A novel decomposition approach for holistic airline optimization

Airlines face many different planning processes until the day of operation. These include Fleet Assignment, Tail Assignment and the associated control of ground processes between consecutive flights, called Turnaround Handling. All of these planning problems have in common that they often need to be reoptimized on the day of execution due to unplanned events. In many cases, this is still done manually with the expertise of airline operators in the airline operations center. In order to automate this process and to be able to find the best possible reoptimization solution, the medium-term aircraft assignments and the short-term plannable turnaround processes should be optimized in an integrated way. For this purpose we provide a new integrated mixed integer program, which combines Fleet Assignment, Tail Assignment and Turnaround Handling. Due to the size of the model, it cannot be solved efficiently as a complete model. Therefore, we develop a new decomposition algorithm that alternately solves the combined assignment problems and the turnaround model, and projects the turnaround costs into the objective function of the combined assignment model. In a computational study with realistic airline data, it is shown that for many problem instances this method yields feasible and optimal solutions much faster than comparable benchmark algorithms.





View A novel decomposition approach for holistic airline optimization