具觀測器之PID基因適應型控制器

A GA-Tuned Adaptive PID Controller with Observer

溫宏康(Hong-Kang Wen)

指導教授 ： 姚立德 ； 王偉彥

國立臺北科技大學/電資學院/電機工程系所/碩士(2008年)

摘要

基因演算法(GA)是一個最佳化的搜尋工具，近二十年來發展的很迅速，在許多的實際應用上也已經很廣泛的被使用來尋找最佳解。基因演算法有全域的搜尋能力，是靠著它隨機方向的搜尋方式，以及移民滅種的手段來跳脫區域解來找到全域最佳解。而本論文使用了簡化型基因演算法(RGA)來做線上PID控制器參數調整。
PID控制器有著簡單的架構，以及不錯的性能，所需要設計的參數也很少，所以很適合來做線上控制。然而，PID控制器也有一些缺點，像是對時變系統或非線性系統的控制性能較差。而且傳統的PID控制器都是透過專家的經驗或使用錯誤嘗試法來調整參數，這樣的做法非常耗費精力與時間，因此結合基因演算法來設計控制器是很好的做法，也越來越多這樣的設計。
適應性控制可以有很好的穩定性與強健性，並且可以有系統地更新來達到近似的功能，所以常常拿來加入控制器中，來使控制器更強健。本論文提出一個以觀測器為基礎的輸出回授直接適應性PID控制器，來控制一個未知的受控廠，使用Lyapunov方法設計出一個適合的適應函數再利用RGA來做線上參數的調整，調整PID的三個參數。並且加入一個監督控制器 來使系統穩定。

關鍵字

基因演算法 ； PID控制器 ； 適應控制 ； 監督控制

並列摘要

Recently, since genetic algorithms have good capabilities of direct random search for global optimization with techniques mutation and extinction, they have drawn significant attention in various fields. In this thesis, we use a reduced form genetic algorithm to tune the parameters of PID controller online.
PID controller, which is simple in structure and have good performance, is suitable to use in online control. However, PID controller have some drawbacks, for examples, in time-varying systems and nonlinear systems. The parameters of PID controller have traditionally been adjusted by expert experience and trial and error, and this takes time and our energy. Therefore, genetic algorithms can help this, and there are more and more design from this way.
Adaptive control of systems has been an active area of research. Some adaptive control schemes for nonlinear systems via feedback linearization have been proposed. In this thesis, we proposed an online GA-Based direct adaptive PID controller . We tuned the parameters of PID controller by using the fitness function designed by the SPR-Lyapunov approach. Finally, we add a supervisory controller in order to guarantee the stability.

並列關鍵字

genetic algorithm ； adaptive control ； PID ； supervisory control.

參考文獻

[1] L. A. Zadeh, “Fuzzy Sets,” Information and Control, vol. 8, pp. 338-353, 1965.

[2] L. A. Zadeh, “The Concept of Linguistic Variable and Its Application to Approximate Reasoning, Parts 1,2,” Inform. Sci., vol. 8, pp. 119-249, 1975; vol. 8, pp. 301-357, 1975.

[3] L. O. Hall, S. Szabo, and A. Kandel, “On the Derivation of Memberships for Fuzzy Sets in Expert Systems,” Inform. Sci., vol. 40, pp. 39-52, 1986.

[5] J. H. Holland, Adaptation in Natural and Artificial System, Ann Arbor, MI: Univ. of Michigan Press, 1975.

[6] Z. Michalewicz, Genetic Algorithms + Data Structures = Evolution Programs, Springer-Verlag, 1992.