頁籤選單縮合
題 名 | 類神經網路結合半變異圖模式於地下水水位預測之研究=A Study on Ground Water Level Forecasting by Combining Neural Networks and Semi-variogram Model |
---|---|
作 者 | 陳清田; 陳儒賢; 黃建演; 黃上竹; | 書刊名 | 農業工程學報 |
卷 期 | 61:1 2015.03[民104.03] |
頁 次 | 頁14-28 |
分類號 | 443.67 |
關鍵詞 | 地下水預測; 倒傳遞類神經網路; 半變異圖模式; 試誤法; Groundwater level forecasting; Back-propagation neural network; Semi-variogram model; Try and error; |
語 文 | 中文(Chinese) |
中文摘要 | 本研究嘗試結合倒傳遞神經網路(back-propagation neural network; BPN)、克利金法(kriging technique)之半變異圖(semi-variogram)模式與試誤法(try and error),建立一地下水位預測模式,並將此模式應用於台南地區二十個地下水位測站之月平均地下水位預測上。研究結果顯示,資料若先經過克利金法所計算出之影響範圍(influence range)篩選後,的確有助於BPN模式之地下水位預測,然本研究地區之影響範圍並不明顯,故本研究再經由試誤的方式篩選影響範圍內之地下水位資料。由研究結果得知,經過試誤法篩選後,有十八個測站BPN模式之效率係數(coefficient of efficiency; CE)值大於0.7,且有九個測站之CE值提高10%以上。此外,輸入BPN模式之相關測站數目也由原本平均8站減少到平均2站。上述結果顯示,經由半變異圖模式所計算出之影響範圍內之測站資料,若再經由試誤法篩選後,不僅可以去除影響模式精確度之干擾項,提高預測結果之精確度,且可以有效地減少模式訓練時間以及資料成本的使用。 |
英文摘要 | In this study, based on the combination of the the back-propagation neural network (BPN), semi-variogram model and try and error method to propose a groundwater level forecasting model. The proposed model is applied to actual groundwater data from 20 groundwater stations of the Tainan area. The study results show that data of selected stations within influence range obtained by kriging method is indeed helpful to improve performance of BPN model in forecasting groundwater level. However, the influence range in the study area is unobvious. Try and error method is used to select groundwater level data within the influence range. The results show that there are 18 stations achieve the coefficient efficiency, CE, of over 0.7. Moreover, the 9 stations yield an average CE upgradation 10%. In addition, the average number of related stations is reduced eight to two. The results show that the proposed method can not only delete noise to the network and produce reasonable forecasts, but also can save training time of model and decrease costs of data collection and management. |
本系統中英文摘要資訊取自各篇刊載內容。