This paper proposes an improved ant colony optimization-based path planning method for UAV inspection in substations. Considering the equipment partition characteristics and no-fly zone constraints, a two-dimensional inspection scenario model is constructed with typical equipment areas, inspection points, a depot, and no-fly zones. The fixed partition with the nearest-neighbor method is used as the baseline, and the basic ACO algorithm is introduced for global path search. To further improve path quality, candidate neighborhood selection, elite pheromone updating, integrated turning and obstacle-avoidance costs, and local optimization are incorporated into the improved ACO. Simulation results based on 30 independent runs show that the improved ACO achieves an average path length of 1694.08 m and an average estimated flight time of 372.27 s in the 24-point scenario, reducing these two metrics by 22.30% and 20.89%, respectively, compared with the baseline method. Compared with the basic ACO, the improved ACO further reduces the average path length and estimated flight time by 2.28% and 2.41%, respectively, with statistically significant differences. Comparisons with GA and PSO and scalability experiments under different inspection point scales further demonstrate the effectiveness of the proposed method.
Guo et al. (Fri,) studied this question.