由感知到动作决策一体化的类脑导航技术研究现状与未来发展
杨闯1, 刘建业1,2,3, 熊智1,2,3, 赖际舟1,2,3, 熊骏1
1. 南京航空航天大学 自动化学院 导航研究中心, 南京 211106;
2. 先进飞行器导航、控制与健康管理工业和信息化部重点实验室, 南京 211106;
3. 卫星通信与导航江苏高校协同创新中心, 南京 210016
YANG Chuang1, LIU Jianye1,2,3, XIONG Zhi1,2,3, LAI Jizhou1,2,3, XIONG Jun1
1. Navigation Research Center, College of Automation Engineering, Nanjing University of Aeronautics and Astronautics, Nanjing 211106, China;
2. Key Laboratory of Navigation, Control and Health-Management Technologies of Advanced Aerocraft, Ministry of Industry and Information Technology, Nanjing 211106, China;
3. Satellite Communication and Navigation Collaborative Innovation Center, Nanjing 210016, China
摘要: 随着脑与神经科学以及人工智能技术的持续发展,昆虫和哺乳动物大脑导航机理启发下的感知/认知/路径规划/动作决策一体化类脑导航技术得到了较大发展,可以实现由原始感知信息输入到导航动作决策的直接输出,呈现出接近动物端到端面向目标导航的智能行为,具有提高密集型无人机集群导航鲁棒性、准确性、实时响应动作、自主智能性以及计算效率的潜力。阐述了昆虫和哺乳动物大脑导航机理及其互补对称关系,以及昆虫和哺乳动物大脑导航机理启发的端到端类脑导航技术内涵;论述了类脑导航技术研究进展,包括类脑环境感知、类脑空间认知、面向目标类脑导航;分析了类脑导航向智能化、神经形态系统以及群体导航发展的新趋势;最后讨论了类脑导航技术应用于无人机密集集群系统时存在的挑战。
关键词: 类脑导航, 感知与动作决策一体化, 大脑导航, 神经形态系统, 类脑集群导航, 无人机密集集群, 人工智能
Abstract: With the sustained development of brain and neuroscience as well as artificial intelligence, brain-like navigation technology inspired by insect & mammal has been greatly developed, integrating the abilities of perception, cognition, path planning and action decision-making. The brain-inspired navigation can directly produce action decision-making of navigation based on the perception input from raw sensory cues, with emerging intelligence similar to the end-to-end character of goal-directed navigation of animal, which can potentially improve the performance of robustness, accuracy, real-time response to action decision-making, autonomy & intelligence, and computing efficiency of the dense UAV swarm navigation. In this paper, we first make a brief discussion of the neural basis of insect & mammal navigation, the complementary symmetry between insect swarm navigation and mammal navigation, and the concept of brain-like navigation technology inspired by insect & mammal. After that, we emphatically review the state of the art of the brain-inspired end-to-end navigation, including brain-inspired perception, spatial cognition, and goal-directed navigation. Then we describe the new trends of brain-inspired navigation towards intelligence, neuromorphic system and swarm navigation. Finally, we analyse the challenges of applying brain-inspired navigation technology to dense UAV swarm.
Key words: brain-inspired navigation, integration of perception and action decision, brain navigation, neuromorphic system, brain-inspired swarm navigation, dense UAV swarm, artificial intelligence