Robust Path Planning for Multiple UAVs Considering Position Uncertainty

With the widespread application of unmanned aerial vehicles (UAVs), the issue of path planning has become increasingly significant in search of suitable paths for UAVs. However, positioning errors may exist in the system carried by the UAV in practical situations, leading to the suboptimal or even unsafe path execution. In view of this, we construct the multi-UAV robust path planning model under position uncertainty by incorporating several important considerations. This can be expressed as a complicated robust optimization problem, aiming to obtain a robust optimal path for each UAV in the presence of positioning errors. Based on this, we introduce the corresponding overall cost function and its expected expression for robust evaluation. Then, we propose a novel robust particle swarm optimization (PSO) algorithm, which employs the scale-free topology to characterize the individual interactions in the swarm. And an improved explicit sampling technique is developed by introducing a sampling coefficient, where the number of samples increases proportional to the degree value for a particle in PSO, allowing effective robustness evaluation for each solution. The proposed algorithm shows great advantages on benchmark functions, compared with some other robust PSO algorithms. Further, we present the specific implementation of the multi-UAV robust path planning method based on the proposed algorithm. Finally, simulation experiments on various path planning scenarios and comparison results indicate the superiority of the developed method, which can plan a robust and effective path for each UAV.