Collision Free Path Planning for Unmanned Aerial Vehicles in Environments with Dynamic Obstacles

Abstract

A relevant problem in unmanned aerial vehicles flight is planning paths in 3D environments. The problem can be solved using different behaviour constraints and performance criteria. The goal of this work is to design a collision-free path that allows an unmanned aerial vehicle to fly from a starting pose to a destination one in an 3D environment with dynamic obstacles. A very important phase in completing the proposed goal is environment modelling. Based on an original contribution made by the authors, the free space in the environment is modeled by using a cuboid representation. Accounting information about the velocity magnitude and direction of the moving objects, a novel 3D cell decomposition planning algorithm is proposed. Testing and evaluation through numerical simulation show the modification in path design when comparing static parts of the environment with dynamic obstacle events.