Evaluation of reinforcement learning autonomous navigation systems for a NOMAD 200 mobile robot

被引：0

作者：

Ortiz, M ^{[1
]}

Zufiria, PJ ^{[1
]}

机构：

[1] Univ Politecn Madrid, Dept Matemat Aplicada Tecnol Informac, ETSI Telecomunicac, E-28040 Madrid, Spain

来源：

INTELLIGENT AUTONOMOUS VECHICLES 1998 (IAV'98) | 1998年

关键词：

learning systems; embedded systems; CMAC; radial base function network; autonomous mobile robots;

D O I：

暂无

中图分类号：

TP [自动化技术、计算机技术];

学科分类号：

0812 ;

摘要：

In this paper reinforcement learning is employed to provide autonomous navigation capabilities to a mobile robot NOMAD 200. The system is based on the Actor/Critic architecture in the context of Reinforcement Learning. The context-action function is learned by means of Williams REINFORCE algorithm. Two context coding approaches, CMAC and Radial Basis Functions are compared from the point of view of learning capabilities, resource requirements and plasticity. Results obtained in simulations as well as in experiments with the real robot are presented. Copyright (C) 1998 IFAC.

引用

页码：309 / 314

页数：6

共 11 条

[1]

Barto A G., 1983, IEEE Trans, on Systems, Man, and Cybernetics, V13, P835

[2] Reinforcement learning: A survey [J].

Kaelbling, LP ;

Littman, ML ;

Moore, AW .

JOURNAL OF ARTIFICIAL INTELLIGENCE RESEARCH, 1996, 4 :237-285

[3]

KROSE BJA, 1992, ARTIF NEURAL NETWORK, V2, P619

[4] CMAC-BASED ADAPTIVE CRITIC SELF-LEARNING CONTROL [J].

LIN, CS ;

KIM, H .

IEEE TRANSACTIONS ON NEURAL NETWORKS, 1991, 2 (05) :530-533

[5]

MILLAN JR, 1996, IEEE T SYST MAN CYB

[6]

PRESCOTT AJ, 1993, THESIS SHEFFIELD U

[7]

SANTAMARIA JC, 1996, EXPT REINFORCEMENT L

[8]

SANTHARAM G, 1997, IEEE T SYST MAN CYB

[9]

SINGH SP, 1993, ADV NEURAL INFORMATI

[10]

Sutton RS, 1996, ADV NEUR IN, V8, P1038

← 1 2 →