A Fuzzy Logic Approach to Collision Avoidance in Smart UAVs

In recent years, there has been a rise in attention given to the issue of collision avoidance techniques for unmanned aerial vehicles(UAVs). Many approaches have been proposed but very few have been tested in high density, fixed speed test conditions. This approach successfully uses fuzzy logic to determine an appropriate avoidance maneuver after a possible collision is detected. Initially, fuzzy logic was also used to detect possible collisions, the reasons this level was removed from the final results are detailed within. This approach was chosen for its adaptability, ease of implementation, and robustness in dealing with questionable sensing techniques. The methods used are explained in detail and the simulated results are presented after testing on two different field sizes, each with four levels of plane density. Overall, this approach was able to reduce collisions by a minimum of 98% in all cases.