請用此 Handle URI 來引用此文件:
http://tdr.lib.ntu.edu.tw/jspui/handle/123456789/36898
標題: | 模糊理論於地下水採樣規劃之研究 The Sampling Planning for Groundwater Risk Assessment using Fuzzy Theory |
作者: | Jen-Wei Huang 黃仁緯 |
指導教授: | 馬鴻文 |
關鍵字: | 不確定性,模糊集理論,條件模擬,地下水污染, Uncertainty,Fuzzy set theory,Conditional simulation,Groundwater contamination, |
出版年 : | 2005 |
學位: | 碩士 |
摘要: | 採樣為地下水整治系統中不確定性降低的方法之一,研究中嘗試利用模糊理論所提供的方法將不確定性量化,利於決策者以風險不確定性的降低為基礎,進行後續的整治工作。提出的方法為將採樣資訊量的增加所反應出風險值變異係數的降低,藉由模糊化的過程轉換成為語言項的變數,即風險評估結果的可信度,再經由模糊綜合評判的方式進行不確定性的量化。在研究中同時考慮多個水文地質參數,並配合條件模擬的方式消除空間變異的影響,探討最佳採樣點的配置方式以降低最大的不確定性。
結果顯示,在所選用的參數中,對結果不確定性影響最大的參數為水力傳導係數,其次為總體密數、有效孔隙率及縱向延散係數。而在以此四項參數為基礎對不確定性進行評判時,部份的評判模型會對風險評估結果的可信度造成不適當的描述,造成此影響的主要原因為參數權重分配所致。另外,以「場」的概念對參數進行設定時,在後續蒙地卡羅的模擬上,需至少進行5000次的模擬,始可於後續風險值機率分布的推估上,達到模糊評判的可信度結果。 在採樣位置的配置上,若以風險不確定性的降低做為採樣策略擬定的基礎,研究結果顯示,採樣點位主要集中在污染源與污染接受者附近。除此之外,產生隨機變域場(N)乃為降低參數空間變異的影響,而N的數目會造成採樣點配置順序的不同,但對於整體採樣配置並無明顯的差異存在。 Sampling is a method to value the uncertainty in groundwater remediation system. This study presents a method using fuzzy set theory to quantify the uncertainty of parameters and facilitate the subsequent remediation. This method firstly takes the coefficient of variation (CV) of risk as inputs of membership function and then transfers it to linguistic variables, credibility, and finally quantifies the uncertainty by Fuzzy-Based Comprehensive Assessment Theory. Besides, in order to reduce the influences of space variations and uncertainty, we couple random field generation and conditional simulation procedure to obtain several conditional realizations to allocate the optimal sampling positions in the study. The case study shows that the major parameters affecting cancer risk are hydraulic conductivity, bulk density, effective porosity and longitudinal dispersivity, according to priority. And some fuzzy models would have inappropriate credibility assessment results. Then, when the parameters are considered as “fields” instead of “values”, it suggests that Monte Carlo simulations should be modeled more than 5000 times to make the probability distribution of risk stable. Based on the reduction of risk uncertainty, the results also show that the optimal sampling positions are better located in the pollution sources and pollution acceptors or the neighborhood. In addition, the number of random fields would not influence the overall sampling configuration but the sequence of sampling positions. |
URI: | http://tdr.lib.ntu.edu.tw/jspui/handle/123456789/36898 |
全文授權: | 有償授權 |
顯示於系所單位: | 環境工程學研究所 |
文件中的檔案:
檔案 | 大小 | 格式 | |
---|---|---|---|
ntu-94-1.pdf 目前未授權公開取用 | 1.04 MB | Adobe PDF |
系統中的文件,除了特別指名其著作權條款之外,均受到著作權保護,並且保留所有的權利。