轮式机器人玩链球游戏的策略与评价
Wheeled Robots playing Chain Catch: Strategies and Evaluation
摘要:
机器人玩人类擅长的游戏是一个挑战。我们研究了机器人代理玩链捕捉游戏作为一个多智能体系统(MAS)。我们的游戏从一个传统的追赶游戏开始,类似于追逐逃避,并进一步扩展它,形成一个不断增长的捕食者代理链,以追逐剩余的猎物。因此,Chain Catch是两个挑战的结合-追求域和机器人链形成。这些游戏要求机器人代理团队相互合作,并通过快速决策与其他代理团队竞争。在本文中,我们提出了一个链捕捉模拟器,使我们能够将游戏规则,设计策略和模拟游戏。我们为每个Escapee,Catcher和Chain开发了成本模型驱动策略。研究结果表明,在链捕中,对于逃逸者来说,滑动斜坡策略是最佳策略,而对于链的运动来说,标记策略是最佳策略。我们还使用生产质量的机器人在物理环境中实现游戏,并在真实机器人上分析游戏策略。我们在不同场景中的真实机器人实现表明,游戏策略如预期的那样工作,并且在每个游戏中成功地形成完整的链。
英文摘要:
Robots playing games that humans are adept in is a challenge. We studied robotic agents playing Chain Catch game as a Multi-Agent System (MAS). Our game starts with a traditional Catch game similar to Pursuit evasion, and further extends it to form a growing chain of predator agents to chase remaining preys. Hence Chain Catch is a combination of two challenges - pursuit domain and robotic chain formation. These are games that require team of robotic agents to cooperate among themselves and to compete with other group of agents through quick decision making. In this paper, we present a Chain Catch simulator that allows us to incorporate game rules, design strategies and simulate the game play. We developed cost model driven strategies for each of Escapee, Catcher and Chain. Our results show that Sliding slope strategy is the best strategy for Escapees whereas Tagging method is the best method for chain s movement in Chain Catch. We also use production quality robots to implement the game play in a physical environment and analyze game strategies on real robots. Our real robots implementation in different scenarios shows that game strategies work as expected and a complete chain formation takes place successfully in each game.
机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。