无人机安全间隔的确定方法

The Method for Determining the Safety Separation of Unmanned Aerial Vehicles

为准确描述无人机空中运行风险,并确定安全间隔,建立了一种综合考虑碰撞概率和冲突率的碰撞风险模型。首先根据无结构空域中两种不同的运行条件,分别建立了碰撞概率模型,针对空域内无法获取动态且无避撞措施的无人机,建立了结合航向的不确定性和位置误差的碰撞概率模型,计算碰撞概率对应的安全间隔;针对空域内可以获取动态且采取避撞措施的无人机,基于其机动性能以及位置误差,系统误差和人为因素,确定避撞所需的安全间隔并计算相应碰撞概率。然后,根据多旋翼无人机在无结构空域运行的特点,使用气体模型以确定空域内的冲突率,将前一步计算得到的安全间隔作为冲突距离,输入进气体模型,计算空域内的冲突频率。最后结合安全间隔对应的碰撞概率和冲突率,得到空域内的碰撞风险。以物流无人机为例,使用上述方法计算无结构空域内碰撞风险和安全间隔,并对两种不同模式进行比较分析,结果表明在无法获取空域内其他无人机的动态信息且无避撞措施时,5%碰撞概率对应的安全间隔为最佳;在可获取动态信息并采取避撞措施时,碰撞风险随着冲突条件下的碰撞概率减少而减少。两种模型均为可行的和有效的,可以为在此两种条件下运行的无人机安全间隔提供参考。

To accurately describe the operational risks of UAV(unmanned aerial vehicles) in air and determine safety separations,a collision risk model was established that considers both collision probability and collision rate.Based on the characteristics of multi rotor UAVs operating in unstructured airspace,firstly,for unmanned aerial vehicles that cannot obtain dynamic and have no collision avoidance measures in the airspace,a collision probability model was established that combines the uncertainty of heading and position error,and different collision probabilities were used as safety target levels to determine safety separations.Then,for other UAVs in the airspace that can obtain dynamics and take collision avoidance measures,based on their maneuverability,position error,and system error,determine the safety separation and calculate the collision probability.Then,the safety separation was used as the conflict distance to calculate the conflict frequency in the airspace,and finally,the collision risk in the airspace was obtained by combining the collision probability.Taking the logistics UAV as an example,the collision risk and safety separation were calculated using the above method,and two different modes were compared and analyzed.The results show that when it is impossible to obtain dynamic information of other drones in the airspace and there are no collision avoidance measures,the safety separation corresponding to a 5% collision probability is the best option.When dynamic information can be obtained and collision avoidance measures can be taken,the collision risk decreases as the collision probability under conflict conditions decreases.Both models are feasible and effective,and can provide reference for the safety separation of unmanned aerial vehicles operating under these two conditions.