Solving the allocation problem of the airport parking stands

thumbnail

Tutor / Supervisor

Student

Busquets I Trias, Nil

Document type

Bachelor thesis

Date

2024

rights

Open AccessOpen Access

Publisher

Universitat Politècnica de Catalunya



Abstract

The stand allocation problem is a main problem in all airports and has to be solved every day without exception. The rapid increase in the air traffic and its continued growth during the last years has made this problem a major issue in all airports. This is mainly caused because in most of the cases the airports are not growing as fast as the air traffic. In addition, in the past years, airlines had tried to reduce the time that aircraft are on ground in order to maximize their benefits causing the stand allocation to become a major concern. This has produced the necessity of having a system capable of solving this problem taking into account different constraints and preferences that can vary on each airport or between the different airlines. In this project, two algorithms are designed and compared. The purpose is to see if it's possible to create a software capable of solving a stand allocation problem using one of this two algorithms. The algorithms that are used are a mixed-integer linear programming algorithm and a genetic algorithm. The development of the linear programming algorithm has been more or less straightforward due to the simplicity of the used library. In contrast, the implementation of the genetic algorithm has not been that easy because some changes on the basic functions of the algorithm had to be made. In order to do all these changes some easy problems were designed and tested increasing the complexity gradually in order to find the strong and weak points of the algorithm. After designing the algorithms and improving them, the algorithms were tested with different invented and real airport cases. The genetic algorithm was able to find solutions for all the cases. In contrast, the linear programming algorithm stuck in problems involving big amounts of data. It can be said that the initial objective has been accomplished for the genetic algorithm and, less successfully, for the linear programming algorithm.
user

Participating teacher

Files