Coverage Path Planning of Heterogeneous Unmanned Aerial Vehicles Based on Dual-Stage Genetic Algorithm

To address the increasing complexity of modern autonomous applications, heterogeneous unmanned aerial vehicles(UAVs) with solid robustness and high parallelism have emerged to compensate for the limited capabilities and insufficient performance of single UAVs. In this study, we propose a dual-stage genetic algorithm to solve cooperative coverage path planning in complex boundary regions, specifically in the context of cooperative coverage search by heterogeneous UAVs. First, considering that UAVs have different flight attitudes and sensor performances, the complex boundary region is uniformly discretized. A regional coverage path planning model is established and solved using a linear programming formulation. Next, a two-layer coding mechanism represents the region allocation results and coverage path sequences. A density clustering strategy is used to generate the initial solution, and a dual-stage alternating optimization strategy is utilized to enhance the performance of the solution set, providing the optimal point-to-point flight paths between the regions for each UAV. The coverage paths are generated through the decoding scheme. Finally, the performance of the proposed method is evaluated experimentally in terms of execution time, task completion time, and deviation ratio.