臺灣博碩士論文加值系統

許多光碟機中用來控制光學讀寫頭的控制器，如PID控制器，相位領先-落後補償器，大部分都是由工程師用離線(off-line)的方式調整控制器參數，來改變讀寫頭的動態響應，但是重複利用離線的方式來取得參數，不但浪費時間也浪費成本。因此本文利用MATLAB程式軟體發展出可以線上自動調變控制器參數的智慧型的機構來解決此問題。市售主要的光學讀寫頭致動器只有聚焦（Focusing）、循軌（Tracking）兩個動態方向，並由四根弦(Wire Spring)支撐的整個機構體(Bobbin),並經由永久磁鐵產生磁場進而驅動光學讀寫頭.而本文裡所運用的是四弦型三軸光學讀寫頭，如Fig.1所示,。其中最主要的差別在於多出來的第三軸傾角（Tilting）的控制線圈，用來控制第三軸 方向,並藉由第三軸傾斜的修正，來增加讀寫頭的讀取資料時速度以及精密度，以因應Blu-ray、AOD等藍光雷射光碟機之高密度光資料存取設備的要求。計畫目標將採用業界使用雙級相位領先控制器為基本架構，再加入智慧型模糊參數自動調變系統。此控制器將以線上調變之方式自動尋找最佳之控制器參數，以適應每顆四弦型光學讀寫頭不同之動態特性差異。此動態特性差異為無可避免之製造誤差所造成。為了克服動態差異，在傳統之CDs和DVDs設計中可藉由提高控制器強健性及頻寬以克服外界振動，進而達成定軌目標，但在面對下世代光資料存取設備如HD-DVD、藍光雷射及近場光碟機具有較小之軌距之情況下，所設計之強健控制器往往需過大控制輸出或過高頻寬來克服外界振動，以致無法實現。所以本文將設計及實現一智慧型模糊參數自動調變控制器以克服讀寫頭動態差異。

Parameters of the controllers for optical pickup control in optical disc drives, such as for PID controllers or phase lead-lag compensators, are in common practice tuned off-line by engineers to adapt to various pickup dynamics. The process of off-line tuning is time-consuming and not economic. To remedy the aforementioned problem, the control parameters are automatically tuned on-line in this study by MATLB within some intelligent frameworks. A intelligent tuning methods are developed and employed. The usage of a fuzzy logic controller (FLC) to change the parameter in the double-lead compensator to tune the loop gain parameter in the double-lead compensator. And the design parameters of them are searched by genetic algorithm (GA) for attaining the best performance. Simulations is conducted and the results show that better controller performance can be achieved by the intelligent controllers as compared to the classical design methods. Furthermore, the fuzzy double-lead controller owns fast convergent speed and low overshoot.
One of the main characteristics of GA algorithm is to prevent the local optimization problem. It can be easily designed without an expert’s knowledge base. So how to application of GA algorithm is become an important research field recently. The main contribution of this thesis is that we use GA algorithm to look for optimal parameter of fuzzy double-lead controller and then achieve the better simulation result. It is shown by computer simulation that GA algorithm can find better fitness value with faster convergent rate. The control performances of the fuzzy double-lead controller are greatly improved especially when compared with fuzzy double-lead controller without GA algorithm.

Abstract...................................1
Figure Captions............................5
Table Titles...............................7
Nomenclature...............................8
1. Introduction...........................11
2. Mathematical Modeling..................13
2.1 Dynamic Model.........................13
2.2 System Identification.................16
3. Controller Design......................18
3.1 Double-Lead Compensator...............18
3.2 Auto-Tuning Algorithm.................23
3.2.1 Fuzzy Logic controller..............23
3.2.1.1 Design Inputs and Outputs of FLC..24
3.2.1.2 Fuzzification.....................24
3.2.1.3 Rule Table ........................25
3.2.1.4 Defuzzification...................26
3.2.3 Genetic Algorithm...................27
3.2.3.1 Genetic Algorithm for Fuzzy Double-Lead Controller................................27

4. Numerical Results and Discussion.......30
4.1 Numerical Results.........30
4.2 Discussion................31

5. Conclusions and Future Works...........32
5.1 Conclusions.................32
5.2 Future Works................33
References................................33
Figures...................................35
Fig.1 Construction of 3-Axis Optical Pickups...35
Fig.2 The dynamic model of the bobbin..........36
Fig.3 Moment generation of the four-wire-type optical pickup...37
Fig.4 Experiment devices for system identification.....38
Fig.5 Frequency responses of the real model and identified model....39
Fig.6 The system block diagram....40
Fig.7 The uncompensated system...41
Fig.8 The system within the first lead compensator.....42
Fig.9 The system within the double-lead compensator....43
Fig.10 The structure of the fuzzy double-lead controller....44
Fig.11 The s-plane of the phase lead compensator....45
Fig.12 The triangular membership functions....46
Fig.13 The typical step response....47
Fig.14 A flow chart of a genetic algorithm....48
Fig.15 Definition of triangular membership functions....49
Fig.16 (a) Fitness value of fuzzy double-lead controller; (b) Integral of absolute error of fuzzy double-lead controller....50
Fig.17 The Membership functions (a) The error; (b) Increment of the error; (c) The parameter “a”....51
Fig.18 The step response of the double-lead compensator....52
Fig 19.Response of double-lead controller simulate result(step time 0.02 sec)....53
Fig 20.Response of double-lead controller with fuzzy simulate result(step time 0.02 sec)....54
Fig 21.Response of fuzzy double-lead controller with GA simulate result(step time 0.02 sec)...55

Tables....................................56
Table1 Rule table.....56
Table 2 The parameters of genetic algorithm....57

簡 歷...................................58

[1]M. Akgül and Ö. Morgül, “Fuzzy Controller Design for Parametric Controllers,” Proceedings of the 1997 IEEE International Symposium on Intelligent Control, pp. 67-62, 16-18 July, 1997.
[2]H. P. Hong, S. J. Park, S. J. Han, K. Y. Cho, Y. C. Lim, J. K. Park and T. G. Kim, “A Design of Auto-tuning PID Controller Using Fuzzy Logic,” Proceedings of the 1992 International Conference on Industrial Electronics, vol. 2, pp. 971-976, 9-13 Nov., 1992.
[3]A. Visioli, “Fuzzy Logic Based Set-Point Weight Tuning of PID Controllers,” IEEE Transactions on Systems, Man and Cybernetics, Part A, vol. 29, Issue 6, pp. 587-592, Nov., 1999.
[4]Y. P. Hsu and C. C. Tsai, “Autotuning for Fuzzy-PI Control Using Genetic Algorithm,” Proceedings of the 1996 IEEE IECON 22nd International Conference on Industrial Electronics, vol. 1, pp. 602-607, 5-10 Aug., 1996.
[5]W. R. Hwang and W. E. Thompson, “Design of Intelligent Fuzzy Logic Controllers Using Genetic Algorithms,” Proceedings of the Third IEEE Conference on Fuzzy Systems, vol. 2, pp. 1383-1388, 26-29 June, 1994.
[6]L. Jia and J. Jingping, “The Model Reference Adaptive Control Based on the Genetic Algorithm,” IEEE International Conference on Neural Networks, vol. 2, pp. 783-787, 9-12 June, 1997.
[7]L. M. L. Porter II and K. M. Passino, “Genetic Model Reference Adaptive Control,” Proceedings of the 1994 IEEE International Symposium on Intelligent Control, pp. 219-224, 16-18 Aug., 1994.
[8]P. C-.P. Chao, J. S. Huang and C. L. Lai, “Nonlinear Dynamic Analysis and Actuation Strategy for a Three-DOF Four-wire Type Optical Pickup,” Sensor & Actuators: A. Physical, pp. 171-182, 2003.
[9]江衍科，「改良型基因演算模糊控制器之設計與運用」，私立華梵大學機電工程研究所碩士論文，民國92年。
[10]周鵬程，「遺傳演算法原理與應用—活用MATLAB」，台北：全華科技圖書股份有限公司，民
國90年。
[11]李繼瑜, 「光學讀寫頭之自動調變演算法設計與實作驗證」，私立中原大學機械工程研究所碩士論文，民國94年。