Logistics drone scheduling model based on delivery service modes and routing strategies

REN Xinhui; RUI Yuqi

doi:10.11860/j.issn.1673-0291.20230083

Abstract：This study addresses the challenges of long-range urban logistics via Unmanned Aerial Ve-hicle (UAV) delivery by introducing a two-level multi-hub UAV delivery service model. By compre-hensively taking into account factors such as customer time preferences, UAV recharging necessities,and delivery routing strategies, a two-level multi-hub urban UAV delivery scheduling model is estab-lished to minimize aggregate delivery expenses. Solution methodologies are meticulously evaluated through benchmark case studies, comparing the efficacy of Cplex optimization software with heuristics including the Whale Optimization Algorithm (WOA), Genetic Algorithm (GA), and Grey Wolf Opti-mization (GWO). Subsequently, a two-phase algorithmic resolution strategy is devised. The efficacy of the model is exemplified through its application in Tianjin's central urban area, accompanied by sensitivity analyses meticulously examining three pivotal facts: demand node scale, UAV types, and time window constraints. Results reveal that compared with closed routing strategy, the semi-open routing strategy achieves savings of approximately 11.09％ in total delivery costs and 60.45％ in time window penalty costs, with cost savings increasing with delivery scale. Compared with on-demand models, the proposed fixed-route schedule mode demonstrates over 18％ reductions in both variable costs and overall costs, alongside a nearly 28％ decrease in required UAV numbers. Temporal-spatial diagrams illustrating flight schedules and routes highlight the advantages and disadvantages of different approaches. This model offers practical insights for refining long-range urban UAV logistics scheduling schemes and delivery patterns, providing valuable guidance for UAV-based logistics planning for enterprises.