Trajectory Planning for UAVs in 3D Environments Using a Moving Band in Potential Sigmoid Fields

An interesting aspect of unmanned aerial vehicles is to determine an appropriate trajectory in order them to be moved from an initial point to a target position. In addition, it is important to guarantee a collision-free trajectory, because the workspace generally has obstacles. An algorithm to propose trajectories is presented, it uses sigmoidal functions to model obstacles with arbitrary shapes. Then, the algorithm uses a model of band composed by moving points, which spread out from the initial point to the target and change their position because of the action of the potential field generated by the obstacles. The algorithm is a contribution to this field as it simultaneously overcomes some problems such as: local minima, oscillations in the trajectory and destination points close to any obstacle. Additionally, the algorithm takes into account arbitrary shapes of the objects in the environment. Moreover, it allows the processing in parallel. Furthermore, the algorithm allows its implementation in 3D environments, where it is also possible to avoid an obstacle by going over or below it.