Gain Adaptation of Fuzzy Controller Design

增益調整之模糊控制器設計

陳珍源(Jen-Yang Chen) ； 蔡樸生(Pu-Sheng Tsai)

《中華技術學院學報》 27期 (2003/05) Pp. 83-93

https://doi.org/10.7095/JCIT.200305.0083

摘要

本文針對TSK(Taskagi-Sugeno-Kang) 模糊系統提出一基於線上增益調節之適應性控制器設，其中適應律設計係採直接式架構，以產生適當的控制信號。所提方法之主要優點是：在此適應性控制架構下，不但原TSK模糊系統的推論法則及歸屬函數無須更動外，且僅需調適單一參數。一般而言，當模糊控制器之輸入變數量大時，在未引入專家知識的情況下，很難描述口語化型式的模糊控制法則及其相關之歸屬函數分佈。本研究，首先即在未引入任何關於受控系統之經驗知識下，建構了一個TSK模擬控制器。然後藉由所提之適應律機構對此控制器的輸出作線上增益調節，以獲得令人滿意的閉迴路性能。最後，以一個非線性系統之適應性模糊控制為例，來驗證所提方法之可行性。

關鍵字

模糊控制 ； 模糊集合 ； 歸屬函數 ； 調適性增益 ； 適應性模糊控制

並列摘要

This paper proposes a Takagi-Suegeno-Kang (TSK) fuzzy controller design using scaling gain adaptation. The proposed adaptation law for the TSK fuzzy controller resulting from the direct adaptive approach is used to appropriately determine the control signal of the controller. The primary advantage of the proposed method is that the parameters of the fuzzy controller, including its rules and membership functions; do not need to be changed under control, particularly when only one adapted parameter is used. Usually, it is not easy to characterize the linguistic control rules and the membership function distribution without expert knowledge when the number of input variables of fuzzy controller is large. In this study, we first organize the TSK fuzzy controller without any specific experience information about the controlled system. Then, the proposed adaptation law is used to adapt the controller’s output scaling gain so that the satisfactory system performance of the closed-loop control system can be achieved. Finally, we use the proposed fuzzy controllers to control a nonlinear system, verifying its effectiveness.

並列關鍵字

Fuzzy Control ； Fuzzy Sets ； Membership Functions ； Scaling Gain ； Adaptive Fuzzy Control