Full Text:   <155>

CLC number: TP242.6

On-line Access: 2022-07-21

Received: 2021-12-31

Revision Accepted: 2022-07-21

Crosschecked: 2022-05-23

Cited: 0

Clicked: 122

Citations:  Bibtex RefMan EndNote GB/T7714

 ORCID:

Hanqi DAI

https://orcid.org/0000-0002-3724-4389

Weining LU

https://orcid.org/0000-0002-0927-1259

-   Go to

Article info.
Open peer comments

Frontiers of Information Technology & Electronic Engineering  2022 Vol.23 No.7 P.1069-1076

http://doi.org/10.1631/FITEE.2100597


Cooperative planning of multi-agent systems based on task-oriented knowledge fusion with graph neural networks


Author(s):  Hanqi DAI, Weining LU, Xianglong LI, Jun YANG, Deshan MENG, Yanze LIU, Bin LIANG

Affiliation(s):  Department of Automation, Tsinghua University, Beijing 100084, China; more

Corresponding email(s):   dhq19@mails.tsinghua.edu.cn, luwn@tsinghua.edu.cn

Key Words:  Multi-agent system, Cooperative planning, GraphSAGE, Task-oriented knowledge fusion


Hanqi DAI, Weining LU, Xianglong LI, Jun YANG, Deshan MENG, Yanze LIU, Bin LIANG. Cooperative planning of multi-agent systems based on task-oriented knowledge fusion with graph neural networks[J]. Frontiers of Information Technology & Electronic Engineering, 2022, 23(7): 1069-1076.

@article{title="Cooperative planning of multi-agent systems based on task-oriented knowledge fusion with graph neural networks",
author="Hanqi DAI, Weining LU, Xianglong LI, Jun YANG, Deshan MENG, Yanze LIU, Bin LIANG",
journal="Frontiers of Information Technology & Electronic Engineering",
volume="23",
number="7",
pages="1069-1076",
year="2022",
publisher="Zhejiang University Press & Springer",
doi="10.1631/FITEE.2100597"
}

%0 Journal Article
%T Cooperative planning of multi-agent systems based on task-oriented knowledge fusion with graph neural networks
%A Hanqi DAI
%A Weining LU
%A Xianglong LI
%A Jun YANG
%A Deshan MENG
%A Yanze LIU
%A Bin LIANG
%J Frontiers of Information Technology & Electronic Engineering
%V 23
%N 7
%P 1069-1076
%@ 2095-9184
%D 2022
%I Zhejiang University Press & Springer
%DOI 10.1631/FITEE.2100597

TY - JOUR
T1 - Cooperative planning of multi-agent systems based on task-oriented knowledge fusion with graph neural networks
A1 - Hanqi DAI
A1 - Weining LU
A1 - Xianglong LI
A1 - Jun YANG
A1 - Deshan MENG
A1 - Yanze LIU
A1 - Bin LIANG
J0 - Frontiers of Information Technology & Electronic Engineering
VL - 23
IS - 7
SP - 1069
EP - 1076
%@ 2095-9184
Y1 - 2022
PB - Zhejiang University Press & Springer
ER -
DOI - 10.1631/FITEE.2100597


Abstract: 
cooperative planning is one of the critical problems in the field of multi-agent system gaming. This work focuses on cooperative planning when each agent has only a local observation range and local communication. We propose a novel cooperative planning architecture that combines a graph neural network with a task-oriented knowledge fusion sampling method. Two main contributions of this paper are based on the comparisons with previous work: (1) we realize feasible and dynamic adjacent information fusion using graphSAGE (i.e., Graph SAmple and aggreGatE), which is the first time this method has been used to deal with the cooperative planning problem, and (2) a task-oriented sampling method is proposed to aggregate the available knowledge from a particular orientation, to obtain an effective and stable training process in our model. Experimental results demonstrate the good performance of our proposed method.

基于融合任务信息图神经网络的多智能体系统协同规划

戴汉奇1,2,芦维宁2,李祥隆3,杨君1,孟德山4,刘衍泽5,梁斌1
1清华大学自动化系,中国北京市,100084
2清华大学北京信息科学与技术国家研究中心,中国北京市,100084
3天津大学科学技术发展研究院,中国天津市,300350
4中山大学航空航天学院,中国深圳市,518107
5诺丁汉大学电气与电子工程系,中国宁波市,315154
摘要:协同规划是多智能体系统博弈领域的关键问题之一。本文聚焦每个智能体只有一个局部观测范围和局部通信情况下的协作规划。提出一种新型协同规划框架,该框架将图神经网络与融合任务信息采样方法相结合。本文的两个主要贡献是基于与前期工作的比较:(1)使用图采样与聚合方法(GraphSAGE)实现动态近邻智能体信息融合,这是该方法首次用于处理协同规划问题;(2)提出一种面向任务的采样方法,从特定方向聚合知识,使所提模型获得高效、稳定的训练过程。实验结果证明了所提方法的有效性。

关键词:多智能体系统;协同规划;图采样与聚合(GraphSAGE);融合任务信息

Darkslateblue:Affiliate; Royal Blue:Author; Turquoise:Article

Reference

[1]Barer M, Sharon G, Stern R, et al., 2014. Suboptimal variants of the conflict-based search algorithm for the multi-agent pathfinding problem. Proc 21st European Conf on Artificial Intelligence, p.961-962.

[2]Dresner K, Stone P, 2008. A multiagent approach to autonomous intersection management. J Artif Intell Res, 31(1):591-656.

[3]Enright JJ, Wurman PR, 2011. Optimization and coordinated autonomy in mobile fulfillment systems. Proc 9th AAAI Conf on Automated Action Planning for Autonomous Mobile Robots, p.33-38.

[4]Hamilton WL, Ying R, Leskovec J, 2017. Inductive representation learning on large graphs. Proc 31st Int Conf on Neural Information Processing Systems, p.1025-1035.

[5]Krizhevsky A, Sutskever I, Hinton GE, 2017. ImageNet classification with deep convolutional neural networks. Commun ACM, 60(6):84-90.

[6]Li QB, Gama F, Ribeiro A, et al., 2020. Graph neural networks for decentralized multi-robot path planning. IEEE/RSJ Int Conf on Intelligent Robots and Systems, p.11785-11792.

[7]López J, Pérez D, Zalama E, 2011. A framework for building mobile single and multi-robot applications. Robot Autonom Syst, 59(3-4):151-162.

[8]Ortega A, Frossard P, Kovačević J, et al., 2018. Graph signal processing: overview, challenges, and applications. Proc IEEE, 106(5):808-828.

[9]Prorok A, Kumar V, 2017. Privacy-preserving vehicle assignment for mobility-on-demand systems. IEEE/RSJ Int Conf on Intelligent Robots and Systems, p.1869-1876.

[10]Sartoretti G, Kerr J, Shi YF, et al., 2019. Primal: pathfinding via reinforcement and imitation multi-agent learning. IEEE Robot Autom Lett, 4(3):2378-2385.

[11]Singhal V, Dahiya D, 2015. Distributed task allocation in dynamic multi-agent system. Int Conf on Computing, Communication and Automation, p.643-648.

[12]Standley T, Korf R, 2011. Complete algorithms for cooperative pathfinding problems. Proc 22nd Int Joint Conf on Artificial Intelligence, p.668-673.

[13]van den Berg J, Lin M, Manocha D, 2008. Reciprocal velocity obstacles for real-time multi-agent navigation. IEEE Int Conf on Robotics and Automation, p.1928-1935.

[14]van den Berg J, Guy SJ, Lin M, et al., 2011. Reciprocal n-body collision avoidance. In: Pradalier C, Siegwart R, Hirzinger G (Eds.), Robotics Research. Springer, Berlin, Heidelberg, p.3-19.

[15]van den Berg JP, Overmars MH, 2005. Prioritized motion planning for multiple robots. IEEE/RSJ Int Conf on Intelligent Robots and Systems, p.430-435.

[16]Veloso M, Biswas J, Coltin B, et al., 2015. Cobots: robust symbiotic autonomous mobile service robots. 24th Int Joint Conf on Artificial Intelligence, p.4423-4429.

[17]Verma JK, Ranga V, 2021. Multi-robot coordination analysis, taxonomy, challenges and future scope. J Intell Rob Syst, 102:10.

[18]Wu ZH, Pan SR, Chen FW, et al., 2021. A comprehensive survey on graph neural networks. IEEE Trans Neur Netw Learn Syst, 32(1):4-24.

[19]Wurman PR, D'Andrea R, Mountz M, 2007. Coordinating hundreds of cooperative, autonomous vehicles in warehouses. Proc 19th National Conf on Innovative Applications of Artificial Intelligence, p.1752-1759.

[20]Yu JJ, LaValle SM, 2013. Structure and intractability of optimal multi-robot path planning on graphs. Proc 27th AAAI Conf on Artificial Intelligence, p.1443-1449.

[21]Zhou J, Cui GQ, Hu SD, et al., 2020. Graph neural networks: a review of methods and applications. AI Open, 1:57-81.

Open peer comments: Debate/Discuss/Question/Opinion

<1>

Please provide your name, email address and a comment





Journal of Zhejiang University-SCIENCE, 38 Zheda Road, Hangzhou 310027, China
Tel: +86-571-87952783; E-mail: cjzhang@zju.edu.cn
Copyright © 2000 - 2022 Journal of Zhejiang University-SCIENCE