程序控制之研究 ─ Fuzzy控制器設計與參數調適□

Abstract

模糊控制(Fuzzy Control)是近代興起的一項控制理論，由於它具有不依賴受控系統(plant)數學模型之特色，僅需根據人的操作經驗、知識與控制法則推論即可決定控制命令的優點，因此受到廣泛的應用。 本篇論文敘述模糊控制理論及模糊控制器原理，而藉比率因子與PID型模糊控制器的關係，說明FLC(模糊邏輯控制器)與傳統PD型控制器、PI控制器的等效性，亦探討FLC穩態誤差之形成與平衡穩態控制輸入(Balanced Steady-state Control Input，BSCI)之關係。本論文提出與BSCI觀念結合之PD+I型FLC，具有響應快速且零穩態誤差之特性；亦提出一種簡單且有效之自動調適法，以系統性能為依據，逐次調整比例因子以達性能要求。 最後，利用此PD+I型FLC運用到液位系統的高度控制，驗證其對一階系統具有良好之控制效能，且抗干擾特性亦佳，對驅動器具有飽和(saturation)限制之系統比傳統PI控制器有較佳之穩態控制。

Fuzzy control is a rather popular control theorem recently. It depended on experiences, knowledge and inferences of operators only rather than mathematical model of the controlled system, and the simplicity of fuzzy control extend its applications widely. This thesis shows the theorem and the principle of fuzzy control and states the equivalents between FLC (Fuzzy Logic Controller) and traditional PD or PI controller. The relations of FLC steady-state error with Balanced Steady-state Control Input (BSCI) were studied. System using the PD+I like FLC and including the BSCI compensation has the advantage of rapid response with zero steady-state error. A simple and effective auto-tuning method is proposed to tune all the parameters step by step according to the performances of the controlled system. Finally, a water tank system is controlled by PD+I like FLC to reach an assigned level. It verifies the theorem and shows that the proposed above method is effective and practical. When compared with the traditional PI controller, its performance is better the later.