Autonomous Landing of Fixed Wing Unmanned Aerial Vehicle with Reactive Collision Avoidance

摘要

Auto landing of unmanned aerial vehicles with reactive collision avoidance is presented in this paper. Unmanned aerial vehicle follows approach phase, glide slope phase and flare phases during auto landing maneuver. This auto landing unmanned aerial vehicle encounters another unmanned aerial vehicle which comes across its glide path and poses a threat of possible collision. The other vehicle is termed as an 'obstacle' for the auto landing vehicle. The prediction logic is used to compute the minimum distance between two vehicles. If the minimum predicted distance is less than a predefined safety distance then the possible collision is predicted. The collision cone based approach is used to compute the aiming point. The velocity vector is deflected using guidance strategies to reach the aiming point. The differential geometric guidance is used to reach the aiming point in the available time to go. The auto landing vehicle senses the obstacle approaching towards it through stereo vision sensing and performs a collision avoidance maneuver. During collision avoidance maneuver, auto landing vehicle changes its speed as well as direction. However, the speed reduction is subjected to stall speed of the vehicle. The non cooperative scenario is considered where only the auto landing vehicle performs the collision avoidance and the obstacle follows its nominal path. The cooperative scenario is also considered where both the vehicles perform collision avoidance maneuver and rejoin their respective path post collision avoidance maneuver. The six degree of freedom model of both the unmanned aerial vehicles is used for simulation.