Research On Collaboration Delivery By Vehicles And UAVs In The Last Mile

As online shopping seeps into People’s Daily life,the last mile of logistics distribution,as the final link of consumption,is crucial to improve the service quality of logistics distribution and customer service experience.Unmanned aerial vehicle(UAV)become the last mile discovery of a new research field,it has the flight speed,high efficiency,low transportation cost advantages,the vehicle has the advantages of the long distance transportation,vehicles and drones complement each other.Therefore,this paper proposes a cooperative delivery scheme between vehicle and UAV to achieve better and faster logistics distribution service.First,the last mile logistics as the research background,the advantages and disadvantages of the existing distribution mode is compared and this paper puts forward to improve customer satisfaction with logistics distribution solutions,namely vehicles and drones cooperative delivery,through the analysis of the advantages and algorithms of drones delivery,further to the development of vehicles and drones cooperative delivery provides a theoretical basis.On this basis,two kinds of cooperative delivery modes and structures are designed: vehicles carrying a single drone cooperative delivery,vehicles carrying multiple drones cooperative delivery.Secondly,the collaborative delivery route optimization between vehicles and UAVs is studied,considering the time and space cooperation between drones and vehicles as well as the constraints of drone carrying capacity,to minimize operating costs as the optimization objective,two mixed integer programming models are constructed.When solving the problem of vehicles routing problems with time windows with a drone(VRPTW-D),GA-VNS is designed and the CPLEX solver is used to solve the small-scale case accurately,and the scheme of individual vehicle distribution is compared and analyzed to prove the validity of the model.When solving the problem of vehicles routing problems with time windows with drones(VRPTWD),two-phase algorithm is designed,the improved K-means clustering algorithm is used to find the vehicle stopping point,PSO-VNS is used to optimize the vehicles and drones routing.The traditional K-means algorithm and CPLEX are used to solve the smallscale calculation example accurately,and the vehicle individual distribution scheme is compared and analyzed.The experiment shows that the model and algorithm are feasible.Finally,combined with the example experiment,the cost types and cost sources of each mode are analyzed by comparing the two modes of single vehicle delivery and collaborative vehicle and UAV delivery,which further verifies that drones participate in vehicle delivery can effectively reduce the total cost.Through sensitivity analysis experiment,the influence of variation of drone carrying weight and maximum flight distance on operating cost and delivery mode is analyzed.It is concluded that vehicles carrying a single drone mode is suitable for the situation where demand points are scattered,while vehicles carrying multiple drones mode is suitable for the situation where demand points are densely distributed and UAV can deliver more customers.It provides reference for the practical application of the cooperative delivery mode of vehicle-drone in the last mile.