臺灣博碩士論文加值系統

焚化廢氣中極微量的戴奧辛(dioxin)污染物即會對人類及其環境造成極大的危害，焚化系統中戴奧辛之生成與排放的控制已是刻不容緩。然而，戴奧辛在焚化系統之生成機制甚為複雜，文獻研究探討雖多，對其形成過程仍然未窺全貌，無法經由數學模式之建立(deterministic approach)做有系統之模擬與控制。本研究擬以類神經網路(neural network)模擬焚化系統戴奧辛生成，其終極目的在發展出一套具人工智慧之控制系統。
本研究以類神經網路分別完成模擬兩座文獻中有關焚化系統中戴奧辛生成之探討，一為試驗規模焚化爐；另一為商業較規模焚化爐。使用工具為美國NeuralWare公司所出版的NeuralWorks Professional Ⅱ/PLUS。研究結果均顯示類神經網路模擬結果優於文獻中利用其他統計方法之結果，能夠更精確建立焚化系統戴奧辛生成之因果關係。顯然地，本研究充分顯示類神經網路模擬焚化系統中戴奧辛之生成深具潛力，未來將廣泛蒐集戴奧辛之生成資料，計劃實驗，以類神經網路為模擬工具做更進一步之探討。同時佐以合宜之控制策略，建立具人工智慧之控制系統。

Trace amounts of dioxins emitted from incineration facilities present a great threat to human beings and the environment. An efficient control system based on deterministic approach is, however, not available due to the lack of knowledge regarding the formation of dioxins during waste incineration. The purpose of this study is to render the neural networks approach to train and propose a relationship between the formations of dioxins to the operating parameters of an incineration system. The ultimate concern is to develop an expert system to monitor and control the dioxin emissions.
In this study, a pilot-scale incinerator and a commercial-scale one are simulated using neural networks. Simulated results are compared with that of the original authors to testify the feasibility of using neural networks for such analysis. The NeuralWorks Professional II/PLUS by American NeuralWare Company was used for all the computations. Results from this study indicate that neural networks shows better stability and convergence than other statistical approach used by the original authors. More case studies, however, are required to further testify these findings.

目錄
頁次
中文摘要………………………………………………………….……………iii
英文摘要………………………………………………………….…………….iv誌謝……………………………………………………………………………...v
目錄………………………………………………………………...…………...vi表目錄……………………………………………………………….…..……...ix
圖目錄………………………………………………………………..……….…x
第一章緒論………………………………………………………….…………1
1-1 研究緣起...……………………………………………………………...1
1-2 研究動機與方法…...……………………………………….…………..6
第二章文獻回顧…………………………………………………….…………7
2-1 戴奧辛……………………………………………………….………….7
2-1-1 戴奧辛產生之生成………………………………….………….8
2-1-2 戴奧辛之濃度單位…………………………….……………….9
2-1-3 戴奧辛之特性……….…………………………….…………..10
2-1-4 戴奧辛在環境中之移動性………………………….…………11
2-1-5 戴奧辛對人體危害性及其排放標準…...…………………….11
2-2 焚化操作條件與戴奧辛生成之影響……….….………...…………...12
2-3 類神經網路……………………………………..……………………..14
第三章 戴奧辛生成模擬……………………………..……………………….17
3-1 類神經網路模擬流程………………………………...……………….17
3-1-1 問題分析…………………..…………….……………………...18
3-1-2 資料收集和整理………………………………………………..19
3-1-3 類神經網路架構………………………………………………..19
3-1-4 系統整合和維護………………………………………………..20
3-2 模擬個案………………………………………………..…………….20
3-2-1 試驗規模焚化系統………..……………………………………20
3-3-2 商業規模焚化系統………..……..…………..………………....26
第四章 結果與討論………………………………….………………………..29
4-1 試驗規模焚化系統之模擬……………………………………………29
4-1-1 六個操作條件－（爐床溫度、氧氣濃度、boiler出口溫度，鹽酸濃度、水的含量及熱交換器中滯留時間）………..………29
4-1-2 五個操作條件－（氧氣濃度、boiler出口溫度，鹽酸濃度、水的含量及熱交換器中滯留時間）……………….……..…...….30
4-1-3 四個操作條件－（氧氣濃度、boiler出口溫度，鹽酸濃度及熱交換器中滯留時間）…………. …...……………………….….30
4-1-4 三個操作條件－（氧氣濃度、boiler出口溫度及熱交換器中滯留時間）……..………….………………..……….…………….31
4-1-5 二個操作條件－（boiler出口溫度及熱交換器中滯留時間）……………….……………………………………………..31
4-2 商業規模焚化系統……………………..……………………………..34
第五章 結論與建議……………………………………………………….…..46
5-1 結論…………………………………………………………….……...46
5-2 建議…………………………………………………………………....47
參考文獻……………………………………………………………………….48
附錄一. 類神經網路…………………………………………………………..51
附錄二. 類神經網路特性分析………………………………………………..58
附錄三. 試驗規模焚化系統之類神經網路轉換成C語言網路型式………..60
附錄四. 商業規模焚化系統之類神經網路轉換成C語言網路型式………..64

中綱科技股份有限公司，”類神經網路建構工具” (1995).
行政院環境保護署環境衛生及毒物管理處，”戴奧辛環境毒理簡介” (2000).
美國環保署，http://www.epa.gov/ncea/dei.htm
環保署全球資訊網，http://www.epa.gov.tw
Addink , R. , and K. Olie , “Role of Oxygen in Formation of Polychlorinated Dibenzo-p-dioxins / Dibenzofurans from Carbon on Fly Ash,” Environ. Sci. Technol. , 29 (6) , 1586 (1995).
Addink , R. , W. C. M. Bakker , and K. Olie , ”Influence of HCl and Cl2 on the Formation of Polychlorinated Dibenzo-p-dioxins / Dibenzo-furans in a Carbon / Fly Ash Mixture, ” Environ. Sci. Technol. , 29(8) , 2055 (1995).
Baughman , D. R. , and Y. A. Liu , Neural Networks in Bio-processing and Chemical Engineering , Academic Press. Inc. (1995).
Bhat , N. V. , P. A. Minderman , and T. J. McAvoy , “Use of Neural Nets for Dynamic Modeling and Control of Chemical Process Systems,” Comput. Chem. Eng. , 14 , 573 (1990).
Deshpande , P. B. , and R. H. Ash , Computer Process Control with Advanced Control Applications , Instrument Society of America (1988).
Fängmark , I. , B. van Bavel , S. Marklund , C. Rappe , B. Strömberg , and N. Berge , ” Influence of Combustion Parameters on the Formation of Polychlorinated Dibenzo-p-dioxins , Dibenzofurans , Benzenes , and Biphenyls and Polyaromatic Hydrocarbons in a Pilot Incinerator,” Environ. Sci. Technol. , 27(8) , 1602 (1993).
Fängmark , I. , C. Rappe , B. Strömberg , and N. Berge , “Influence of Post- combustion Temperature Profiles on the Formation of PCDDs , PCDFs , PCBzs , and PCBs in a Pilot Incinerator,” Environ. Sci. Technol. , 28(4) , 624 (1994).
Hoskins , J. C. , and D. M. Himmelblau , “Artificial Neural Network Models of Knowledge Representation in Chemical Engineering,” Comput. Chem. Eng. , 12 , 881 (1988).
Huang , C. -T. , and Y. -S. Lin , “Turning PID Controller for Open-Loop Unstable Processes with Time Delay,” Chem. Eng. Comm. , 133 , 11 (1995).
Huang , C. -T. , C. -J. Chou , and J. -L. Wang , “Turning of PID Controllers Based on the Second-Order Model by Calculation,” J. Chin. I. Ch. E. , 27(2) , 107 (1996).
Ishikawa , R. , A. Buekens , and K. Watanabe , “ Influence of Combustion Conditions on Dioxin in an Industrial-Scale Fluidized-Bed Incinerator : Experimental Study and Statistical Modeling,” Chemosphere , 35(3) , 465 (1997).
Lee , M. , and S. Park , “A New Scheme Combing Neural Feedforward Control with Model-Predictive Control,” AICHE J. , 38 , 193 (1992).
McFarland , R. C. , and D. E. Rivera , “Identification of Distillation Systems,” in W. L. Luyben , Editor , Practical Distillation Control , Van Nostrand Reinhold (1992).
Oh , J. E. , K. T. Lee , J. W. Lee , and Y. S. Chang , “The Evaluation of PCDD/Fs from Various Korean Incinerators,” Chemosphere , 38(9) , 2097 (1999).
Pollard , J. F. , M. R. Broussard , D. B. Garrison , and K. Y. San , “Process Identification Using Neural Networks,” Comput. Chem. Eng. , 16 , 253 (1992).
Prett , D. M. , T. A. Skrovenek , and J. F. Pollard , “Process Identification-Past , Present , Furture,” Shell Process Control Workshop (1986).
Savkovic-Stevanovic , J. , “Neural Networks for Process Analysis and Optimization : Modeling and Application,” Comput. Chem. Eng. , 18 , 1149 (1994).
Simpson , P. K. , Artificial Neural Systems , Pergamon Press. New York (1990).
Song , J. J. , and S. Park , “Neural Model Predictive Control for Nonlinear Chemical Process,” J. Chem. Eng. Japan , 26 , 347 (1993).
Su , H. T. , and T. J. McAvoy , “Integration of Multilayer Preceptron Networks and Linear Dynamic Models : A Hammerstein Modeling Approach,” Ind. Eng. Chem. Res. , 32 , 1927 (1993).
Tagashira , K. , I. Torii , K. Myouyou , and K. Takeda , “Combustion Characteristics and Dioxin Behavior of Waste Fired CFB,” Chem. Eng. Sci. , 54 , 5599 (1999).
Venkatasubramanin , V. , R. Vaidyanathan , and Y. Yamamoto , “Process Fault Detection and Diagnosis Using Neural Network-I. Steady-State Processes,” Comput. Chem. Eng. , 14 , 699 (1990).