Please use this identifier to cite or link to this item: http://hdl.handle.net/11455/1762
標題: 可調式模糊控制於多輸入多輸出系統之應用
The Application of Neural-fuzzy to MIMO System
作者: 陳述豪
Chen, chu-hau
關鍵字: 可調式模糊控制
Neural-Fuzzy
多輸入多輸出
MIMO
出版社: 機械工程學系
摘要: 本文提出一複合可調式模糊控制法,使內含雙層質量振動之多輸入多 輸出系統,在兩質量間有耦合之干擾狀況下,仍能達到位置控制的目的, 進一步要求系統在受到不規則之干擾訊號時,依然可以到達我們所期望的 位置。 複合之可調式模糊控制器,是由兩個可調式模糊控制器組合而 成,多增加的一個可調式模糊控制器是用來解決系統耦合的問題;因為可 調式模糊控制器具有學習的功能,所以利用線上學習求得最佳的控制參數 ,達到我們所希望的控制目標。 當系統受到不規則的干擾誤差訊號, 會有嚴重的振動,而且有向上或向下之偏移的問題;本文將利用複合可調 式模糊法來抑制系統的振盪與偏移,使系統可以保持穩定。
In this thesis, a compound neural-fuzzy controller is proposed to control a multi-input/muliti-output(MIMO) vibrating system with a two-degree-of-freedom-spring-lumped mass. The objective of the controller is to eliminate coupling effect and irregular disturbance input so the desired position can be reached. The compound neural-fuzzy controller is composed of two single neural-fuzzy controller. The extra neural-fuzzy controller is designed to eliminate coupling effect of the system. Since the neural-fuzzy controller has the learning ability, the parameters of the neural-fuzzy controller can be tuned on-line to achieve the desired position. If the system has irregular disturbance input, it will vibrate seriously and has some upward and downward move. This thesis will use the compound neural-fuzzy controller to suppress the vibration, erase the move, and keep the system stable.
URI: http://hdl.handle.net/11455/1762
Appears in Collections:機械工程學系所

文件中的檔案:

取得全文請前往華藝線上圖書館



Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.