Truck-drone collaborative delivery method based on three-stage adaptive ant colony optimization

Abstract

To address the inefficiency of “last-mile” delivery in urban and rural logistics caused by road condition constraints, this paper proposes a collaborative truck-drone routing optimization scheme. First, a mixed-integer programming model is constructed with the objective of minimizing the total delivery time, comprehensively considering physical constraints such as drone endurance, payload capacity, and spatial-temporal synchronization of both vehicles at rendezvous points. Second, to overcome the defect of the traditional ant colony algorithm falling into local optima when solving complex routes, an improved ant colony algorithm based on a three-stage adaptive mechanism is designed. This algorithm divides the iteration process into exploration, balance, and convergence stages, and coordinates with dynamic parameter adjustment and a dedicated route decoder to achieve efficient task allocation between the truck and drone. Finally, simulation experiments based on the modified Solomon benchmark show that at a 50-node scale, the total delivery time of the proposed TS-ACO algorithm is reduced by 13.9% compared to the traditional ant colony algorithm. Compared with the truck-only delivery mode, the collaborative delivery mode can save 28.4% of the total time and reduce the truck driving distance by 38.1%. This study provides a practical theoretical model and solution tool for the optimization of terminal logistics networks in complex terrains.