Verification and Evaluation of Collision Avoidance Algorithms

Unmanned aerial vehicles (UAVs) are being utilized in a growing variety of military and civil situations. One major issue in situations that require multiple UAVs to share the same air space is collision avoidance. A successful collision avoidance system should allow a UAV to prevent all collisions without greatly impacting flight path efficiency. This study reviews literature centered on collision avoidance, and selects a variety of avoidance algorithms for verification and evaluation. Each of these approaches were implemented and simulated on a common infrastructure. This process revealed the strengths and weaknesses of each individual approach as well as their overall success and efficiency. Among the algorithms chosen for evaluation, an inverse proportional navigation approach to collision avoidance appeared to be the most effective approach, resulting in the least collisions, and greatest path efficiency.