摘要
针对移动机器人在非结构环境下的导航任务,受哺乳动物空间认知方式的启发,提出一种基于生物认知进行移动机器人路径规划的方法.结合认知地图特性,模拟海马体的情景记忆形成机理,构建封装了场景感知、状态神经元及位姿感知相关信息的情景认知地图,实现了机器人对环境的认知.基于情景认知地图,以最小事件距离为准则,提出事件序列规划算法用于实时导航过程.实验结果表明,该控制算法能使机器人根据不同任务选择最佳规划路径.
Inspired by spatial cognition of mammals, a path planning method of mobile robots based on biological cognition is proposed for navigation tasks of mobile robots in the unstructured environment. Combined with characteristics of the cognitive map, an episodic cognitive map encapsulating the information of scene perception, state neurons and pose perception is built through simulating the formation mechanism of the episodic memory in the hippocampus, and the environment cognition is realized by the robot. Based on the episodic cognitive map, an event sequence planning algorithm for real-time navigation is put forward according to the minimum distance between events. Experimental results show that the mobile robot can choose the best planning path for different tasks with the proposed control algorithm.
出处
《机器人》
EI
CSCD
北大核心
2018年第6期894-902,共9页
Robot
基金
国家自然科学基金(61503057)
