Predicting possible conflicts in hierarchical planning for multi-agent systems

Toshiharu Sugawara, Satoshi Kurihara, Toshio Hirotsu, Kensuke Fukuda, Toshihiro Takada

Research output: Contribution to conferencePaper

1 Citation (Scopus)

Abstract

This paper proposes a learning method to select the most appropriate abstract plans during hierarchical planning in the context of multi-agent systems (MAS). In hierarchical planning, a plan is first created at the most abstract level, and is then refined to a more concrete plan, level by level. Thus, selecting an appropriate plan at the abstract level is very important because the selected plan restricts the scope of lower concrete-level plans. This restriction can enable agents to create plans efficiently, but if all the plans under the selected plan contain serious and difficult-to-resolve conflicts with other agents' plans, the resulting plan does not work well or is of low quality. We propose a method in which, from the conflict pattern among agents' plans, an agent learns which abstract plans will cause conflicts with less probability or which conflicts are easy to resolve, thus inducing probabilistically higher-utility concrete plans after conflict resolution. We also show some experimental results to evaluate our method, with the results suggesting structures of resources where tasks are executed.

Original languageEnglish
Pages939-955
Number of pages17
Publication statusPublished - 2005 Dec 1
Externally publishedYes
Event4th International Conference on Autonomous Agents and Multi agent Systems, AAMAS 05 - Utrecht, Netherlands
Duration: 2005 Jul 252005 Jul 29

Other

Other4th International Conference on Autonomous Agents and Multi agent Systems, AAMAS 05
CountryNetherlands
CityUtrecht
Period05/7/2505/7/29

Fingerprint

Multi agent systems
Planning
Concretes

Keywords

  • Cooperation
  • Coordination
  • Learning
  • Planning

ASJC Scopus subject areas

  • Engineering(all)

Cite this

Sugawara, T., Kurihara, S., Hirotsu, T., Fukuda, K., & Takada, T. (2005). Predicting possible conflicts in hierarchical planning for multi-agent systems. 939-955. Paper presented at 4th International Conference on Autonomous Agents and Multi agent Systems, AAMAS 05, Utrecht, Netherlands.

Predicting possible conflicts in hierarchical planning for multi-agent systems. / Sugawara, Toshiharu; Kurihara, Satoshi; Hirotsu, Toshio; Fukuda, Kensuke; Takada, Toshihiro.

2005. 939-955 Paper presented at 4th International Conference on Autonomous Agents and Multi agent Systems, AAMAS 05, Utrecht, Netherlands.

Research output: Contribution to conferencePaper

Sugawara, T, Kurihara, S, Hirotsu, T, Fukuda, K & Takada, T 2005, 'Predicting possible conflicts in hierarchical planning for multi-agent systems' Paper presented at 4th International Conference on Autonomous Agents and Multi agent Systems, AAMAS 05, Utrecht, Netherlands, 05/7/25 - 05/7/29, pp. 939-955.
Sugawara T, Kurihara S, Hirotsu T, Fukuda K, Takada T. Predicting possible conflicts in hierarchical planning for multi-agent systems. 2005. Paper presented at 4th International Conference on Autonomous Agents and Multi agent Systems, AAMAS 05, Utrecht, Netherlands.
Sugawara, Toshiharu ; Kurihara, Satoshi ; Hirotsu, Toshio ; Fukuda, Kensuke ; Takada, Toshihiro. / Predicting possible conflicts in hierarchical planning for multi-agent systems. Paper presented at 4th International Conference on Autonomous Agents and Multi agent Systems, AAMAS 05, Utrecht, Netherlands.17 p.
@conference{cda130fa3ce04140aac30a1d453016ea,
title = "Predicting possible conflicts in hierarchical planning for multi-agent systems",
abstract = "This paper proposes a learning method to select the most appropriate abstract plans during hierarchical planning in the context of multi-agent systems (MAS). In hierarchical planning, a plan is first created at the most abstract level, and is then refined to a more concrete plan, level by level. Thus, selecting an appropriate plan at the abstract level is very important because the selected plan restricts the scope of lower concrete-level plans. This restriction can enable agents to create plans efficiently, but if all the plans under the selected plan contain serious and difficult-to-resolve conflicts with other agents' plans, the resulting plan does not work well or is of low quality. We propose a method in which, from the conflict pattern among agents' plans, an agent learns which abstract plans will cause conflicts with less probability or which conflicts are easy to resolve, thus inducing probabilistically higher-utility concrete plans after conflict resolution. We also show some experimental results to evaluate our method, with the results suggesting structures of resources where tasks are executed.",
keywords = "Cooperation, Coordination, Learning, Planning",
author = "Toshiharu Sugawara and Satoshi Kurihara and Toshio Hirotsu and Kensuke Fukuda and Toshihiro Takada",
year = "2005",
month = "12",
day = "1",
language = "English",
pages = "939--955",
note = "4th International Conference on Autonomous Agents and Multi agent Systems, AAMAS 05 ; Conference date: 25-07-2005 Through 29-07-2005",

}

TY - CONF

T1 - Predicting possible conflicts in hierarchical planning for multi-agent systems

AU - Sugawara, Toshiharu

AU - Kurihara, Satoshi

AU - Hirotsu, Toshio

AU - Fukuda, Kensuke

AU - Takada, Toshihiro

PY - 2005/12/1

Y1 - 2005/12/1

N2 - This paper proposes a learning method to select the most appropriate abstract plans during hierarchical planning in the context of multi-agent systems (MAS). In hierarchical planning, a plan is first created at the most abstract level, and is then refined to a more concrete plan, level by level. Thus, selecting an appropriate plan at the abstract level is very important because the selected plan restricts the scope of lower concrete-level plans. This restriction can enable agents to create plans efficiently, but if all the plans under the selected plan contain serious and difficult-to-resolve conflicts with other agents' plans, the resulting plan does not work well or is of low quality. We propose a method in which, from the conflict pattern among agents' plans, an agent learns which abstract plans will cause conflicts with less probability or which conflicts are easy to resolve, thus inducing probabilistically higher-utility concrete plans after conflict resolution. We also show some experimental results to evaluate our method, with the results suggesting structures of resources where tasks are executed.

AB - This paper proposes a learning method to select the most appropriate abstract plans during hierarchical planning in the context of multi-agent systems (MAS). In hierarchical planning, a plan is first created at the most abstract level, and is then refined to a more concrete plan, level by level. Thus, selecting an appropriate plan at the abstract level is very important because the selected plan restricts the scope of lower concrete-level plans. This restriction can enable agents to create plans efficiently, but if all the plans under the selected plan contain serious and difficult-to-resolve conflicts with other agents' plans, the resulting plan does not work well or is of low quality. We propose a method in which, from the conflict pattern among agents' plans, an agent learns which abstract plans will cause conflicts with less probability or which conflicts are easy to resolve, thus inducing probabilistically higher-utility concrete plans after conflict resolution. We also show some experimental results to evaluate our method, with the results suggesting structures of resources where tasks are executed.

KW - Cooperation

KW - Coordination

KW - Learning

KW - Planning

UR - http://www.scopus.com/inward/record.url?scp=33644801546&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=33644801546&partnerID=8YFLogxK

M3 - Paper

AN - SCOPUS:33644801546

SP - 939

EP - 955

ER -