Multi-Agent Deep Reinforcement Learning For Multi-Robot Systems: A Survey Of Challenges And Applications

Abstract

Multi-agent deep reinforcement learning has emerged as a powerful approach for addressing coordination and collaboration challenges in multi-robot systems. This survey provides a comprehensive overview of the current state of research in this domain. This paper covers key methodological challenges, such as the non-stationarity of the environment and the heterogeneity of the agents, as well as emerging approaches to address these challenges, including attention mechanisms and multi-agent reinforcement learning algorithms. Additionally, this paper reviews the practical applications of multi-agent deep reinforcement learning in multi-robot systems, such as navigation, cooperative manipulation, and distributed task allocation. The main purpose of this paper is to present the latest developments in the field and provide a clear understanding of the current multi-agent reinforcement learning strategy training methods and their potential for advancing multi-robot systems. The survey aims to serve as a comprehensive resource for researchers and practitioners working in the domain of multi-agent deep reinforcement learning for multi-robot systems.