集群智能与协同控制

多智能体多耦合任务混合式智能决策架构设计

  • 王雪鉴 ,
  • 文永明 ,
  • 石晓荣 ,
  • 张宁宁 ,
  • 刘洁玺
展开
  • 北京控制与电子技术研究所,北京 100038
.E-mail: mely0110@sina.com

收稿日期: 2023-10-26

  修回日期: 2023-11-21

  录用日期: 2023-12-20

  网络出版日期: 2024-01-04

Design of hybrid intelligent decision framework for multi⁃agent and multi⁃coupling tasks

  • Xuejian WANG ,
  • Yongming WEN ,
  • Xiaorong SHI ,
  • Ningning ZHANG ,
  • Jiexi LIU
Expand
  • Beijing Institute of Control & Electronics Technology,Beijing 100038,China
E-mail: mely0110@sina.com

Received date: 2023-10-26

  Revised date: 2023-11-21

  Accepted date: 2023-12-20

  Online published: 2024-01-04

摘要

针对多智能体在实际复杂应用场景下面对的任务分配和路径规划等多任务相互耦合问题及其决策问题,提出了一种多智能体多耦合任务混合式智能决策架构设计方法。首先,结合单智能体多任务混合式架构和多智能体分布式协同控制的优点,设计了多智能体面向多耦合任务的混合式智能决策架构;其次,对架构的策略网络以及策略网络训练控制器进行设计,并提出了基于耦合关系的耦合关系矩阵,实现了多智能体多任务在面对协同决策问题时的高效训练;最后,在仿真环境下进行建模、算法训练与仿真,并通过与传统方法进行对比试验,验证了所提方法的有效性和先进性。

本文引用格式

王雪鉴 , 文永明 , 石晓荣 , 张宁宁 , 刘洁玺 . 多智能体多耦合任务混合式智能决策架构设计[J]. 航空学报, 2023 , 44(S2) : 729770 -729770 . DOI: 10.7527/S1000-6893.2023.29770

Abstract

To address the coupling problem and decision-making problem of multiple tasks such as task allocation and path planning of multi-agents in complex application scenarios, a design method of hybrid intelligent decision-making framework for multi-agent and multi-coupling tasks is proposed. Firstly, the advantages of single agent multi-task hybrid framework and multi-agent distributed collaborative control, a hybrid intelligent decision-making framework for multi-agent and multi-coupling tasks is designed. Secondly, the strategy network of the framework and the training controller for the strategy network are designed and a coupling relationship matrix based on coupling relationships is proposed to achieve efficient training of multi-agents and multi-tasks in face of collaborative decision-making problems. Finally, this paper modeled, trained algorithm and simulated in simulation environment,and compared with the tradition method to verifies the effectiveness and advantages of the proposed method.

参考文献

1 VERMA J K, RANGA V. Multi-robot coordination analysis, taxonomy, challenges and future scope[J]. Journal of Intelligent & Robotic Systems2021102(1): 10.
2 林萌龙, 陈涛, 任棒棒, 等. 基于多智能体深度强化学习的体系任务分配方法[J]. 指挥与控制学报20239(1): 93-102.
  LIN M L, CHEN T, REN B B, et al. Task assignment method of operation system of systems based on multi- agent deep reinforcement learning[J]. Journal of Command and Control20239(1): 93-102 (in Chinese).
3 TANG J, DUAN H B, LAO S Y. Swarm intelligence algorithms for multiple unmanned aerial vehicles collaboration: A comprehensive review[J]. Artificial Intelligence Review202356(5): 4295-4327.
4 GUO J G, HU G J, GUO Z Y, et al. Evaluation model, intelligent assignment, and cooperative interception in multimissile and multitarget engagement[J]. IEEE Transactions on Aerospace and Electronic Systems202258(4): 3104-3115.
5 刘庆周, 吴锋. 多智能体路径规划研究进展[J]. 计算机工程202046(4): 1-10.
  LIU Q Z, WU F. Research progress of multi-agent path planning[J]. Computer Engineering202046(4): 1-10 (in Chinese).
6 RAHMAN M, ALAM M A, ISLAM M M, et al. An adaptive agent-specific sub-optimal bounding approach for multi-agent path finding[J]. IEEE Access109210: 22226-22237.
7 刘文兵, 王艺栋. 多无人机协同搜索多目标的路径规划问题研究[J]. 电光与控制201926(3): 35-38, 73.
  LIU W B, WANG Y D. Path planning of multi-UAV cooperative search for multiple targets[J]. Electronics Optics & Control201926(3): 35-38, 73 (in Chinese).
8 丁文俊, 柴亚军, 侯冬冬, 等. AUV&UAV跨域协同搜索与跟踪路径规划[J]. 航空学报202344(21): 128471.
  DING W J, CHAI Y J, HOU D D, et al. Path planning for AUV & UAV cross-domain collaborative search and tracking[J]. Acta Aeronautica et Astronautica Sinica202344(21): 128471 (in Chinese).
9 刘畅, 谢文俊, 张鹏, 等. 多基地多无人机航迹避障任务规划[J]. 计算机工程201945(11): 275-280.
  LIU C, XIE W J, ZHANG P, et al. Mission planning for multi-base multi-UAV obstacle avoidance[J]. Computer Engineering201945(11): 275-280 (in Chinese).
10 赵飞虎, 李哲, 王宁, 等. 面向战场的多无人机协同打击航迹规划[J]. 电光与控制202330(9): 9-14, 91.
  ZHAO F H, LI Z, WANG N, et al. Path planning for multi-UAV cooperative strike in battlefield environments[J]. Electronics Optics & Control202330(9): 9-14, 91 (in Chinese).
11 陈明强, 冯树娟, 李奇峰. 基于改进人工势场的物流无人机三维航迹规划[J]. 无线电工程202353(10): 2352-2359.
  CHEN M Q, FENG S J, LI Q F. Three-dimensional trajectory planning of logistics UAV based on improved artificial potential field[J]. Radio Engineering202353(10): 2352-2359 (in Chinese).
12 贾高伟, 王建峰. 无人机集群任务规划方法研究综述[J]. 系统工程与电子技术202143(1): 99-111.
  JIA G W, WANG J F. Research review of UAV swarm mission planning method[J]. Systems Engineering and Electronics202143(1): 99-111 (in Chinese).
13 李璐璐,朱睿杰,隋璐瑶,等.智能集群系统的强化学习方法综述[J/OL].计算机学报. .
  LI L L, ZHU R J, SUI L Y, et al. The reinforcement learning methods for intelligent collective system: A survey[J/OL]. Chinese Journal of Computers. .
14 马卫华. 导弹/火箭制导、导航与控制技术发展与展望[J]. 宇航学报202041(7): 860-867.
  MA W H. Review and prospect of missile/launch vehicle guidance, navigation and control technologies[J]. Journal of Astronautics202041(7): 860-867 (in Chinese).
15 文永明, 石晓荣, 黄雪梅, 等. 一种无人机集群对抗多耦合任务智能决策方法[J]. 宇航学报202142(4): 504-512.
  WEN Y M, SHI X R, HUANG X M, et al. An intelligent decision-making method for multi-coupling tasks of UAV cluster countermeasure[J]. Journal of Astronautics202142(4): 504-512 (in Chinese).
16 刘全, 翟建伟, 章宗长, 等. 深度强化学习综述[J]. 计算机学报201841(1): 1-27.
  LIU Q, ZHAI J W, ZHANG Z C, et al. A survey on deep reinforcement learning[J]. Chinese Journal of Computers201841(1): 1-27 (in Chinese).
17 LOWE R, WU Y, TAMAR A, et al. Multi-agent actor-critic for mixed cooperative-competitive environments[C]∥ Proceedings of the 31st International Conference on Neural Information Processing Systems. New York: ACM, 2017: 6382–6393.
18 SUKHBAATAR S, SZLAM A, FERGUS R. Learning multiagent communication with backpropagation[C]∥ Proceedings of the 30th International Conference on Neural anInformation Processing Systems. New York: ACM, 2016: 2252–2260.
19 王子豪, 张严心, 黄志清, 等. 部分可观测下基于RGMAAC算法的多智能体协同[J]. 控制与决策202338(5): 1267-1277.
  WANG Z H, ZHANG Y X, HUANG Z Q, et al. Multi- agent collaboration based on RGMAAC algorithm under partial observability[J]. Control and Decision202338(5): 1267-1277 (in Chinese).
20 孙英博, 苗国英, 庄亚楠. 基于改进的深度强化学习多智能体协作方法[J]. 传感器与微系统202342(9): 25-29.
  SUN Y B, MIAO G Y, ZHUANG Y N. Multi-agent collaboration method based on improved deep reinforcement learning[J]. Transducer and Microsystem Technologies202342(9): 25-29 (in Chinese).
文章导航

/