An implementation efficient learning algorithm for adaptive control using associative content addressable memory

Author

Hu, Yendo ; Fellman, Ronald D.

Author_Institution

Dept. of Electr. & Comput. Eng., California Univ., San Diego, La Jolla, CA, USA

Volume

25

Issue

4

fYear

1995

fDate

4/1/1995 12:00:00 AM

Firstpage

704

Lastpage

709

Abstract

Three modifications to the Boxes-ASE/ACE reinforcement learning improves implementation efficiency and performance. A state history queue (SHQ) eliminates computations for temporally insignificant states. A dynamic link table only allocates control memory to states the system traverses. CMAC state association uses previous learning to decrease training time. Simulations show a 4-fold improvement in learning. The SHQ in a hardware implementation of the pole-cart balancer reduces computation time 11-fold

Keywords

adaptive control; computational complexity; content-addressable storage; learning (artificial intelligence); Boxes-ASE/ACE reinforcement learning; CMAC state association; adaptive control; associative content-addressable memory; computation time; dynamic link table; implementation efficient learning algorithm; pole-cart balancer; state history queue; Adaptive control; Algorithm design and analysis; Associative memory; Computational modeling; Control systems; Decoding; Equations; Hardware; History; Learning;

fLanguage

English

Journal_Title

Systems, Man and Cybernetics, IEEE Transactions on

Publisher

ieee

ISSN

0018-9472

Type

jour

DOI

10.1109/21.370204

Filename

370204