Towards a PDE-based large-scale decentralized solution for path planning of UAVs in shared airspace

摘要

Recently, there has been a tremendous increase of interest in utilizing Unmanned Aerial Vehicles (UAVs) for a number of civilian applications. With this increased interest, it is imperative that these UAVs are able to operate in shared airspace for enhanced efficiency. Multi-UAV systems are inherently safety-critical systems, which means that safety guarantees must be made to ensure no undesirable configurations, such as collisions, occur. This paper proposes a decentralized method based on a Partial Differential Equation (PDE) to generate collision-free 3D trajectories for multiple UAVs operating in a shared airspace. This method exploits the dynamical properties of multi-phase fluids flowing through a porous medium by modeling the porosity values as a function of the risk of collision. To highlight the feasibility for onboard implementation, we propose propose a machine learning technique for obtaining computationally efficient solutions of the PDE describing flow movements in porous medium. This method has been compared via a simulation study to two other path planning strategies, centralized and sequential planning, and the advantages of this method are presented. Furthermore, results from an experiment using three UAVs have been presented to demonstrate the applicability of the proposed method to real world implementation. (c) 2020 Elsevier Masson SAS. All rights reserved.