强化学习在无人车领域的应用与展望

Applications and Prospect of Reinforcement Learning in Unmanned Ground Vehicles

  • 摘要: 无人车(UGV)可替代人类自主地执行民用和军事任务,对未来智能交通及陆军装备发展有重要战略意义。随着人工智能技术的日益成熟,采用强化学习技术成为了无人车智能决策领域最受关注的发展趋势之一。本文首先简要概述了强化学习的发展历程、基础原理和核心算法;随后,分析总结了强化学习在无人车智能决策中的研究进展,包括障碍物规避、变道与超车、车道保持和道路交叉口通行四种典型场景;最后,针对基于强化学习的智能决策面临的问题和挑战,探讨并展望了未来的研究工作与潜在的研究方向。

     

    Abstract: Unmanned ground vehicle (UGV)can replace human to conduct civilian and military missions, which is of great strategic significance to the construction of intelligent transportation system and the development of army equipment. With the development of artificial intelligence technology, the reinforcement learning technology is regarded as one of the key technologies for UGV to realize intelligent decision making. Firstly, this paper briefly summarizes the development process, basic principles and main algorithms of reinforcement learning. Then, the research progress of reinforcement learning in intelligent decision-making of UGV is analysed and summarized, including obstacle avoidance, lane changing and overtaking, lane keeping, and intersection traffic. Finally, in view of the problems and challenges faced by intelligent decision making based on reinforcement learning, the future research work and potential research directions are discussed and prospected.

     

/

返回文章
返回