臺灣博碩士論文加值系統

雖然傳統PID控制器現在仍廣泛的應用在現代工業界中，這是因為其結構簡單而且具有強韌性，但可惜的是PID無法對時延系統作有效的控制。Shinskey 於1994年提出一個新的PID延遲控制方法
(PID dead-time controller)， 其架構與功能類似於以模式為基控制器(model-based controller)。本文之目的在於針對好幾種不同的程序設計最佳控制器， 使其誤差平方積分最小化(ISE)。 另外，為了避免時間應答出現振盪現象， 我們也選用ITSE做為目標函數設計最佳控制器。 為
了與傳統PID控制器作比較， 我們也作傳統PID最佳控制器的設計， 並與之比較。
控制器的設計第一個要考慮的是控制系統的穩定性， 一定要能確保調諧模式系統的閉環路能穩定操作。 可是如果轉移函數含有超越函數的話在數學上並無解析式解決， 所以無法用代數方法，如Routh法判斷穩定性，在此我們以程式技巧追蹤Nyquist路途來做穩定性判斷
，克服在時延系統中穩定性判定的難題。
本文是應用微分演化法(Differential evolution algorithms)搜尋最佳PID延遲控制器的參數，DEA很成功的應用在最適化這方面的問題， 避免像其它方法因陷入局部最佳而失敗。除了一階與二階時延系統， 在後面章節以熱交換之分佈系統為例來做設計，實例證明PID延遲控制器的控制確實效能良好。

Although the conventional PID controllers are still widely used in modern processes because of their simple structure and performance robustness.However,it is not so effective for the processes with time delays. A new method of PID dead-time control is introduced by Shinskey in 1994,while functionally much similar to model based controller and more effective for lag-dominat processes. The purpose of this thesis is to design optimal PID dead-time controller for different types of processes with time delays such the integral error is minimized. Considering a matter in the oscillations of time response in the systems,we also tune the controller to minimize the performance index ITSE.In order to compare the performance of PID dead-time controller with conventional PID ccontroller,
we also design it at the same time.
Since the stability of system is the major concern in the controller design,it is required that the
designed tuned-model control system will be stable for the closed loop operation. There is no analytical form in math for that the transfer function has transcendental terms
and we can't use the direct methods,like Routh algorithm,to do the satbility test. So it is diffcult to solve the stability problems for the systems with time delays. For this reason,we use the computer programs to trace the Nyquist path so it is easily to obtain the stable parameters.
This thesis is also to investigate the use of differential evolution algorithms(DEA) for optimization in PID dead-time control. DEA have been successfully applied to a variety of
complex optimization problems where other techniques have often failed. To illustrate the proposed optimal PID dead-time controller design,besides first and second order plus time
delays processes,a example of a heat transfer process which is known as distributed system will be worked out. The examples prove that PID dead-time controller is effective.

中文摘要…………………………………………………………………i
英文摘要…………………………………………………………………ii
圖表目錄…………………………………………………………………iii
第一章 緒論…………………………………………………………………1
1.1 研究動機…………………………………………………………………1
1.2 文獻探討…………………………………………………………………2
1.3 PID延遲控制器…………………………………………………………………3
1.4 章節組織…………………………………………………………………4
第二章 性能指標介紹與計算…………………………………………………………………7
2.1 前言…………………………………………………………………7
2.2 ISE之數值計算法…………………………………………………………………8
2.3 變異ISE計算法…………………………………………………………………10
2.4 結論…………………………………………………………………11
第三章 微分演化法…………………………………………………………………12
3.1 簡介…………………………………………………………………12
3.1-2 微分演化法特點…………………………………………………………………13
3.2 演化機制…………………………………………………………………14
3.2-1 建構起始族群…………………………………………………………………15
3.2-2 突變…………………………………………………………………15
3.2-3 交配…………………………………………………………………16
3.2-4 選擇…………………………………………………………………17
3.3 演化流程…………………………………………………………………17
3.4 結論…………………………………………………………………18
第四章 最適模式為基控制器設計實例…………………………………………………………………20
4.1 前言…………………………………………………………………20
4.2 Nyquist穩定判斷法則…………………………………………………………………20
4.3 增益邊線與相位邊限…………………………………………………………………22
4.4 最適控制器參數設計流程…………………………………………………………………22
4.5 一階與二階時延系統控制器設計實例…………………………………………………………………23
4.6 結論…………………………………………………………………27
第五章 分佈系統之控制器最適設計…………………………………………………………………56
5.1 前言…………………………………………………………………56
5.2 問題陳述…………………………………………………………………56
5.3 分佈系統之簡化模式…………………………………………………………………60
5.4 擾動排除之參數設定 60
5.5 結論…………………………………………………………………61
第六章 總結與未來展望…………………………………………………………………69
6.1 總結…………………………………………………………………69
6.2 未來展望…………………………………………………………………69
參考文獻…………………………………………………………………71

石延平著, "程序控制",友寧出版社
蔡文嘉,"分佈系統之最適PID控制器設計,"中正大學化工研究所,87級碩士論文.
D. W. Luse,"A Nyquist-Type Stability Test for Multivariable Distributed Systems,"IEEE Trans. Automatic Control,Vol. AC-9,pp. 107-109
F.G. Shinskey,"Feedback Controllers for the process industries,"
McGraw-Hill,Inc.,1994
F.G. Shinskey,"Process Control Systemsŗd. ed.,
McGraw-Hill,Inc.},1988}
F.G. Shinskey,"PID-deadtime control of distributed processes,"
IFAC Process Control Consultant,pp. 14-18,2000
Hwang, C., Chuang, Y. H., and Shih, Y. P., "Numerical computation of quadratic cost functions for systems with multiple delays,"Control and Computers,Vol. 22, No. 3, pp. 90-95,1994
J. E. Marshall and K. Walton,"A review of Analytical Methods for Control od Time-Delay Systems,"IFAC Control of Distributed Parameter Systems,Los Angeles,California,pp. 485-490,1986
Katsuhiko Ogata,"Modern Control Engineering,"Prentice Hall,1990
Price., K., Storn, R.,"Diferential evolution,"IFAC Control of Distributed Parameter Systems,Los Angeles,California,pp. 485-490,1986
Price., K. V.,"Diferential evolution:a fast and simple numerical optimizer,"Proc. of North American Fuzzy Information Processing Conf.,pp. 524-527,1997
Shih-Lian Cheng and Chiy Hwang,"Designing PID Controllers with A Minimum IAE Criterion by A Differential Evolution Algorithm ,"
IEEE Trans. Automatic Control,Vol. AC-9,pp. 107-109
Sobol,I. M.,"On the Systematic Search in a Hypercube,"
IFAC SIAM J. Numer. Anal.vol. 16,no. 5,pp. 790-793,1979
Yuwana, M and D. E. Seborg,"A New Method for On-line Controller Turning,"AIChE J.,28. 434,1982
Zigler,J.G.,and N.B. Nichols,"Optimum Settings for Automatic Controllers,"Trans.ASME,pp. 759-768,1942