无人机认知防碰撞系统安全边界研究

Study on safety boundary of cognitive anti-collision system for unmanned aerial vehicle

无人机认知防碰撞控制是借鉴人类认知智能,建立复杂动态环境无人机防碰撞控制的新思路,其认知决策的安全边界是系统设计的重要问题.本文整体考虑环境威胁与无人机的相互作用关系,建立了“认知无人机-环境系统”模型.进而定义了系统的防碰撞稳定性,并导出了认知无人机-环境系统的防碰撞稳定条件.在此基础上,通过分析防碰撞稳定的安全特征量,解算出面向动态正向碰撞的无人机认知防碰撞系统的安全边界.仿真实验分析了“云雀”无人机的安全边界特性和防碰撞控制要求,并与其他方法进行了对比.

The cognitive anti-collision control method is a novel idea to build anti-collision control of UAV in complex dynamic environment by using human cognition intelligence.The safety boundary of cognitive decision-making is an important issue in system design.This paper considers the interaction between obstacles and UAVs as a whole,and establishes a“cognitive UAV-environment system”model.Furthermore,the anti-collision stability of the system is defined,and the anti-collision stability condition of the cognitive UAV-environment system is derived.On this basis,by analyzing the safety feature of anti-collision stability,the safety boundary of the UAV cognitive anti-collision system for dynamic forward collision is solved.The simulation experiment analyzes the safety boundary characteristics and anti-collision control requirements of the“Lark”UAV,and the comparation with other methods is carried out at last.