Research On Ground Holding Policy For Flights’ Arrival And Departure Under Uncertain Capacity

Airport capacity change is one of the important factors leading to flight delays.Among the many traffic management methods,the ground holding policy is an effective means of traffic management.In order to convert the high-cost air-waiting into the low-cost ground-based waiting,it needs to redistribute the time slots for the flights that involved in the case of uncertain capacity.Due to the different types of flights and the different tasks performed by some flights,different flights have different degrees of sensitivity to time,which leads to different degrees of influence corresponding to different flight adjustments at the same time.At the same time,the cost of different flight delay is also different,the delay caused by different time losses are also different.How to allocate time slots for flights with capacity uncertainty makes the combination of time cost and other delay costs reach an optimal state is an main issue of the study.In order to solve the above problems,this paper takes the ground holding policy in air traffic flow management as the main research object,and studies the issue of time slot allocation for flight arrival and departure under uncertain capacity conditions,and establishes a two-stage stochastic programming model with recourse.At the same time,three recourse functions have been established-"simple recourse","time link recourse",and "equivalent flight recourse".The algorithm has been studied in depth.First,a general algorithm was used to solve the model.The optimal solution was simulated using simulation data and the results were analyzed.Then the existing problems of the model are improved.The model is solved by genetic algorithm.Two corresponding evaluation indexes are established-“deterministic comprehensive cost” and “uncertain comprehensive cost”.The simulation data is used to analyze the model.Simulation verification has achieved the desired results.Finally,two applications were developed for the above two models,and the interfaces were carefully designed to implement a multi-functional interface integrating input,output,and data analysis.