The Important Role of Global State for Multi-Agent Reinforcement Learning

Li，Shuailong1,2,3; Zhang，Wei1,3; Leng，Yuquan4,5; Wang，Xiaohui1,2,3

2022

10.3390/fi14010017

Future Internet   影响因子和分区

1999-5903

Environmental information plays an important role in deep reinforcement learning (DRL). However, many algorithms do not pay much attention to environmental information. In multi-agent reinforcement learning decision-making, because agents need to make decisions combined with the information of other agents in the environment, this makes the environmental information more important. To prove the importance of environmental information, we added environmental information to the algorithm. We evaluated many algorithms on a challenging set of StarCraft II micromanagement tasks. Compared with the original algorithm, the standard deviation (except for the VDN algorithm) was smaller than that of the original algorithm, which shows that our algorithm has better stability. The average score of our algorithm was higher than that of the original algorithm (except for VDN and COMA), which shows that our work significantly outperforms existing multi-agent RL methods.

Deep reinforcement learning Environmental information Multi-agent reinforcement learning

ESCI ; EI

英语

通讯

National Natural Science Foundation of China[51805237];National Natural Science Foundation of China[52175272];

WOS:000759104500001

20220211454055

Decision making ; Multi agent systems ; Reinforcement learning

Ergonomics and Human Factors Engineering:461.4 ; Artificial Intelligence:723.4 ; Management:912.2

2-s2.0-85122545113

Scopus

1.State Key Laboratory of Robotics,Shenyang Institute of Automation,Chinese Academy of Sciences,Shenyang,110016,China
2.Institutes for Robotics and Intelligent Manufacturing,Chinese Academy of Sciences,Shenyang,110169,China
3.University of Chinese Academy of Sciences,Beijing,100049,China
4.Shenzhen Key Laboratory of Biomimetic Robotics and Intelligent Systems,Department of Mechanical and Energy Engineering,Southern University of Science and Technology,Shenzhen,518055,China
5.Guangdong Provincial Key Laboratory of Human-Augmentation and Rehabilitation Robotics in Universities,Southern University of Science and Technology,Shenzhen,518055,China

Li，Shuailong,Zhang，Wei,Leng，Yuquan,et al. The Important Role of Global State for Multi-Agent Reinforcement Learning[J]. Future Internet,2022,14(1).

Li，Shuailong,Zhang，Wei,Leng，Yuquan,&Wang，Xiaohui.(2022).The Important Role of Global State for Multi-Agent Reinforcement Learning.Future Internet,14(1).

Li，Shuailong,et al."The Important Role of Global State for Multi-Agent Reinforcement Learning".Future Internet 14.1(2022).