Mobile Robot Navigation Using Reinforcement Learning Based on Neural Network with Short Term Memory

被引：0

作者：

Gavrilov, Andrey V. ^{[1
]}

Lenskiy, Artem ^{[2
]}

机构：

[1] Novosibirsk State Tech Univ, Dept Prod Automat Machine Engn, Karl Marx Av 20, Novosibirsk 630092, Russia

[2] Korea Univ Technol & Educ, Sch Elect Elect & Commun Engn, Byeongcheon 330708, Dongnam Cheonan, South Korea

来源：

ADVANCED INTELLIGENT COMPUTING | 2011年 / 6838卷

关键词：

neural networks; mobile robots; reinforcement learning; ARCHITECTURE;

D O I：

暂无

中图分类号：

TP18 [人工智能理论];

学科分类号：

081104 ; 0812 ; 0835 ; 1405 ;

摘要：

In this paper we propose a novel bio-inspired model of a mobile robot navigation system. The novelty of our work consists in combining short term memory and online neural network learning using history of events stored in this memory. The neural network is trained with a modified error back propagation algorithm that utilizes reward and punishment principal while interacting with the environment.

引用

页码：210 / +

页数：3

共 19 条

[1]

Amosov N.M., 1975, 4 INT JOINT C ART IN, V9, P1

[2] DRAMA, a connectionist architecture for control and learning in autonomous robots [J].

Billard, A ;

Hayes, G .

ADAPTIVE BEHAVIOR, 1999, 7 (01) :35-63

[3] A ROBUST LAYERED CONTROL-SYSTEM FOR A MOBILE ROBOT [J].

BROOKS, RA .

IEEE JOURNAL OF ROBOTICS AND AUTOMATION, 1986, 2 (01) :14-23

[4]

Carpenter G.A., 1991, PATTERN RECOGN

[5] DISTRIBUTED REPRESENTATIONS, SIMPLE RECURRENT NETWORKS, AND GRAMMATICAL STRUCTURE [J].

ELMAN, JL .

MACHINE LEARNING, 1991, 7 (2-3) :195-225

[6]

Gavrilov A., 2007, IEEE INT C SMC UK RI

[7]

Gavrilov A.V., 2004, MECHATRONICS AUTOMAT, V8, P30

[8]

Gavrilov A, 2007, LECT NOTES ARTIF INT, V4682, P182

[9]

Gavrilov A, 2006, LECT NOTES COMPUT SC, V3971, P707

[10]

Gavrilov AV, 2009, NESTER 2009: NETWORKED EMBEDDED AND CONTROL SYSTEM TECHNOLOGIES: EUROPEAN AND RUSSIAN R&D COOPERATION, PROCEEDINGS, P94

← 1 2 →