UAVs Autonomous Collision Avoidance in Non-segregated Airspace using Dynamic Alerting Thresholds

Abstract

There is a pressing operational requirement to migrate Unmanned Aerial Vehicles (UAVs) outside the confines of segregated airspace. The Sense and Avoid capacity is one of the key elements to open the access to airspace for UAVs Referred to layered model of the airspace surrounding the UAV; a collision avoidance method for use in UAVs in nonsegregated airspace is outlined. As to the sense problem, the neighboring vehicles are supposed to exchange their dynamics through the ADS-B system at each sample time to predict the conflicts. Layered zones around the UAV are defined corresponding with hierarchical alarm levels. Dynamic alerting thresholds contribute to reducing fake alarms and neglected alarms by calculating the alerting thresholds based on relative position, azimuth and speed of two aircrafts. The detail algorithm of deriving dynamic alerting thresholds based on the UAV local reference frame is illustrated. Aiming at modifying the algorithm for head-on and tail-chase encounters, right-of-way rules as the maneuver logic are considered in the study. In the simulation, two encounter scenarios were modeled including an intruder with a level flight path intersection, another intruder climbing towards the UAV. Our simulation shows that different guidance advisory associated with hierarchical alarm levels can be real-time presented to ensure separation and avoid collision.