题名 |
攔砂壩系對小流域不同時間尺度逕流係數影響研究 |
并列篇名 |
Study on the Effect of Check Dam System on the Runoff Coefficient of Small Watershed in the Loess Plateau |
DOI |
10.6937/TWC.201906_67(2).0010 |
作者 |
袁水龍(SHUI-LONG YUAN);李占斌(ZHAN-BIN LI);李鵬(PENG LI);田珂(KE TIAN) |
关键词 |
攔砂壩系 ; 逕流係數 ; 灰色關聯度分析 ; BP神經網路 ; Check dam system ; Runoff coefficient ; Grey relational analysis ; BP neural network |
期刊名称 |
台灣水利 |
卷期/出版年月 |
67卷2期(2019 / 06 / 01) |
页次 |
83 - 93 |
内容语文 |
繁體中文 |
中文摘要 |
為科學認識黃土高原攔砂壩系建設對小流域不同時間尺度逕流係數影響,本文通過灰色關聯度分析篩選影響小流域多尺度逕流係數的關鍵因素,並基於BP神經網路模型對小流域逕流係數進行模擬和預測。研究表明:(1)1988到2010年,韭園溝流域骨幹壩系總庫容和已淤庫容逐漸增大,剩餘庫容先增大,後保持穩定,最後逐漸減小;(2)韭園溝流域年逕流係數和次降雨逕流係數影響因素的關聯度排序一致,連通性指數、總庫容、剩餘庫容、位置指數4個指標是影響不同時間尺度小流域逕流係數的關鍵因素,能夠較為全面的描述小流域攔砂壩系特徵;(3)BP神經網路模型能夠較好的類比預測年逕流係數和次降雨逕流係數,預測年逕流係數和次降雨逕流係數的均方誤差均較小,年逕流係數和次降雨逕流係數的預測值和實測值線性回歸的相關係數R^2分別為0.99,0.93。研究成果以期為黃土高原攔砂壩系建設提供科學參考。 |
英文摘要 |
In order to scientifically understand the effect of the construction of check dam system on the runoff coefficient of small watershed in the loess plateau at different time scales, this paper screens the main factors which affect the runoff coefficient of small watersheds by grey relational grade analysis, as well as simulates and predicts the coefficients in small watershed based on BP neural network model. The results showed that: (1) From 1988 to 2010, the total storage capacity of the key check dams and the accumulated storage capacity increase gradually, and the remaining storage capacity increases firstly, and then remains stable, while finally decreases in the Jiuyuangou watershed. (2) The correlation degree ranking of the influence indexes of annual runoff coefficient and individual rainfall runoff coefficient is consistent in the Jiuyuangou watershed. Connectivity index, total storage capacity, remaining storage capacity and position index are the key factors affecting the runoff coefficient of small watershed at different time scales, which express the effect of check dam system on runoff coefficient. (3) BP neural network model can simulate and predict annual runoff coefficient and individual rainfall runoff coefficient effectively, which means square error is small. BP. The correlation coefficient R^2 of the predicted and measured values of annual runoff coefficient and rainfall runoff coefficient are 0.99 and 0.93 respectively. The research results are intended to provide a scientific reference for the benefit evaluation of the construction of check dams on the loess plateau. |
主题分类 |
工程學 >
水利工程 |
参考文献 |
|