Research On UAV Intelligent Scheduling Method For Logistics Distribution

Logistics distribution generally refers to the distribution service which from the logistics center to the customers.It is the end link of the whole logistics,including delivery,takeout,medical drugs,emergency relief supplies and other distribution scenarios.In recent years,people have a higher requirements for the timeliness of logistics distribution.Nowadays,logistics distribution and delivery mainly adopt the way of the rider distribution or the vehicle distribution.However,with the increasingly complex urban environment,there usually occur multiple detours and the traffic congestion,these traditional methods often lead to delivery time prolonged and cannot meet the delivery timeliness requirements.With the rapid development and the function improvement of Unmanned Aerial Vehicle(UAV),its straight-line carrying capacity and avoid complex ground traffic's ability have become the focus of research institutions and related companies in the field of logistics and distribution.The existing UAV logistics distribution is mostly applied to the fixed two-point distribution.The UAV can only operate on the fixed route instead of automatically schedule with the mission changing.This way cannot tap the greater potential of the limited transportation resources.The main reason is the lack of mature UAV scheduling mathematical model and applicable intelligent scheduling algorithm.To our knowledge,there are very few mature and efficient UAV intelligent dispatching system.Therefore,it is necessary to study the UAV intelligent dispatching method in the logistics distribution scenario to enhance the utilization of transportation resources.This paper,firstly,analyzes the application scenarios of UAVs in the field of logistics distribution.It takes the total time cost of the order forms and the timeout penalty as the objective function,and constructs the UAV scheduling model with considering the constraints of UAV collision risk and airport state time window.Secondly,based on the existing single parent genetic algorithm,optimize and improve the coding design and gene operation stage which is aimed to be applied to the UAV scheduling model,and make spatial search with the optimal execution order of the order forms.Thirdly,a time window allocation strategy based on time slice search is developed.According to order of the task execution,the scheduling results of the unmanned aircraft which is executing each task and the specific execution time are obtained.The fitness function of the current task execution sequence is calculated and return to the improved single parent genetic algorithm for selection iteration.Then the actual take-away distribution data is selected for abstract analysis.The validity and feasibility of the model and improved algorithm established in this paper are verified in the simulation environment.Finally,the semi-physical simulation experiment platform for UAV scheduling is established,and simulating the real scene to model and add order dynamically.The data of pre-order no-legacy task,the pre-ordered legacy order and the multi-batch continuous input order are selected respectively to make the UAV dispatching flight experiment,and compared with the traditional prioritization method to verify the superiority of the improved algorithm.At the same time,compared with the rider's distribution method to prove the improvement of the timeliness of the UAV scheduling model and intelligent algorithm established in this paper in practical logistics distribution.