透過您的圖書館登入
IP:3.145.74.54
  • 學位論文

結合時頻分析於地下水集塊系統模式之建立與應用

Development and Application of a Groundwater Lumped System Model by Combining Time Frequency Analysis

指導教授 : 徐年盛

摘要


本研究使用地下水庫概念建立一地下水集塊系統模式,並結合訊號分析方法使用疊代高斯濾波及刻痕濾波從地下水觀測資料找出代表人為一天一次之抽水行為,稱之為抽補強度,接著使用最佳化法優選出模式之參數。以台中盆地為範例,蒐集其地面與地下水資料建立模式,比較優選參數後之模式結果,最後將其應用於補遺地下水位。   本模式是假設地下水為一封閉系統,考慮其出流量與入流量對水位之影響,並將模式中物理參數公式化,並經由測試找出各參數限制式之範圍來優選出合理之參數,其物理因子分別考慮地下水流失、降雨補注之入滲、人為抽水行為。   在建置台中盆地之四種不同模式中,分為線性模式、非線性模式、非線性配合乾濕季模式探討、線性配合乾濕季模式探討,結果以非線性配合乾濕季模式所模擬出地下水位最佳,尤其將參數分為乾濕季兩組更能使得模式明顯改善,接著把此模式應用補遺地下水位,以台中建平水位站做為範例。為了測試補遺之效果引進統計之概念,將已知資料假設為缺漏並隨機分布在時間序列中,配合著蒙地卡羅試驗中結果可發現,在5%~15%之缺漏資料情形下,補遺可達到相近的精度,亦即本模式在15%缺漏下能提供穩定的結果。

並列摘要


In this study, iterative Gaussian filter and Notch filter are applied to groundwater level data to find the human effect with the frequency of pumping per day, called pumping recovering strength (PRS). Once the PRS is determined, we develop a model based on the concept of groundwater reservoir and optimal process. In order to validate this model, we take the groundwater data from Taichung basin as example. We compare different types of model, and furthermore, we apply this model to groundwater data supplementing. We assume the groundwater is a closed system ,only consider the effect of ouflow and inflow,and formulate the physical parameters in the model. The meaning of the parameters including groundwater outflow、groundwater recharge、effect of human pumping. In the case of Taichung basin, we develop four different types of model. The first is linear model, the second is nonlinear model, and the third is nonlinear model comnibed with dry and wet season, the last is linear model comnibed with dry and wet season. And the best results of these models is nonlinear model comnibed with dry and wet season. Then,we used one of the groundwater staions in Taichung basin which is called Jianping to apply nonlinear model comnibed with dry and wet season to groundwater data supplementing. According to Monte Carlo method, we can find that the consequences after data supplementing are close in the range of 5% to 15% missing data . Therefore,we consieder this model have a good supplementing in 15% data missing.

參考文獻


14. 鄭筱玫,「抽水井群最佳化操作方式之研究」,國立臺灣大學工學院土木工程學系 ,碩士論文,2013。
12. 廖玲琬、 洪銘堅、王逸民、徐年盛、 游雅淳、劉宏仁,「以頻譜分析法評量地下水位時空變動」,工程環境會刊 28期, P.1 – 14,2012。
6. 陳文福、曹鎮、 林志憲,「頻譜與濾波分析應用於地下水位變動研究」,中興工程,P.79 – 85,2007
10. 陳良榮,「使用迭代式高斯法與傾斜極值篩選法解決經驗模態分解法中的混波現象」,國立臺灣師範大學機電科技學系,碩士論文,2011。
15. 戴瑜萱,「地下水位變化與地表高程變動之關聯性分析—以濁水溪沖積扇為例¬¬¬」,國立臺灣大學工學院土木工程學系 ,碩士論文,2015。

延伸閱讀