论文部分内容阅读
基于增强学习的多机器人系统优化控制是近年来机器人学与分布式人工智能的前沿研究领域.多机器人系统具有分布、异构和高维连续空间等特性,使得面向多机器人系统的增强学习的研究面临着一系列挑战,为此,对其相关理论和算法的研究进展进行了系统综述.首先,阐述了多机器人增强学习的基本理论模型和优化目标;然后,在对已有学习算法进行对比分析的基础上,重点探讨了多机器人增强学习理论与应用研究中的困难和求解思路,给出了若干典型问题和应用实例;最后,对相关研究进行了总结和展望.
The optimization control of multi-robot system based on reinforcement learning is a frontier research field of robotics and distributed artificial intelligence in recent years.Multi-robot system has such characteristics as distribution, heterogeneity and high-dimensional continuous space that makes multi-robot system enhanced learning Facing a series of challenges, this paper systematically reviews the progress of its related theories and algorithms.Firstly, the basic theoretical models and optimization objectives of multi-robot enhanced learning are expounded. Then, Based on this, the difficulties and solutions in multi-robot enhancement learning theory and application research are discussed. Some typical problems and application examples are given. Finally, the related research is summarized and prospected.