DocumentCode :
2692819
Title :
A simple and converged structure of addressing technique for CMAC_GBF
Author :
Chiang, Ching-Tsan ; Chiang, Tung-sheng ; Li, Chien-Kuo
Author_Institution :
Dept. of Electr. Eng., Ching Yun Univ., Jung-Li, Taiwan
Volume :
7
fYear :
2004
fDate :
10-13 Oct. 2004
Firstpage :
6097
Abstract :
This paper proposed a simple and converged structure of addressing technique for CMAC_GBF (cerebellar model articulation controller with general basis function). The structure of addressing technique for original CMAC_GBF has two major problems: one is the structure of addressing technique does not have a fixed procedure to find it, the other is the memory size growing too large in the high dimension input space systems. In this paper, the procedure to find the simple structure is developed. The mathematical basis for this simple structure is derived and the learning convergence is guaranteed. Simulations for the simple structure of CMAC_GBF are performed to demonstrate the learning performance and the capability in reducing the memory size of system.
Keywords :
cerebellar model arithmetic computers; learning (artificial intelligence); mathematical analysis; neurocontrollers; storage allocation; CMAC_GBF; addressing technique; cerebellar model articulation controller; converged structure; general basis function; high dimension input space systems; learning convergence; mathematical basis; Control systems; Convergence; Electrical equipment industry; Fuel storage; Geology; Hypercubes; Industrial control; Manipulators; Quantization; Robot control;
fLanguage :
English
Publisher :
ieee
Conference_Titel :
Systems, Man and Cybernetics, 2004 IEEE International Conference on
ISSN :
1062-922X
Print_ISBN :
0-7803-8566-7
Type :
conf
DOI :
10.1109/ICSMC.2004.1401355
Filename :
1401355
Link To Document :
بازگشت