固體廢棄物管理系統整合模式之修正研究

An Improved Integrated Model for Solid Waste Management

蘇俊賓(Jun-Pin Su)

指導教授 ： 馬鴻文

共同指導教授 ： 於幼華(Yue-Hwa Yu)

國立臺灣大學/工學院/環境工程學研究所/博士(2009年)

https://doi.org/10.6342/NTU.2009.00895

摘要

近年來，隨著環境變遷，固體廢棄物產出的質與量不斷改變，其所衍生的相關問題也日益多元，雖有許多廢棄物管理技術、工具已陸續發展，相關工具是否能妥適解決日益複雜的問題，已日益受到關注。如何提供妥適的固體廢棄物管理政策，邁向循環型社會，為本研究所關注的核心問題。
本研究考量政策之環境面、社會面、技術面以及兼顧衝突特性等因素進而提出固體廢棄物管理系統整合模式之修正。此模式較其他固體廢棄物管理模式不同之修正特色有三，第一為本模式以模糊多準則評估模式為架構，建立政策衝擊潛勢分析方法，輔助既有之社會面評估方法。第二為本模式為了有利區域之間同類型固體廢棄物政策之經驗移轉，提出以區域資料庫為基礎，結合類神經網路與複回歸推估技術，作為評估之依據。第三為本模式增加中長程整合性推估與決策功能，修正目前之廢棄物管理模式較欠缺整合性之問題。
本研究整合上述三項修正，在決策規則輔助下建立具有中長程整合預測功能、區域經驗能有效移轉以及以政策衝擊潛勢方法輔助社會面評估之固體廢棄物政策支援系統。並以桃園縣垃圾費隨袋徵收、台北市垃圾焚化飛灰再利用、以及台北縣廢棄物中長程規劃等三個案例進行模式之案例研究，驗證本方法之實用性。
案例研究結果顯示，政策衝擊潛勢分析適度輔助決策支援系統，可供決策者較多元之參考依據，而在區域資料庫較完備的情形下，類神經網路結合複回歸預測，的確有助於不同區域之間廢棄物政策之經驗移轉，而面對廢棄物政策之整合性需求，本研究所建立之中長程廢棄物評估方法為可考量之決策評估方法。本研究以系統軟體Stella為工具，建立永續廢棄物政策決策支援系統。

關鍵字

多準則決策 ； 模糊層級分析法 ； 垃圾費隨袋徵收 ； 類神經網路 ； 廢棄物管理 ； 決策支援系統

並列摘要

Both the quantity and quality of municipal solid waste (MSW) have changed with time. The problems of municipal solid waste management (MSWM) become multiple. While many technologies and models on waste management have been developed, more and more attentions have been paid to whether these tools are appropriate to solve increasingly complex problems. The core concern of this research is to deal with the issues of how to plan a sustainable waste management policy for Taiwan for the purpose of approaching.
An improved integrated model for solid waste management has been provided in this study. This model has considered and integrated the environmental, social, and technical aspects of waste policies as well as the characteristic of potential conflicts. Compared to other models, this model has three characteristics as follows. First, fuzzy MCDM has been used in this model as a basic framework to build the policy impact potential analysis (PIPA) method to assist in the assessment of influences from social conflict toward each case. Second, the SOM neural network is combined with multiple regression for estimating MSW based on a regional database to assist in the transfer of regional experience. Third, in order to enhance integration, the mid- and long-term planning is also incorporated in the model.
This study tried to set up a policy decision-making support system. The system can provide a long-term forecast function, an model on transfer of regional experiences, and an analysis of potential policy impacts. This study examined the practicability of the model by testing case studies, including “fly ash regeneration in Taipei city”, “regional PAYT policy assessment in Taoyuan county”, and “mid-term and long-term regional waste treatment policies in Taipei county”.
The results of the case studies showed that policy impact potential analysis (PIPA) can be valuable to the decision-making support system. It can provide decision-makers with guidelines of communication when other information was not available. The SOM neural network and the multiple regression for estimating MSW was useful for the transfer of regional experience if the regional database was sufficient. The integrated long-term model has also been demonstrated to be one of the models to idenfify the key points of long-term waste management. This study used Stella as tool for building the improved Decision Support System of Sustainable Waste Management.

並列關鍵字

Multiple Criteria Decision Making ； Fuzzy Analytic Hierarchy Process (Fuzzy AHP) ； PAYT ； SOM neural network ； waste management ； decision support system

參考文獻

行政院衛生署，「都市垃圾處理計畫」，1986年。

行政院環境保護署，「一般廢棄物(垃圾)收費技術研究」」，1997年。

桃園縣政府環保局，「桃園縣一般廢棄物處理工作調整暨清除處理成本費率計算及徵收計畫」，2002年。

AbuQdais, H.A., Hamoda, M.F., Newham, J. Analysis of residential waste at generation sites. Waste Management and Research 1997; 15 (4): 395–406.

Bach, H., Mild, A., Natter, M., Weber, A. Combining Socio-demographic and logistic factors to explain the generation and collection of waste paper. Resources, Conservation and Recycling 2004; 41 (1): 65–73.