Multiselection genetic algorithm for UAV-based coverage path planning

Due to its excellent maneuverability and wide-area coverage, unmanned aerial vehicle (UAV) has become essential tool in forest search and rescue missions. However, UAV is limited by its minimum turning radius when performing area coverage tasks. Frequent turns can create redundant paths and prolong rescue times, resulting in personnel not being rescued in time. To address this issue, this paper proposes a multi-selection genetic algorithm (MSGA) to solve the path planning problem in area coverage. First, the initial population generation mechanism is proposed, which combines a distance-based heuristic and a greedy algorithm strategy. This enhances the algorithm's diversity and strengthens its search capabilities in complex environments. Second, a multi-selection crossover operator is proposed to improve the diversity and convergence speed of the solution. Simulation results demonstrate that the proposed method reduces flight path length and improves the effectiveness of the crossover operator. The experimental results show that, compared with the comparative algorithms (SPC, GA, and DEGA), the proposed algorithm reduces the flight path length within the convex pentagonal region by 17.9%, 16.2%, and 8.3%, respectively. In addition, the box plot demonstrates that the improved crossover operator provides better stability than the comparison algorithms.