Path planning of unmanned delivery vehicles based on GMACO algorithm in complex environments

Abstract In the logistics and transportation industry, unmanned delivery vehicles are gradually replacing manual labor to complete sorting, distribution, and other tasks. They can autonomously plan feasible routes to complete transportation tasks after acquiring environmental information via sensors. The ant colony algorithm, as a global path planning approach, has been widely applied in this field due to its strong global search and optimization capabilities. In this paper, an improved ant colony algorithm based on a guidance mechanism (GMACO) is proposed to address problems such as slow convergence, susceptibility to local optima, and complex parameter adjustment in the traditional algorithm. Firstly, the angle function is introduced to provide directional information and prevent large-scale blind searches. Secondly, strategies of inward pheromone expansion and pheromone concentration limitation prevent premature convergence and local optima while ensuring the quality of subsequent search paths. Finally, to streamline parameter adjustment, this paper introduces an adaptive pheromone volatilization factor and determines the optimal parameter combination through pre-experimental processing. The convergence and feasibility of the GMACO algorithm are demonstrated through mathematical theory. To meet diverse delivery requirements, this paper not only applies the GMACO algorithm to complex single-target environments for comparative simulation, but also integrates it with the traveling salesman problem (TSP) for multi-target delivery tasks. The simulation results demonstrate that the GMACO algorithm exhibits faster convergence, shorter path planning, and stronger adaptability, outperforming traditional methods in transportation tasks.