臺灣博碩士論文加值系統

在先前的研究案 (李元平, 2005) 中我們以化學還原法合成奈米銀微粒，並以實驗數據建構的模式尋找最適化製造條件。
上述的實驗成果建立在實驗室級實驗數據，也就是在200 cc的量瓶中配藥100 cc在恆溫振盪器中反應取得。當我們要將此實驗室奈米級銀粉製造系統放大(生產2公升或更大量)，攪拌混合狀況的改變，如何以原先建立之數據驅動模式為依據，快速而有效的以放大生產的少量試驗數據找到此放大系統的最適化生產條件是本研究的主要目標。本研究中以一個假設性的例子以及合成奈米銀粒子的實驗作為測試的系統，並由此測試系統可證實此最適化演算法是可行的。

In our previous work (Lee, 2005), optimization on the synthesis of nanosized silver particles by chemical reduction based on the data-driven model was studied. The results of our previous work were carried out based on the laboratory experimental data (loading 100 cc reacting mixtures in a 200 cc beaker and shaking in an isothermal shaker). In this work, the scale-up production (2 liter or more) of nanosized silver particles is desirable. Extension of the identified data-driven model from the laboratory experimental data to find the optimal operating conditions for scale-up production is achieved by the tailored experiment design. A reliable optimum condition for operating a process can be determined on the basis of the sequential constructed data-driven model originated from the identified model of a small-scale system. Only limited experiments are required. A simulated example and the experimental synthesis of nanosized silver particles were adopted as the testing systems. The proposed heuristic optimization algorithm was examined and proven to be applicable.

ACKNOWLEDGMENTS i
ABSTRACT (in English) ii
ABSTRACT (in Chinese) iii
CONTENTS iv
LIST OF FIGURES vi
LIST OF TABLES xi
NOTATION xiii
CHAPTER 1 INTRODUCTION 1
CHAPTER 2 THE EXPERIMENT DESIGN METHODS 6
2.1 Uniform Design 6
2.2 Sequential Pseudo-Uniform Design 10
CHAPTER 3 STATISTICAL ANALYSIS AND OPTIMIZATION OF THE IDENTIFIED MODELS 14
3.1 Regression Model 14
3.2 Artificial Neural Network 17
3.3 Statistics Analysis of the Identified Model 20
3.4 Optimization on the Identified Model by the Random Search Method 27
3.5 Product or Process Optimization Algorithm via the SPUD Method 31
CHAPTER 4 ALGORITHM FOR EXTENDING THE IDENTIFIED MODEL FROM A SMALL-SCALE SYSTEM TO A SCALE-UP SYSTEM 36
CHAPTER 5 APPLICATION OF THE OPTIMIZATION METHOD TO A SCALE-UP SYSTEM FOR THE SYNTHESIS OF NANOSIZED SILVER PARTICLES 65
5.1 Experimental System 65
CHAPTER 6 CONCLUSION 104
REFERENCES 106

1.Chou, K. S.; Chiang, Y. R. Synthesis of Nanosized Silver Particles by Chemical Reduction Method. Materials Chemistry and Physics. 2000, 64, 241.
2.Chang, J. S.; Lin, J. P. Product and Process Development via Pseudo-Sequential Uniform Design. Ind. Eng. Chem. Res. 2004, 43, 4278.
3.Fang, K. T. Uniform Design: Application of Number-Theoretic Methods in Experimental Design. Acta Math. Appl. Sin. 1980, 3, 363.
4.Fang, K. T. and Ma, Z. X. Orthogonal and Uniform Experimental Design; Hong Kong Baptist University: HK, 2000.
5.Hagan, M. T.; Demuth, H. B. Neural Network Design; Pws Publishing: Boston, 1995.
6.Jang, J. S. R.; Sun, C. T.; Mizutani, E. Neural-Fuzzy and Soft Computing; Prentice-Hall International: USA, 1997.
7.Lee, Y. P. Synthesis and Optimization of Nanosized Silver Particles via Sequential Pseudo-Uniform Design Method. Master Thesis, Tatung University, Taipei, Taiwan, 2005.
8.Liang, Y. Z.; Fang, K. T.; Xu, Q. S. Uniform Design and Its Applications in Chemistry and Chemical Engineering. Chem. Int. Lab. Syst. 2001, 58, 43.
9.Solis, F. J.; Wets, J. B. Minimization by random search technique. Math. Opera. Res. 1981, 6, 19.