A study on path optimization of construction machinery by fusing ant colony optimization and artificial potential field

Abstract

With the progress of manufacturing capability and the requirements of modern project construction clustering and safety, unmanned technology has become a major development direction for construction machinery. The study analyzes the defects of the ACO applied to path planning, uses the basic ant colony method combined with the APF algorithm to construct comprehensive heuristic information to improve the efficiency of the initial iteration, and adds a pseudo-random proportional transfer rule to improve its transfer rule, in addition to proposing an optimal ant pheromone release rule to ensure that the pheromone update can effectively improve the effect of the later iteration. In the 20 × 20 raster map, the optimal path of both the basic and potential field ant colony algorithms is 31.52, while the basic ant colony algorithm does not converge to the optimal path after 200 iterations in the 30 × 30 raster map, and the comparison algorithm converges to the optimal path after 156 iterations, while the potential field ant colony algorithm converges to the optimal path after 25 iterations at 47.52. And the potential field ant colony The potential field ant colony algorithm converges to the optimal path of 47.52 after 25 iterations, and the number of turns in both raster maps of different sizes is only about 50% of the basic ant colony algorithm, which is also better than that of the comparison algorithm. The proposed algorithm improves the ability of the ant colony algorithm to explore and exploit raster maps, enhances its path planning capability, and improves the construction efficiency well in practical applications.