|Table of Contents|

Improvement of Speedy Algorithm of CMAC for Gauss Basis Functions and Its Application


Research Field:
Publishing date:


Improvement of Speedy Algorithm of CMAC for Gauss Basis Functions and Its Application
QI Hai-longLI Xiu-juan
College of Automation,Nanjing University of Aeronautics and Astronautics, Nanjing 210016,China
CMAC Gauss basis funct ion genet ic algorithm learning rate servo systems
The speedy algorithm of Gauss Basis CMAC( Cerebellar Model Articulat ion Controleer) is improved. Inview of its problem of learning rate. s selection, an optimizat ion of learning rate based on Genet icAlgorithm is presented. This method makes the learning rate. s select ion be optimal. The feasibility of this method is discussed. A tactic of separat ing the process of the parameter opt imization from the pract ical cont rol process is presented. The simulat ion results in a servo system show that this improved method can avoid the uncertainty of learning rate selected by experience and improve the speed of convergence of the CMAC.


[ 1] Albus J S. Data stor ag e in the cerebellar model articulation controller( CMAC) [ J] . Journal of Dynamic Systems, Measurement and Control, Transactions of the ASME, 1975, 97( 3) : 228- 233.
[ 2] Broomhead D S, Lowe D. Multivariable functional interpolat ion and adaptive networks [ J] . Complex Systems, 1998, 2( 4) : 312- 355.
[ 3] Lee S, Kil R M. A Gaussian potential function netw ork w ith hier ar chically sel-f organizing learning [ J ] . Neur al Netw orks, 1991, 4( 7) : 207- 224.
[ 4] Liu Jun, Tang Liang , Xu Xiaoming . Speedy algor ithm of CMAC model for Gauss basis functions [ J] . Journal of Shang hai Jiaotong University, 1998, 32( 8) : 63- 65.[ 5] 孙增圻. 智能控制理论与技术[ M] . 北京: 清华大学出版社, 1997.
[ 6] Holland J H. Adaptive in nature and ar tificial systems [M ] . Michig an: The Univ ersity of Michigan Press, 1975.
[ 7] 王小平, 曹立明. 遗传算法) ) ) 理论、应用与软件实现 [ M] . 西安: 西安交通大学出版社, 2003.
[ 8] 刘金琨. 先进PID 控制及其MATLAB 仿真[M] . 北京: 电子工业出版社, 2003.


Last Update: 2013-05-23