Research On Optimization Of ’truck And Uav’ Collaborative Delivery Path Considering The Impact Of COVID-19

In January 2020,the COVID-19 broke out,endangering the lives and health of the people.In order to reduce the spread of the virus among the population and control the epidemic as soon as possible,China adopts the method of sealing and isolating epidemic gathering areas,and divides them into low,medium,and high risk areas based on the actual situation to avoid cross infection caused by the movement of personnel and vehicles.Reducing human contact also blocks the normal operation of vehicles.In order to ensure the timely supply of materials to the people who have been in closed isolation for a long time,drones with the advantages of being unaffected by geographical and environmental conditions and fast flight speed bring new opportunities to the distribution of materials in medium to high-risk areas.In this context,the development prospects of unmanned delivery have been broadened,and new delivery models mainly based on drones have gradually entered the research field.In order to provide data sources for case solving to classify risk levels,this paper first takes the reality of the COVID-19 as the background,considers the latency and infectivity of the virus,and constructs an epidemic diffusion model considering the infection rate.The model parameters are reasonably estimated using the least square method,and the prediction solution uses the double simulated annealing algorithm.Use this model to fit and predict the infection of COVID-19 in Shiyan City,Xiangyang City and other regions in Hubei Province.With February 2020 as the fitting date and March1 to March 10 as the prediction date,verify the model fitting effect,and provide data basis for the division of risk level in the region where the customer point is located.That is,according to the different infection prediction in the first 14 days of a day,the region where the customer point is located can be divided into low,medium High risk areas.Then,considering the risk level of the epidemic,drones must fly to medium to high risk areas for delivery.A collaborative delivery model between vehicles and drones is constructed,with the objective function of minimizing delivery time.For the constructed model,a hybrid particle swarm optimization genetic algorithm based on fitness variance is designed.The coding adopts floating point coding.When the variance change of fitness of continuous iterations is less than a certain threshold,the introduction of genetic algorithm can ensure the continuity of excellent particles and jump out of the local optimal solution.Finally,taking the epidemic situation in Hubei Province in early 2020 as a design case,the predicted COVID-19 infection situation was reasonably divided into low,medium and high risk areas according to the relevant policy requirements for the region where the customer points are located,and then the constructed vehicle and UAV collaborative distribution model was substituted,and the hybrid particle swarm optimization genetic algorithm was used to solve the path optimization.Sensitivity analysis was conducted by adjusting relevant parameters,focusing on the dynamic impact of the epidemic and the impact of drone flight speed on delivery time.The results show that it is feasible and efficient to use unmanned aerial vehicles and vehicles for collaborative delivery when considering the level of epidemic risk.The hybrid particle swarm genetic algorithm has better convergence than particle swarm algorithm or genetic algorithm in solving such problems.The introduction of drones can not only provide timely access to materials for people in medium to high risk areas who are under lockdown and isolation,but also shorten delivery time,reduce the risk of cross infection,and meet the different needs of different regions.It is of great significance for material distribution during the epidemic.