Genetic Algorithm for Optimal Aircraft Routing

Airlines take the extra effort required to figure out what is the best route that their aircrafts should take so as to minimize additional costs from non-revenue flights or idle time. One way to di this is to first generate large numbers of feasible routes and then assign flights to a subset of them so as to cover all flight legs. It is clear that the quality of the resulting solution depends highly on both the number of routes we generate anf also the diversity among the routes. Classical deterministic approaches to the first phase are slow; especially when the number of flights increases. This project aims to maximize these factors in a quick and efficient manner by developing a Genetic Alogrithm. The stochastic nature of this algorithm will speed up the route generation process and the GA operators used will ensure high quality and diverse routes. Bombardier Inc., can offer a valuable service to its clients by reducing the cost of non-revenue flights and providing the opportunity to optimally stagger individual aircraft utilization to smooth out the cost of major maintenance events.

Intern: 
Vipin Mathew
Faculty Supervisor: 
Dr. Kamran Behdinan
Project Year: 
2014
Province: 
Ontario
Program: