Improvement of learning surface by exploiting the interplay between control and mechanical dynamics

Takaomi Hasegawa; Toshihiro Kawakatsu; Akio Ishiguro

Improvement of learning surface by exploiting the interplay between control and mechanical dynamics

Takaomi Hasegawa, Toshihiro Kawakatsu, Akio Ishiguro

SCI - Physics

Research output: Contribution to conference › Paper › peer-review

Abstract

This study is intended to deal with the interdependency between control and body systems, and to discuss the "relationship as it should be" between these two systems. To this end, a decentralized control of a multi-legged robot is employed as a practical example. The result derived indicates that the convergence of decentralized gait control can be significantly ameliorated by modifying the interaction between the control and body dynamics. We also discuss a property expected to emerge under the "well-balanced coupling" particularly from the viewpoint of learning, by borrowing the idea from the "protein folding problem".

Original language	English
Pages	2931-2935
Number of pages	5
Publication status	Published - 2004 Dec 1
Event	SICE Annual Conference 2004 - Sapporo, Japan Duration: 2004 Aug 4 → 2004 Aug 6

Other

Other	SICE Annual Conference 2004
Country/Territory	Japan
City	Sapporo
Period	04/8/4 → 04/8/6

Keywords

Emergence
Learning Surface
Protein Folding
Well-balanced Design

ASJC Scopus subject areas

Control and Systems Engineering
Computer Science Applications
Electrical and Electronic Engineering

Cite this

@conference{a153e36763784ecfa08a41765c79467c,

title = "Improvement of learning surface by exploiting the interplay between control and mechanical dynamics",

abstract = "This study is intended to deal with the interdependency between control and body systems, and to discuss the {"}relationship as it should be{"} between these two systems. To this end, a decentralized control of a multi-legged robot is employed as a practical example. The result derived indicates that the convergence of decentralized gait control can be significantly ameliorated by modifying the interaction between the control and body dynamics. We also discuss a property expected to emerge under the {"}well-balanced coupling{"} particularly from the viewpoint of learning, by borrowing the idea from the {"}protein folding problem{"}.",

keywords = "Emergence, Learning Surface, Protein Folding, Well-balanced Design",

author = "Takaomi Hasegawa and Toshihiro Kawakatsu and Akio Ishiguro",

year = "2004",

month = dec,

day = "1",

language = "English",

pages = "2931--2935",

note = "SICE Annual Conference 2004 ; Conference date: 04-08-2004 Through 06-08-2004",

}

TY - CONF

T1 - Improvement of learning surface by exploiting the interplay between control and mechanical dynamics

AU - Hasegawa, Takaomi

AU - Kawakatsu, Toshihiro

AU - Ishiguro, Akio

PY - 2004/12/1

Y1 - 2004/12/1

N2 - This study is intended to deal with the interdependency between control and body systems, and to discuss the "relationship as it should be" between these two systems. To this end, a decentralized control of a multi-legged robot is employed as a practical example. The result derived indicates that the convergence of decentralized gait control can be significantly ameliorated by modifying the interaction between the control and body dynamics. We also discuss a property expected to emerge under the "well-balanced coupling" particularly from the viewpoint of learning, by borrowing the idea from the "protein folding problem".

AB - This study is intended to deal with the interdependency between control and body systems, and to discuss the "relationship as it should be" between these two systems. To this end, a decentralized control of a multi-legged robot is employed as a practical example. The result derived indicates that the convergence of decentralized gait control can be significantly ameliorated by modifying the interaction between the control and body dynamics. We also discuss a property expected to emerge under the "well-balanced coupling" particularly from the viewpoint of learning, by borrowing the idea from the "protein folding problem".

KW - Emergence

KW - Learning Surface

KW - Protein Folding

KW - Well-balanced Design

UR - http://www.scopus.com/inward/record.url?scp=12744280761&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=12744280761&partnerID=8YFLogxK

M3 - Paper

AN - SCOPUS:12744280761

SP - 2931

EP - 2935

T2 - SICE Annual Conference 2004

Y2 - 4 August 2004 through 6 August 2004

ER -

Improvement of learning surface by exploiting the interplay between control and mechanical dynamics

Abstract

Other

Keywords

ASJC Scopus subject areas

Other files and links

Fingerprint

Cite this