跳到主要內容

簡易檢索 / 詳目顯示

回結果列表
研究生： 李建民
Chien-Min Lee
論文名稱： 使用進化演算法的模糊化類神經網路等化器
指導教授： 賀嘉律
Chia-Lu Ho
口試委員:
學位類別： 碩士
Master
系所名稱： 資訊電機學院 - 電機工程學系
Department of Electrical Engineering
畢業學年度： 89
語文別： 中文
論文頁數： 76
中文關鍵詞： 模糊化類神經網路 、類神經網路 、符元間干擾 、進化演算法
外文關鍵詞： neuro-fuzzy network, neural network, ISI, EAs
相關次數： 點閱：16  下載：0 
分享至:
查詢本校圖書館目錄 查詢臺灣博碩士論文知識加值系統 勘誤回報

    • 傳統上，等化器的設計非常簡單，但通常僅能處理線性判別區域的信號空間。本論文將介紹結合模糊理論與類神經網路架構的等化器，解決等化器無法處理非線性判別區域的的問題。模糊系統的優點是不需要精確的數學模型，另一方面結合人類的知識於系統的設計上。模糊化的好處是可以提供更佳的推廣性、錯誤容忍度、以及更適合應用於真實世界中的非線性系統。而類神經網路的架構，其複雜度可分割非線性判別區域。論文中並提出一種進化演算法則（Evolutionary Algorithms, EAs），應用於模糊化類神經網路等化器上，進化演算法則是一種隨機最佳化(stochastic optimization)的技術，模仿生物遺傳機制的基因進化概念而來，屬於一種多點平行式的全域搜尋（global search）法則。文中將以模擬的方式比較使用進化演算法與傳統演算法，對於模糊化類神經網路等化器效能表現的優劣。

    目　　錄 第一章　緒論..................................................1 1.1 前言....................................................1 1.2 符元間干擾..............................................2 1.3 等化器..................................................3 第一章類神經網路...........................................5 2.1 建構類神經元模型..........................................6 2.2 判別迴授等化器............................................7 2.3 最佳判別界限.............................................10 2.4 多層感知器...............................................13 第二章模糊化類神經網路....................................17 　3.1 模糊系統...............................................17 3.2 模糊化類神經網路.........................................18 3.3 模糊化類神經等化器.......................................21 第三章學習演算法..........................................24 4.1梯度坡降演算法............................................25 4.1.1 BPN learning algorithm for DFE........................25 4.1.2 BPN learning algorithm for MLP-based equalizer.........26 4.1.3 BPN learning algorithm for neuro-fuzzy equalizer.......33 4.2 進化演算法...............................................39 4.2.1 個體的描述 Individual..................................42 4.2.2 評估 Evaluation function...............................42 4.2.3 交配 Crossover.........................................43 4.2.4 突變 Mutation..........................................44 4.2.5 子代選擇 Offspring Selection...........................45 4.2.6 族群 Population........................................45 4.2.8 風險分析 Risk analysis.................................45 第四章模擬結果 Simulation results.........................47 5.1 收斂特性 Convergence characteristics.....................47 5.1.1 適存度 Fitness.........................................49 5.1.2 均方差 Mean square error...............................53 5.2 判別區間 Decision region.................................60 5.3 位元錯誤率 Bit error rate performance....................66 結論.........................................................76 參考文獻 圖目、表目 圖1.1　檢波程序所產生的符號間干擾.............................3 圖1.2　適應性等化器運作流程圖.................................4 圖1.3　(a).發射端傳送之二元信號 (b).受雜訊干擾之信號 (c).經等化器恢復之信號................................................4 圖2.1　類神經元模型...........................................6 圖2.2　活化函數...............................................6 圖2.3　感知器之架構方塊.......................................7 圖2.4　判別迴授等化器的基本架構圖.............................7 圖2.5　兩個輸入的感知器再不同數目的隱藏層時對判別區域的表現...9 圖2.6　通道在信號雜訊比(signal noise ratio, SNR)為20dB時，不同延遲情況下之均方差(Mean Square Error, MSE).....................11 圖2.7　不同雜訊強度下，通道之最佳判別區域圖..................12 圖2.8　類神經網路的前向傳遞模式..............................13 圖2.9　設計類神經網路的步驟與流程............................14 圖2.10　類神經網路等化器.....................................15 圖3.1　模糊化類神經網路的架構................................19 圖3.2　模糊化類神經等化器的架構..............................21 圖4.1　類神經網路的迴授傳遞模式..............................27 圖4.2　不同的p值下 與 的關係.................................32 圖4.3　進化演算法流程圖......................................41 圖4.4　進化演算法虛擬程式碼..................................41 圖5.1　Population learning curve for neuro-fuzzy model-based on linguistic model（fuzzy rule=3）when population size=(200,200)....................................................49 圖5.2　Population learning curve for neuro-fuzzy model-based on linguistic model（fuzzy rule=6）when population size=(200,200)....................................................50 圖5.3　Population learning curve for neuro-fuzzy model-based on linguistic model（fuzzy rule=9）when population size=(200,200)....................................................51 圖5.4　Population learning curve for neuro-fuzzy model-based on linguistic model（fuzzy rule=12）when population size=(200,200)....................................................52 圖5.5　Simulation results showing relative convergence rate performance for DFE and MLP structure........................53 圖5.6　Neuro-fuzzy model-based on linguistic model simulation results showing relative convergence rate performance for different fuzzy rules at SNR = 20dB..........................54 圖5.7　Simulation results showing relative convergence rate performance by (4,1)DFE with (9,3,1)MLP structure for different population sizes at SNR = 20dB...............................55 圖5.8　Simulation results showing relative convergence rate performance by neuro-fuzzy model-based on linguistic model for different population sizes at SNR = 20dB.....................55 圖5.9　EA搜尋完畢後進入fine turning 程序.....................56 圖5.10　Convergence rate performance for BPN and EAs (search miss) for (4,1)DFE with (9,3,1)MLP structure.................57 圖5.11　Convergence rate performance for BPN and EAs (search success) for (4,1)DFE with (9,3,1)MLP structure..............58 圖5.12　Convergence rate performance for BPN and EAs (search success) for neuro-fuzzy equalizer-based on linguistic model....................................................59 圖5.13　Decision region formed by (2,0)DFE with (9,3,1)MLP structure....................................................61 圖5.14　Decision region formed by (2,0)DFE with (9,3,1)MLP structure after 300 generations by Eas.......................62 圖5.15　Decision region formed by neuro-fuzzy equalizer-based on linguistic model (fuzzy rule=6) after 5000 samples........63 圖5.16　Decision region formed by neuro-fuzzy equalizer-based on linguistic model (fuzzy rule=12) after 5000 samples.......64 圖5.17　Decision region formed by neuro-fuzzy equalizer-based on linguistic model (Fuzzy rule=6) after 300 generations by Eas....................................................65 圖5.18　Performance of (4,1)DFE with (9,3,1)MLP structure, training by BPN learning algorithm...........................66 圖5.19　BER of (4,1)DFE structure, training by LMS,..........67 圖5.20　BER of (4,1)DFE with (9,3,1)MLP structure, training by EAs, population size = (200,200).............................68 圖5.21　BER of (4,1)DFE with (9,3,1)MLP structure, for different learning algorithms................................69 圖5.22　BER of (4,1)DFE with (9,3,1)MLP structure training by EAs with and without fine turning............................70 圖5.23　BER of neuro-fuzzy equalizer-based on linguistic model for different learning algorithms............................71 圖5.24　BER of neuro-fuzzy equalizer-based on linguistic model for different fuzzy rule, population size=(200,200)..........72 圖5.25　BER of neuro-fuzzy equalizer-based on linguistic model for different fuzzy rule, population size=(200,200)..........73 圖5.26　BER of Neuro-Fuzzy Equalizer-based on Linguistic Model with and without fine turning................................74 表2.1　通道之無雜訊座標......................................10

    [1]G.j. Gibson, S. Siu, C.F.N. Cowan, “Multi-layer perceptron structures applied to adaptive equalizers for data communications”, IEEE Proceedings ICASSP Glasgow, Scotland, May 1989, pp. 1183-1186.
    [2]S. Siu, G.j. Gibson, and C.F.N. Cowan, “Decision feedback equalisation using neural network structures and performance comparison with standard architecture”, IEEE Proceedings, Vol.137, PartⅠ, NO.4, pp. 221-225, August, 1990.
    [3]S. Siu, “Non-linear adaptive equalization based on a multi-layer perceptron architecture”, Thesis for doctorate, University of Edinburgh, U.K. 1990.
    [4]G.j. Gibson, S. Siu, and C.F.N. Cowan, “The application of nonlinear structures to the reconstruction of binary signal”, IEEE Trans. On Signal Processing, Vol. 39, No. 8, pp.1877-1884, Aug., 1991.
    [5]Li-Xin Wang, “ADAPTIVE FUZZY SYSTEMS AND CONTROL Design and Stability Analysis”, Prentice-Hall, Inc. 1994
    [6]Ronald R. Yager, Dimitar P. Filev, “Essentials of Fuzzy Modeling and Control”, john Wiley & Sons Inc.
    [7]Qureshi, S.U.H, “Adaptive equalization”, Proc. IEEE. 1985, Vol. 73, No. 9, September 1985.
    [8]S. Siu, and C.F.N. Cowan, “Performance analysis of the norm back propagation algorithm for adaptive equalization” IEE Proc. Part F, Vol.140, No.1, Febr. 1993.
    [9]S. Siu, C.H. Chang, and C.H. Wei, “On the effect of of the norm back propagation algorithm for adaptive equalisation” IEEE Trans. on Circuit and SystemⅡ: Analog and Digital Processing, Sept., 1996, Vol. 42, No. 9, pp.604-607.
    [10]Wael A. Farag, Victor H. Quintana, and Germano Lambert-Torres, “A Genetic-Based Neuro-Fuzzy Approach for Modeling and Control of Dynamical Systems” IEEE Transactions On Neural Networks, Vol. 9, No. 5, September 1998.
    [11]Wael A. Farag, Victor H. Quantana, and Germano Lamber-Torres, “A Genetic-Based Neuro-Fuzzy Approach for Modeling and Control of Dynamical Systems” IEEE Transactions On Neural Networks, Vol. 9, No. 5, September 1998.
    [12]Xiaofeng Qi and Francesco Palmieri, “Theoretical Analysis of Evolutionary Algorithms With an infinite Population size in Continuous Space PartⅠ:Basic Properties of Selection and Mutation” IEEE Transactions On Neural Networks, Vol. 5, No. 1, January 1994.
    [13]Xiaofeng Qi and Francesco Palmieri, “Theoretical Analysis of Evolutionary Algorithms With an infinite Population size in Continuous Space PartⅡ:Analysis of the Diversification Role of Crossover” IEEE Transactions On Neural Networks, Vol. 5, No. 1, January 1994.
    [14]David E. Goldberg, “Genetic Algorithms in Search, Optimization, and Machine Learning”, Addison-Wesley Publishing Company, Inc.
    [15]Bernard Sklar, “DIGITAL COMMUNICATIONS Fundamental and Applications”, Prentice-Hall, Inc. 1988
    [16]Simon Haykin, “Adaptive Filter Theory”, Prentice-Hall, Inc. 1996
    [17]Yoh-Han Pao, “Adaptive Pattern Recognition and Neural Networks”, Addison-Wesley Publishing Company, Inc. 1989.
    [18]Simon Haykin, “COMMUNICATIOM SYSTEM”, John Wiley & Sons Inc.
    [19]Ziemer, Tranter, “PRINCIPLES OF COMMUNICATIONS System, Modulation, and Noise ”, John Wiley & Sons Inc.
    [20]蘇木春,張孝德, “機器學習 類神經網路、模糊系統以及基因演算法則” 全華科技圖書股份有限公司
    [21]林繼洲, “函數連結與模糊適應等化器效能評估”, 元智大學電機工程研究所碩士論文, 1999.
    [22]張吉良, “利用進化演算法在多層感知結構之判別回授等化器”, 中央大學電機工程研究所碩士論文, 2001.
    [23]廖鴻翰, “以基因演算法建構類神經網路模型”, 大葉大學電機工程研究所碩士論文, 2000.
    [24]M. Srinivas and L.M. Patnaik, “Genetic Search: Analysis Using Fitness Moments”, IEEE Transactions on Knowledge and data engineering, Vol. 8, No1, February 1996.
    [25]M. Srinivas and L.M. Patnaik, Fellow, IEEE, “Adaptive Probabilities of Crossover and Mutation in Genetic Algorithms”, IEEE Transactions on System, Man and Cybernetics, Vol. 24, No 4, April 1994.
    [26]林進燈, “模糊類神經網路控制與決策系統”, 國科會專題研究計畫成果報告(交大控工所), 1993

    QR CODE
    :::