Collision Avoidance System for Fixed-Wing UAVs using ADS-B Sensors

The integration of unmanned aerial vehicles (UAVs) into the national airspace system presents itself with a myriad of technical problems. One of the key requirements for this integration is the human equivalent level of safety, which requires the ability to detect and avoid other aircraft/obstacles in their flight path so that the UAVs complete their mission without any loss or damage to other aircraft or property. This presentation talks about the use of ADS-B (Automatic Dependent Surveillance- Broadcast) transponders for detection of collision of other similarly equipped aircraft/UAVs. ADS-B transponders can receive and broadcast global position and velocities among other pertinent information in a 100-nautical mile radius. This research used Ping-2020 ADS-B transponders for the collision detection. The collision detection and avoidance is first tested in software-in-the-loop simulation, which also uses the flight controller, Ardupilot, and ADS-B transponders in the loop. MAVproxoy, a UAV ground station software package, is used to communicate between in the autopilot and simulation environment via MAVLink. FlightGear flight simulator is used to visualize the motion of the UAVs. The research uses two fixed-wing aircraft equipped with Pixhawk autopilots, which allow autonomous waypoint navigation. The collision avoidance algorithms use a three-step system of detect, predict, and avoid. The algorithm calculates and sends the waypoints for collision avoidance to the autopilot. Using the kinematic equations, the UAV velocities can be calculated from the information received from GPS sensors, and future positions can be predicted. The collision avoidance algorithm is tested using the incoming information from real-time aircraft.