Planning And Scheduling: Using Genetic Algorithms To Solve Aircraft Fleet Assignment Problem

Abstract

Airlines face the complex task of optimizing aircraft fleet assignment to maximize profitability and efficiency. This thesis tackles this challenge by proposing a hybrid approach that combines Genetic Algorithms (GA) and Mixed-Integer Linear Programming (MILP). GA's strength lies in its ability to efficiently explore vast solution spaces, while MILP provides precise mathematical modeling and guarantees optimal solutions. The research develops a comprehensive mathematical model for the Aircraft Fleet Assignment Problem (AFAP) and evaluates it using both GA and CPLEX, a high-performance MILP solver, on small-scale instances. Results demonstrate GA's computational efficiency while achieving comparable optimality levels. Furthermore, a large-scale case study simulating a major North American airline showcases the effectiveness of the GA approach in generating significant profit improvements and cost reductions. This research provides airlines with a robust and efficient method to optimize fleet management, leading to enhanced operational performance and increased passenger satisfaction.