灰色系統理論應用於大氣污染評估

Tunghai University Institutional Repository > 工學院 > 環境科學與工程學系所 > 碩博士論文 > Item 310901/5721

Please use this identifier to cite or link to this item: http://140.128.103.80:8080/handle/310901/5721

Title:	灰色系統理論應用於大氣污染評估
Other Titles:	Application of Grey System Theory to Assess Atmospheric Pollution
Authors:	薛鼎翰 Shiue, Ding-Han
Contributors:	程萬里 Cheng, Wan-Li 東海大學環境科學與工程學系
Keywords:	灰色系統理論;灰色預測;灰色關聯分析;灰色規劃;TAPM Grey system theory;GM(1,1);GRA;GP;TAPM
Date:	2006
Issue Date:	2011-05-24T08:45:35Z (UTC)
Abstract:	大氣污染物評估是件相當複雜且困難的任務，它包含著許多不確定性與難以量化的因子，因此造成評估上的困難。而灰色系統理論則是以資訊不確定、少數據為基礎所研發的系統。然而，灰色系統於近年來已廣泛使用於環境系統，其中以灰色預測、灰色關聯分析及灰色規劃最為廣泛所應用。因此在本研究中將分別探討這三個模型的優缺點及適用性，並與傳統的分析方法相互比較。本研究共分為三個議題：1)應用灰色預測模型針對以小時及年為單位的污染物濃度預測其未來變化趨勢，探討以長時間及短時間為單位在進行預測評估上之差別，並與傳統的回歸法作準確性比較。2)應用灰關聯分析模型評估中部空品區之空氣品質監測站間的空氣品質優缺點，並以現今環保署使用之污染物標準指標(PSI)相互比較其特性與優缺點。3)利用空氣品質模式(TAPM)模擬污染源排放二氧化硫後對於受體點的貢獻，在將模擬值代入建立之灰色規劃模型中，計算求取當欲新增工廠時舊有的污染源應減量多少，才能在不改變環境現況下求得最佳排放量；並以灰色規劃之灰數原理納入箱型模式中求取大氣涵容能力，使其求得之評估值能更接近現實狀況。分析結果顯示灰色預測之預測值較回歸法精確，而以長時間為單位進行預測之誤差低於以短時間為單位的評估，原因在於短時間的數據變動性較大，若要進行即時預測評估則須了解污染物的變化趨勢。灰色關聯分析空氣品質監測站較PSI效率高，因為PSI只以最大污染物作為評估，而灰色關聯分析則是將5種副指標污染物一同評估，可將各個污染物對於環境之危害程度皆涵蓋分析。而灰色規劃不但能在複雜的條件下求取最佳解，其灰數之原理更能使得分析結果更為客觀，使得結果較傳統的線性規劃更貼近現實狀況。 Assessment of atmospheric pollution is a complex task, and it involves uncertainty and indefiniteness. Therefore, it is difficult to accurately estimate pollution patterns. However, the grey system theory (GST) is a tool that can analyze uncertain and indefinite data. The GST is extensively used in environmental systems, and grey prediction (GM(1,1)), grey relational analysis (GRA) and grey programming (GL) are used in the practical application of GST. The purpose of this study was to evaluate three simulation models of atmospheric pollution and to compare the differences between them and the traditional method of assessment pollution patterns.This research has three studies: 1) the GM(1,1) model to predict yearly and hourly variations in pollutant concentration and to compare forecast results of GM(1,1) and the regression model; 2) GRA to evaluate pollution areas and to compare the effectiveness of GRA and the Pollutants Standard Index (PSI); and 3) The Air Pollution Model (TAPM) applied to grey linear programming as a means to estimate air quality control in central Taiwan. TAPM was used to simulate point sources of pollution which appeared at specific monitoring points. GP was used to calculate the amount of SO2 emission. The optimum amount of emission from point sources was then determined. The merits and defects of each model were also evaluated.Due to GM(1,1) is more accurate than the regression method, long-term units produce fewer errors than short-term units. Because a jump series might reduce the accuracy of GM(1,1), an understanding of variation tendency is needed in order to evaluate short-term pollutant forecasts. GRA is effective in evaluating air quality, the main disadvantage of the PSI is that it can only identify a single pollutant. The GRA information was derived from five sub-indexes of pollutants (O3, NO2, SO2, CO and PM10), and each level of pollutant damage can be displayed by GRA. GP can solve the problem of complex constrained conditions and the uncertainty factor, whereas the grey number can improve objectivity and provide simulations which are closer to real pollutant conditions than those of traditional linear programming.
Appears in Collections:	[環境科學與工程學系所] 碩博士論文

Files in This Item:

File	Size	Format
index.html	0Kb	HTML	299	View/Open

Loading...