Multi-UAV Coverage Path Planning for Autonomous Aircraft Inspection

This paper addresses the multi-UAV path-planning problem for autonomous aircraft inspection. Building on our previous research on parameterization-based inspection path planning, we apply this methodology to a multi-agent scenario. First, we compute a set of viewpoints using the parameterization-based planning algorithm, which enhances inspection quality. We then formulate and solve a multi-objective multiple traveling salesmen problem (MOMTSP) using our proposed heuristic algorithm, Non-dominated Sorting Genetic Ant Colony Optimization (NSGACO). Additionally, we implement a migration mechanism to identify the optimal depot location, ensuring the repeatability of the inspection mission. We validate the performance of NSGACO through numerical simulations, demonstrating its superiority over other heuristic approaches. Finally, software-in-the-loop simulations reveal significant improvements in inspection quality; specifically, when compared to traditional sampling approaches, the reconstruction model quality is markedly enhanced.