Research On Stands Allocation Algorithms Of Overnight Flight

The allocation of overnight flights is an important part of the airport’s allocation of overnight flights.A good allocation result of overnight flights can improve the airport’s operational efficiency,economic benefits and passenger satisfaction.This thesis establishes a multi-time slot allocation model for overnight flights.The heuristic neighborhood search method is used to solve the problem.The algorithm can obtain a good allocation result under various airport constraints and multiple optimization objectives,which greatly improves the automation degree and allocation quality of airports.In this thesis,a multi-time slot model is deployed to model the allocation of overnight flights.The time of flights at the airport is divided into three timeslots according to the arrival,overnight and departure activities.The stands are allocated to each timeslot and Allocate the boarding gate required for the flight at the same time.The target is to optimize the berth ratio,towing cost and preference of boarding gate etc.The constraints are the size of the slots,the simultaneous exclusiveness of the slots and the logic of towing etc.In order to solve this model,the problem of stands allocation is formulated as a constrained satisfaction optimization problem.The overnight flight’s stands allocation is organized as a constraint satisfying optimization problem,the constraint is expressed as a hard constraint,and the optimization objective is expressed as a soft constraint.For the multi-objective seat allocation problem,a weight value is given for each objective by analytic hierarchy process.The multi-objective problem is transformed into a single-objective optimization problem by linear weighting,and heuristic formulas such as Tabu search and Late acceptance are used.Neighborhood search is used to solve the constraint satisfaction optimization problem.Taking the real data of Urumqi International Airport as input data,the results of various heuristic neighborhood search methods are compared and compared with the results of manual scheduling in the same period.The robustness of overnight flight slot allocation schemes is studied.It is considered that the arrival and departure delays of flights are independent and obey a lognormal distribution.The robustness of the allocation results is measured by Monte Carlo method.Two commonly strategies to improve the robustness of slot allocation are compared by simulation,that is,introducing fixed buffer time and optimizing idle time distribution,and the method of fixed buffer time is more stable.