Please use this identifier to cite or link to this item: http://hdl.handle.net/11455/2045
標題: 類神經網路在啟發式學習控制系統之應用
The Application of Neural Network on Heuristic Learning Control System
作者: 劉適豪
關鍵字: 類神經;啟發式學習
出版社: 機械工程研究所
摘要: 
本論文中,我們將提出一以類神經網路為基礎之啟發式學習控制
系統,本控制系統之特色乃是以類神經網路取代傳統電腦記憶式之學
習控制器(Arimoto,S,1985)。透過類神經網路之高度學習能力與簡單
之啟發式學習策略,學習複雜之控制規則,使受控系統之輸出響應符
合吾人之要求。此啟發式學習策略可簡單表示如下:
ud=u+(kp.e+kv.△e)
其中kp,kz為任意常數。
電腦模擬結果證明本學習控制器能依經驗迅速獲取對任一未知系
統之控制法則,且有極高之強韌性,而模糊化輸入訊號與動態kp之導
入更大大提高本學習控制器之學習速度與其控制性能。

In this paper, a heuristic learning control algorithm based on the neural-network techniques is proposed. Neural-Network with its high learning ability is used as the learning controller that can learn the complex control rule to an unknow system by means of a very simple heuristic learning rule.
The heuristic learning rule can be expressed as:
ud=u+(kp‧e+kv‧Δ)
kp and kv can be selected arbitrarily.
Computer simulations show that the proposed learning controller can successfully obtain the complex control rules to an unknow system in very short period.
Converging speed and performance of the proposed learning controller are largely improved by the fuzzification of the input signals and dynamical process in selected kp.
URI: http://hdl.handle.net/11455/2045
Appears in Collections:機械工程學系所

Show full item record
 

Google ScholarTM

Check


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