Dynamic gait pattern generation with reinforcement learning

Mustafa Suphi Erden; Kemal Leblebicioglu

doi:10.3182/20050703-6-CZ-1902.01295

Dynamic gait pattern generation with reinforcement learning

Mustafa Suphi Erden^*, Kemal Leblebicioglu

^*Corresponding author for this work

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

Abstract

This paper presents the gait pattern generation work performed for the sixlegged robot EA308 developed in our laboratory. The aim is to achieve a dynamically developing gait pattern generation structure using reinforcement learning. For the six legged robot a simplified simulative model is constructed. The algorithm constructs a radial basis function neural network (RBFNN) to command proper leg configurations to the simulative robot. The weights of the RBFNN are learned using reinforcement learning. The developed structure succeeded in learning gait patterns compatible with different speeds of the robot.

Original language	English
Title of host publication	IFAC Proceedings Volumes (IFAC-PapersOnline)
Pages	151-156
Number of pages	6
Volume	16
DOIs	https://doi.org/10.3182/20050703-6-CZ-1902.01295
Publication status	Published - Dec 2005
Event	16th Triennial World Congress of International Federation of Automatic Control - Prague, United Kingdom Duration: 3 Jul 2005 → 8 Jul 2005

Conference

Conference	16th Triennial World Congress of International Federation of Automatic Control
Abbreviated title	IFAC 2005
Country/Territory	United Kingdom
City	Prague
Period	3/07/05 → 8/07/05

Keywords

Gait pattern
Radial basis function neural network
Reinforcement learning
Six-legged robot
Walking

Access to Document

10.3182/20050703-6-CZ-1902.01295

Cite this

@inproceedings{fee55ed412e040d39e92fddb1a22619c,

title = "Dynamic gait pattern generation with reinforcement learning",

abstract = "This paper presents the gait pattern generation work performed for the sixlegged robot EA308 developed in our laboratory. The aim is to achieve a dynamically developing gait pattern generation structure using reinforcement learning. For the six legged robot a simplified simulative model is constructed. The algorithm constructs a radial basis function neural network (RBFNN) to command proper leg configurations to the simulative robot. The weights of the RBFNN are learned using reinforcement learning. The developed structure succeeded in learning gait patterns compatible with different speeds of the robot.",

keywords = "Gait pattern, Radial basis function neural network, Reinforcement learning, Six-legged robot, Walking",

author = "Erden, {Mustafa Suphi} and Kemal Leblebicioglu",

year = "2005",

month = dec,

doi = "10.3182/20050703-6-CZ-1902.01295",

language = "English",

isbn = "008045108X",

volume = "16",

pages = "151--156",

booktitle = "IFAC Proceedings Volumes (IFAC-PapersOnline)",

note = "16th Triennial World Congress of International Federation of Automatic Control, IFAC 2005 ; Conference date: 03-07-2005 Through 08-07-2005",

}

TY - GEN

T1 - Dynamic gait pattern generation with reinforcement learning

AU - Erden, Mustafa Suphi

AU - Leblebicioglu, Kemal

PY - 2005/12

Y1 - 2005/12

N2 - This paper presents the gait pattern generation work performed for the sixlegged robot EA308 developed in our laboratory. The aim is to achieve a dynamically developing gait pattern generation structure using reinforcement learning. For the six legged robot a simplified simulative model is constructed. The algorithm constructs a radial basis function neural network (RBFNN) to command proper leg configurations to the simulative robot. The weights of the RBFNN are learned using reinforcement learning. The developed structure succeeded in learning gait patterns compatible with different speeds of the robot.

AB - This paper presents the gait pattern generation work performed for the sixlegged robot EA308 developed in our laboratory. The aim is to achieve a dynamically developing gait pattern generation structure using reinforcement learning. For the six legged robot a simplified simulative model is constructed. The algorithm constructs a radial basis function neural network (RBFNN) to command proper leg configurations to the simulative robot. The weights of the RBFNN are learned using reinforcement learning. The developed structure succeeded in learning gait patterns compatible with different speeds of the robot.

KW - Gait pattern

KW - Radial basis function neural network

KW - Reinforcement learning

KW - Six-legged robot

KW - Walking

U2 - 10.3182/20050703-6-CZ-1902.01295

DO - 10.3182/20050703-6-CZ-1902.01295

M3 - Conference contribution

AN - SCOPUS:79960747404

SN - 008045108X

SN - 9780080451084

VL - 16

SP - 151

EP - 156

BT - IFAC Proceedings Volumes (IFAC-PapersOnline)

T2 - 16th Triennial World Congress of International Federation of Automatic Control

Y2 - 3 July 2005 through 8 July 2005

ER -

Dynamic gait pattern generation with reinforcement learning

Abstract

Conference

Keywords

Access to Document

Fingerprint

Cite this