Research On Dynamic Scheduling Problem Of Fuel Truck Considering Task Balance

In order to improve the utilization rate of the airport's resources and the punctuality rate of flights,the efficiency of the ground task of the airport should be greatly improved.This topic studies the dynamic dispatch problem of airport refueling vehicles,that is,planning the driving path of refueling vehicles,so that the tankers can efficiently and orderly carry out the ever-arriving flight filling fuel tasks.During the route planning process,you need to know information such as how long your flight will take on the refueling task.The duration of the task is determined by the amount of fuel filled on the flight.However,in practice,the planned amount of fuel for the flight is sometimes quite different from the actual amount of fuel filled.So,one of the research topics is the prediction of the amount of fuel filled on flights.According to the fluctuation of the flight history filling fuel data series,the factors that cause the data series fluctuate are decomposed,and the LSTM-Prophet prediction model is constructed for the data characteristics,which establishes an independent model for the abnormal fluctuation point caused by the holiday factor.Using the historical flight fuel data of Shijiazhuang Airport as a study,the LSTM-Prophet weighted combination prediction model is compared to LSTM,Prophet and ARIMA models with better generalization capability.The second research focus is on the dynamic dispatch of airport refueling vehicles.By establishing a time slice,dynamic problems are divided into consecutive static problems.For each static problem,first,use the dynamic planning time window to capture the flight waiting to be filled with fuel.Secondly,a mathematical model of the problem of airport refueling truck scheduling is constructed for these flights.Then,the problem is solved by the adaptive branch pricing algorithm and the greedy algorithm proposed by the subject.The adaptive branch pricing algorithm uses the Dantzig-Wolfe decomposition principle to reconstruct the mathematical model of the airport refueling truck scheduling problem,and uses the dynamic column generation algorithm to solve the problem with arc-based branch strategy and optimize the target of the problem.The greedy algorithm adds specific heuristic conditions to the airport refueling truck scheduling problem,including evaluation function and the threshold for the end of the route.Finally,in order to verify the validity of the proposed algorithm,the real flight data of the airport is used as an experiment,compared with the existing ant colony algorithm,the proposed algorithm has some advantages in the optimization results of the target and the running time of the algorithm.The research content of the project can also be applied to other ground missions of theairport.