LINDA:multi-agent local information decomposition for awareness of teammates  

作　　者：Jiahan CAO Lei YUAN Jianhao WANG Shaowei ZHANG Chongjie ZHANG Yang YU De-Chuan ZHAN 

机构地区：[1]National Key Laboratory for Novel Software Technology,Nanjing University,Nanjing 210046,China [2]Institute for Interdisciplinary Information Sciences,Tsinghua University,Beijing 100084,China [3]Polixir Technologies,Nanjing 210046,China

出　　处：《Science China(Information Sciences)》2023年第8期148-164,共17页中国科学（信息科学）（英文版）

基　　金：supported by National Natural Science Foundation of China(Grant No.61773198)。

摘　　要：In cooperative multi-agent reinforcement learning(MARL),where agents only have access to partial observations,efficiently leveraging local information is critical.During long-time observations,agents can build awareness for teammates to alleviate the restriction of partial observability.However,previous MARL methods usually neglect awareness learning from local information for better collaboration.To address this problem,we propose a novel framework,multi-agent local information decomposition for awareness of teammates(LINDA),with which agents learn to decompose local information and build awareness for each teammate.We model the awareness as stochastic random variables and perform representation learning to ensure the informativeness of awareness representations by maximizing the mutual information between awareness and the actual trajectory of the corresponding agent.LINDA is agnostic to specific algorithms and can be flexibly integrated with different MARL methods.Sufficient experiments show that the proposed framework learns informative awareness from local partial observations for better collaboration and significantly improves the learning performance,especially on challenging tasks.

关 键 词：AGENT local FLEX 

分 类 号：TP181[自动化与计算机技术—控制理论与控制工程]