Strategies for Energy-Aware Multi-agent Continuous Cooperative Patrolling Problems Subject to Requirements

Lingying Wu; Toshiharu Sugawara

doi:10.1007/978-3-030-33792-6_44

Strategies for Energy-Aware Multi-agent Continuous Cooperative Patrolling Problems Subject to Requirements

Lingying Wu^*, Toshiharu Sugawara

^*この研究の対応する著者

基幹理工学部

研究成果: Conference contribution

抄録

This paper proposes a method of autonomous strategy learning for multiple cooperative agents integrated with a series of behavioral strategies aiming at reduction of energy cost on the premise of satisfying quality requirements in continuous patrolling problems. We improved our algorithm of requirement estimation to avoid concentration of agents since they are given the knowledge of the work environment in advance. The experimental results show that our proposal enables the agents to learn to select appropriate behavioral planning strategies according to performance efficiency and energy cost, and to individually estimate whether the given requirement is reached and modify their action plans to save energy. Furthermore, agents with the new requirement estimation method could achieve fair patrolling by introducing local observations.

本文言語	English
ホスト出版物のタイトル	PRIMA 2019
ホスト出版物のサブタイトル	Principles and Practice of Multi-Agent Systems - 22nd International Conference, Proceedings
編集者	Matteo Baldoni, Mehdi Dastani, Beishui Liao, Yuko Sakurai, Rym Zalila Wenkstern
出版社	Springer
ページ	585-593
ページ数	9
ISBN（印刷版）	9783030337919
DOI	https://doi.org/10.1007/978-3-030-33792-6_44
出版ステータス	Published - 2019
イベント	22nd International Conference on Principles and Practice of Multi-Agent Systems, PRIMA 2019 - Turin, Italy 継続期間: 2019 10月 28 → 2019 10月 31

出版物シリーズ

名前	Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
巻	11873 LNAI
ISSN（印刷版）	0302-9743
ISSN（電子版）	1611-3349

Conference

Conference	22nd International Conference on Principles and Practice of Multi-Agent Systems, PRIMA 2019
国/地域	Italy
City	Turin
Period	19/10/28 → 19/10/31

ASJC Scopus subject areas

理論的コンピュータサイエンス
コンピュータサイエンス（全般）

UN SDG

この成果は、次の持続可能な開発目標に貢献しています

文献へのアクセス

10.1007/978-3-030-33792-6_44

他のファイルとリンク

引用スタイル

Wu, L., & Sugawara, T. (2019). Strategies for Energy-Aware Multi-agent Continuous Cooperative Patrolling Problems Subject to Requirements. In M. Baldoni, M. Dastani, B. Liao, Y. Sakurai, & R. Zalila Wenkstern (Eds.), PRIMA 2019: Principles and Practice of Multi-Agent Systems - 22nd International Conference, Proceedings (pp. 585-593). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 11873 LNAI). Springer. https://doi.org/10.1007/978-3-030-33792-6_44

Strategies for Energy-Aware Multi-agent Continuous Cooperative Patrolling Problems Subject to Requirements. / Wu, Lingying; Sugawara, Toshiharu.
PRIMA 2019: Principles and Practice of Multi-Agent Systems - 22nd International Conference, Proceedings. ed. / Matteo Baldoni; Mehdi Dastani; Beishui Liao; Yuko Sakurai; Rym Zalila Wenkstern. Springer, 2019. p. 585-593 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 11873 LNAI).

研究成果: Conference contribution

Wu, L & Sugawara, T 2019, Strategies for Energy-Aware Multi-agent Continuous Cooperative Patrolling Problems Subject to Requirements. in M Baldoni, M Dastani, B Liao, Y Sakurai & R Zalila Wenkstern (eds), PRIMA 2019: Principles and Practice of Multi-Agent Systems - 22nd International Conference, Proceedings. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 11873 LNAI, Springer, pp. 585-593, 22nd International Conference on Principles and Practice of Multi-Agent Systems, PRIMA 2019, Turin, Italy, 19/10/28. https://doi.org/10.1007/978-3-030-33792-6_44

Wu L, Sugawara T. Strategies for Energy-Aware Multi-agent Continuous Cooperative Patrolling Problems Subject to Requirements. In Baldoni M, Dastani M, Liao B, Sakurai Y, Zalila Wenkstern R, editors, PRIMA 2019: Principles and Practice of Multi-Agent Systems - 22nd International Conference, Proceedings. Springer. 2019. p. 585-593. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). doi: 10.1007/978-3-030-33792-6_44

Wu, Lingying ; Sugawara, Toshiharu. / Strategies for Energy-Aware Multi-agent Continuous Cooperative Patrolling Problems Subject to Requirements. PRIMA 2019: Principles and Practice of Multi-Agent Systems - 22nd International Conference, Proceedings. editor / Matteo Baldoni ; Mehdi Dastani ; Beishui Liao ; Yuko Sakurai ; Rym Zalila Wenkstern. Springer, 2019. pp. 585-593 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).

@inproceedings{135177e35d284831a0dd8f81a0502076,

title = "Strategies for Energy-Aware Multi-agent Continuous Cooperative Patrolling Problems Subject to Requirements",

abstract = "This paper proposes a method of autonomous strategy learning for multiple cooperative agents integrated with a series of behavioral strategies aiming at reduction of energy cost on the premise of satisfying quality requirements in continuous patrolling problems. We improved our algorithm of requirement estimation to avoid concentration of agents since they are given the knowledge of the work environment in advance. The experimental results show that our proposal enables the agents to learn to select appropriate behavioral planning strategies according to performance efficiency and energy cost, and to individually estimate whether the given requirement is reached and modify their action plans to save energy. Furthermore, agents with the new requirement estimation method could achieve fair patrolling by introducing local observations.",

keywords = "Continuous patrolling, Cooperation, Energy efficiency, Learning, Multi-agent systems",

author = "Lingying Wu and Toshiharu Sugawara",

note = "Funding Information: This work was partly supported by JSPS KAKENHI (17KT0044). Publisher Copyright: {\textcopyright} 2019, Springer Nature Switzerland AG.; 22nd International Conference on Principles and Practice of Multi-Agent Systems, PRIMA 2019 ; Conference date: 28-10-2019 Through 31-10-2019",

year = "2019",

doi = "10.1007/978-3-030-33792-6_44",

language = "English",

isbn = "9783030337919",

series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",

publisher = "Springer",

pages = "585--593",

editor = "Matteo Baldoni and Mehdi Dastani and Beishui Liao and Yuko Sakurai and {Zalila Wenkstern}, Rym",

booktitle = "PRIMA 2019",

}

TY - GEN

T1 - Strategies for Energy-Aware Multi-agent Continuous Cooperative Patrolling Problems Subject to Requirements

AU - Wu, Lingying

AU - Sugawara, Toshiharu

PY - 2019

Y1 - 2019

N2 - This paper proposes a method of autonomous strategy learning for multiple cooperative agents integrated with a series of behavioral strategies aiming at reduction of energy cost on the premise of satisfying quality requirements in continuous patrolling problems. We improved our algorithm of requirement estimation to avoid concentration of agents since they are given the knowledge of the work environment in advance. The experimental results show that our proposal enables the agents to learn to select appropriate behavioral planning strategies according to performance efficiency and energy cost, and to individually estimate whether the given requirement is reached and modify their action plans to save energy. Furthermore, agents with the new requirement estimation method could achieve fair patrolling by introducing local observations.

AB - This paper proposes a method of autonomous strategy learning for multiple cooperative agents integrated with a series of behavioral strategies aiming at reduction of energy cost on the premise of satisfying quality requirements in continuous patrolling problems. We improved our algorithm of requirement estimation to avoid concentration of agents since they are given the knowledge of the work environment in advance. The experimental results show that our proposal enables the agents to learn to select appropriate behavioral planning strategies according to performance efficiency and energy cost, and to individually estimate whether the given requirement is reached and modify their action plans to save energy. Furthermore, agents with the new requirement estimation method could achieve fair patrolling by introducing local observations.

KW - Continuous patrolling

KW - Cooperation

KW - Energy efficiency

KW - Learning

KW - Multi-agent systems

UR - http://www.scopus.com/inward/record.url?scp=85076443090&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85076443090&partnerID=8YFLogxK

U2 - 10.1007/978-3-030-33792-6_44

DO - 10.1007/978-3-030-33792-6_44

M3 - Conference contribution

AN - SCOPUS:85076443090

SN - 9783030337919

T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

SP - 585

EP - 593

BT - PRIMA 2019

A2 - Baldoni, Matteo

A2 - Dastani, Mehdi

A2 - Liao, Beishui

A2 - Sakurai, Yuko

A2 - Zalila Wenkstern, Rym

PB - Springer

T2 - 22nd International Conference on Principles and Practice of Multi-Agent Systems, PRIMA 2019

Y2 - 28 October 2019 through 31 October 2019

ER -

Strategies for Energy-Aware Multi-agent Continuous Cooperative Patrolling Problems Subject to Requirements

抄録

出版物シリーズ

Conference

ASJC Scopus subject areas

UN SDG

文献へのアクセス

他のファイルとリンク

フィンガープリント

引用スタイル