Rerouting airline passengers of canceled flights with genetic algorithms

Abstract

This study aims to explore and apply the genetic algorithm to the rerouting problem for passengers in the event of flight cancellations. Initially, the research approaches the problem through the maximum flow problem and then employs the genetic algorithm with various refinements in the population crossover and new population selection steps. Additionally, the study compares three crossover functions and two selection methods through 60 runs of the algorithm on a hypothetical scenario. The experimental results demonstrate that the genetic algorithm has the potential to effectively solve this problem with fast execution times, making it suitable for practical applications. Combining a flow-preserving crossover function with the FUSS selection method yields optimal results, maximizing the number of rerouted passengers, minimizing delays, reducing operational costs, and improving passenger experience