类脑发育无人机防碰撞控制

A brain-like mechanism for developmental UAVs’ collision avoidance

城市无人机的大量普及,对无人机在密集动态环境中的防碰撞问题提出了严峻挑战.论文针对密集动态环境中碰撞威胁模式复杂多变的特点,设计了无人机的自主心智发育型防碰撞控制架构,建立了可自主发育的威胁–规避映射关系,构建了威胁模式知识库和规避策略知识库,进而提出了基于自主心智发育的无人机防碰撞控制方法.通过将有效可信的知识不断导入发育器进行发育,可以持续提高无人机应对密集动态障碍的防碰撞能力.仿真对比实验表明,本文方法不仅能快速构建安全避障航路,而且算法效率随知识库的扩充而提高,能够极大提高无人机在密集动态障碍环境中飞行安全性.

With the popularization and development of unmanned aerial vehicles (UAVs), there are crowded with hundreds of UAVs which, as dense dynamic obstacles, produce serious challenges for UAVs’ safety. To avoid the flight-risk in this complex and changeful environment, we first design a mechanism for autonomous mental development UAVs, then developmental threat-avoid mapping, threat pattern base and avoid strategy base are constructed. So a control strategy on the basis of autonomous mental development is proposed. And UAVs’ ability of avoiding collision in dense and dynamic environment is improved through continuous importing reliable knowledge to developmental unit. Through comparing simulation results of experiments, we find that the proposed method can not only react fast to dense obstacles and design a safe path with avoidance quickly, but also improve the efficiency of algorithm with accumulating of knowledge base, thus improving the safety of UAVs flying in dense dynamic obstacles.