A Study of Modeling Development for High Voltage DMOS Transistors by Using Fuzzy Theory and Neural Network
陳盈德、陳勝利
E-mail: 9019868@mail.dyu.edu.tw
ABSTRACT
In recent years, a semiconductor device dimension in ultra large-scale integrated circuits (ULSI’s) have been kept shrinking in order to achieve higher device/circuit density and to reduce the product costs. But it has brings more nonlinearity effects. These effects resulting in the performance of actual device have some difference as compare with the ideal device. Therefore, the modeling development of the semiconductor device has become more and more important. In this thesis, a new methodology is proposed to modeling the highly voltage DMOS by using Adaptive Neuro Fuzzy Inference System (ANFIS), which combines fuzzy theory and adaptive neural network, and predicting device behaviors in the power DMOS with different channel length and channel width under different bias situations. At first, two groups of power DMOS devices with different channel length and channel width were fabricated separately. In the following, the drain current of all devices under different bias conditions were measured. And then input the measurement data treat as the training data. The experimental results have proven that the power of ANFIS used as a realization of I-V characterizations. The prediction results are also compared with experimental data of the actual devices, eventually, which can be obtained a good agreement.
Keywords : fuzzy theory ; neural network
Table of Contents
封面內頁 簽名頁 授權書 iii 中文摘要
iv 英文摘要 v 誌謝
vi 目錄 vii 圖目錄
ix 表目錄 xi 第一章
緒論 1 第二章功率電晶體的結構與發展 4
2.1功率電晶體的發展 4 2.2功率電晶體的結構 6 第三章 理
論原理 11 3.1 模糊理論 11 3.1.1 模糊
理論的特徵 11 3.1.2 模糊集合及明確集合 12 3.1.3 鐘型歸屬函數之定義
13 3.1.4 模糊集合常用之基本演算 15 3.1.5 模糊理論的功效
18 3.1.6 模糊預測(Prediction Model)的設計 18 3.2 類神經網路理論 19 3.2.1 類神
經網路的應用範圍 19 3.2.2 類神經網路及函數應用 19 3.3 適應性類神經模糊推論系
統 24 第四章 元件製作與量測 28 4.1 元件尺寸
28 4.2 元件量測 30 4.3 ANFIS訓練
33 第五章 結果討論與分析 34 5.1 ANFIS模擬結果
34 5.2 結果討論 43 第六章結論
48 參考文獻 49 附錄
51 REFERENCES
[1]A. Rofougaram and A. A. Abidi, "Table Lookup FET Modelling for Accurate Analogue Circuit Simulation", IEEE Trans. Circuit CAD, CAD-3, p. 72, 1984.
[2]B. J. Sheu, D. L. Scharfetter, P. -K. KO, and M. –C. Jeng, "BSIM: Berkeley Short Channel IG-FET Model for MOS Transistors", IEEE J.
Solid Circuit, SC-22, p.558, 1983.
[3]P.Antognetti and G. Massobrio, Semiconductor Device Modelling with SPICE, McGraw-Hill, New York, 1988.
[4]J. B. Roldan, F. Gamiz, and J. A. Lopez-Villanueva, "A Closed-Loop Evaluation and Validation of a Method for Determining the Dependence of the Electron Mobility on the Longitudinal-Electric Field in MOSFET's," IEEE Trans. Electron Devices ,vol.44,p.1447 , 1997.
[5]HSPICE Dser's Manual 9001, Meta-Software, Inc., 1990.
[6]J-S. R. Jang, C. T. Sun and E. Mizutani, “Neuro-Fuzzy and Soft Computing”, Prentice-Hall, Inc., 1997 [7]李宏俊，“高功率金氧半場效 電晶體製程技術及發展趨勢”，電力電子技術，No.55，pp.10~22，2000 [8]P. Rossel, H. Tranduc, G. Charitat,“Power MOS
Devices:Structures and Modelling”, Microelectronics, 1995. Proceedings., 1995 20th International Conference on Volume: 1 , 1995 , Page(s): 341 -352 vol.1 [9]劉中民，“公元2000年之功率元件技術發展趨勢”，電力電子技術，No.55，pp.10~22，2000 [10]" Neuro-Fuzzy AND Soft Computing" by J. S. R. JANG, C. T. SUN, E. MIZUTANI.
[11]Fuzzy Logic Toolbox for use with MATLAB, The Math Works, Inc., 1995.
[12]Neural Network Toolbox for use with MATLAB, The Math Works, Inc., 1994.
[13]D. Pantic, S. Milenkovic, T. Trajkovic, V. Litovski, N. Stojadinovic,“Inverse Modeling of Semiconductor Manufacturing Processes by Neural Networks”, Microelectronics, 1995. Proceedings., 1995 20th International Conference on Volume: 1 , 1995 , Page(s): 321 -326 vol.1 [14]華建智
，葉豐輝，蔡慧駿，李經綸，"適應性類神經模糊控制系統於盲用點字印表機之應用"，1998自動控制研討會， pp.294-299。
[15]Detlef Nauck, Rudolf Kruse, "A Neural Fuzzy Controller Learning by Fuzzy Error Propagation", NAFIP ’92, Puerto, pp.183-184 Mexico, Dec14-17 .
[16]M.SUGENO and C.T.KANG, "STRUCTURE IDENTIFICATION OF FUZZY MODEL", Elsevier Science Publishers B.V., 1988.
