Abstract

Unmanned aerial vehicles have become a promising platform for inspecting large port machinery; however, quay cranes contain sparse but complex steel-frame structures, multiple inspection points, and narrow collision-constrained spaces, which make efficient inspection path planning difficult. Existing approaches often focus either on global point sequencing or local collision-free search, and conventional global optimizers usually use straight-line distances that do not reflect obstacle-constrained flight costs. This paper proposes an integrated path planning method for quay crane inspection based on vector jump point search and ant colony optimization. In the local path-searching stage, vector-guided preprocessing and path simplification are used to calculate collision-free paths between mission points and construct a path cost matrix. In the global optimization stage, ant colony optimization determines the inspection sequence using the collision-free cost matrix rather than Euclidean distances. Simulation experiments were conducted on a simplified quay crane model of 132 m × 22 m × 70 m with 25 mission points. The results show that the proposed method reduced the average local path-searching time from 5.3392 s to 4.2907 s, corresponding to a 19.6% improvement over jump point search, while reducing the average local path length by 5.1%. The final global inspection path obtained in the experimental case was 516.1 m, which was shorter than those obtained by simulated annealing, genetic algorithm, particle swarm optimization, and the previous method. These results indicate that the proposed method can improve local planning efficiency and provide an effective inspection route for unmanned aerial vehicle-based quay crane inspection.

IPC Classification

Keywords