Understanding and realization of constrained motion - Human motion analysis and robotic learning approaches

Shigeyuki Hosoe, Yuichi Kobayashi, Mikhail Mikhailovich Svinin

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

This paper presents two approaches toward the understanding and realization of constrained motion of the upper limbs. The first approach deals with the analysis on the constrained human movements under the framework of optimal control. It is shown that the combination of the optimality criteria constructed by muscle force change and hand contact force change can explain the formation of the constrained reaching movements. The second approach comes from robotics. It is illustrated by application of reinforcement learning to a robotic manipulation problem. Here, an assumption on the existence of holonomic constraints can accelerate learning by introducing function approximation techniques within model-based reinforcement learning. The idea of estimation (parameterization) of the constraints can relate two different problems under the common perspective of "learning constrained motion".

Original languageEnglish
Title of host publicationIntelligent Autonomous Systems 9, IAS 2006
Pages30-37
Number of pages8
Publication statusPublished - Dec 1 2006
Event9th International Conference on Intelligent Autonomous Systems, IAS 2006 - Tokyo, Japan
Duration: Mar 7 2006Mar 9 2006

Other

Other9th International Conference on Intelligent Autonomous Systems, IAS 2006
CountryJapan
CityTokyo
Period3/7/063/9/06

Fingerprint

Reinforcement learning
Robotics
Parameterization
Muscle
Motion analysis

All Science Journal Classification (ASJC) codes

  • Computer Science(all)
  • Computational Mechanics
  • Control and Systems Engineering

Cite this

Hosoe, S., Kobayashi, Y., & Svinin, M. M. (2006). Understanding and realization of constrained motion - Human motion analysis and robotic learning approaches. In Intelligent Autonomous Systems 9, IAS 2006 (pp. 30-37)

Understanding and realization of constrained motion - Human motion analysis and robotic learning approaches. / Hosoe, Shigeyuki; Kobayashi, Yuichi; Svinin, Mikhail Mikhailovich.

Intelligent Autonomous Systems 9, IAS 2006. 2006. p. 30-37.

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Hosoe, S, Kobayashi, Y & Svinin, MM 2006, Understanding and realization of constrained motion - Human motion analysis and robotic learning approaches. in Intelligent Autonomous Systems 9, IAS 2006. pp. 30-37, 9th International Conference on Intelligent Autonomous Systems, IAS 2006, Tokyo, Japan, 3/7/06.
Hosoe S, Kobayashi Y, Svinin MM. Understanding and realization of constrained motion - Human motion analysis and robotic learning approaches. In Intelligent Autonomous Systems 9, IAS 2006. 2006. p. 30-37
Hosoe, Shigeyuki ; Kobayashi, Yuichi ; Svinin, Mikhail Mikhailovich. / Understanding and realization of constrained motion - Human motion analysis and robotic learning approaches. Intelligent Autonomous Systems 9, IAS 2006. 2006. pp. 30-37
@inproceedings{5f44f192905c466cb9a08c5f8c1618ca,
title = "Understanding and realization of constrained motion - Human motion analysis and robotic learning approaches",
abstract = "This paper presents two approaches toward the understanding and realization of constrained motion of the upper limbs. The first approach deals with the analysis on the constrained human movements under the framework of optimal control. It is shown that the combination of the optimality criteria constructed by muscle force change and hand contact force change can explain the formation of the constrained reaching movements. The second approach comes from robotics. It is illustrated by application of reinforcement learning to a robotic manipulation problem. Here, an assumption on the existence of holonomic constraints can accelerate learning by introducing function approximation techniques within model-based reinforcement learning. The idea of estimation (parameterization) of the constraints can relate two different problems under the common perspective of {"}learning constrained motion{"}.",
author = "Shigeyuki Hosoe and Yuichi Kobayashi and Svinin, {Mikhail Mikhailovich}",
year = "2006",
month = "12",
day = "1",
language = "English",
isbn = "1586035959",
pages = "30--37",
booktitle = "Intelligent Autonomous Systems 9, IAS 2006",

}

TY - GEN

T1 - Understanding and realization of constrained motion - Human motion analysis and robotic learning approaches

AU - Hosoe, Shigeyuki

AU - Kobayashi, Yuichi

AU - Svinin, Mikhail Mikhailovich

PY - 2006/12/1

Y1 - 2006/12/1

N2 - This paper presents two approaches toward the understanding and realization of constrained motion of the upper limbs. The first approach deals with the analysis on the constrained human movements under the framework of optimal control. It is shown that the combination of the optimality criteria constructed by muscle force change and hand contact force change can explain the formation of the constrained reaching movements. The second approach comes from robotics. It is illustrated by application of reinforcement learning to a robotic manipulation problem. Here, an assumption on the existence of holonomic constraints can accelerate learning by introducing function approximation techniques within model-based reinforcement learning. The idea of estimation (parameterization) of the constraints can relate two different problems under the common perspective of "learning constrained motion".

AB - This paper presents two approaches toward the understanding and realization of constrained motion of the upper limbs. The first approach deals with the analysis on the constrained human movements under the framework of optimal control. It is shown that the combination of the optimality criteria constructed by muscle force change and hand contact force change can explain the formation of the constrained reaching movements. The second approach comes from robotics. It is illustrated by application of reinforcement learning to a robotic manipulation problem. Here, an assumption on the existence of holonomic constraints can accelerate learning by introducing function approximation techniques within model-based reinforcement learning. The idea of estimation (parameterization) of the constraints can relate two different problems under the common perspective of "learning constrained motion".

UR - http://www.scopus.com/inward/record.url?scp=84871854050&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84871854050&partnerID=8YFLogxK

M3 - Conference contribution

AN - SCOPUS:84871854050

SN - 1586035959

SN - 9781586035952

SP - 30

EP - 37

BT - Intelligent Autonomous Systems 9, IAS 2006

ER -