reveal that the 1-hr leading rainfall can improve the accuracy ofrainfall-stage forecasting model by using rainfall data from the QPESUMS. The present modelcan provide the reliable and satisfactory flood stages forecast for flash floods or storms. |
本系統之摘要資訊系依該期刊論文摘要之資訊為主。