Three-learning strategy particle swarm optimization for air-ground collaborative logistics transportation scheduling problem with pickup and delivery considering customer priorities.

Based on the air-ground collaborative logistics distribution model using UAVs and vehicles, this paper addresses the logistics scenario of delivering before picking up. Considering factors such as customer priority, vehicle cost, and UAV cost, with the objective of minimizing the total travel cost, we propose the Air-ground Collaborative Logistics Transportation Scheduling Problem with Pickup and Delivery Considering Customer Priority (ALTSPPDCP). Based on the characteristics of the model, the use of heuristic algorithms can efficiently solve such optimization problems, avoiding the time-consuming exhaustive search and improving the quality of the solutions. This paper designs a multi-layer, multi-stage encoding and decoding strategy based on the Three-Learning Strategy Particle Swarm Optimization algorithm, integrating the ascending order sorting method and dynamic segmentation method to transform the particle space into the model space. An intelligence optimization algorithm for solving ALTSPPDCP is proposed. Finally, In the 50-node scenario of the model comparison, the vehicle-UAV schema achieved a total cost that was 14.86% lower than that of the vehicle-only schema. In the algorithm comparison experiment, the optimal solution obtained by TSLPSO reduced costs by 39.99% and 27.94% compared to PSO and RPSO, respectively.