To meet traffic demand predictions, the global air traffic management (ATM) system needs to be changed. Several visions on future ATM operations exist. A commonality between the different visions is 4D Trajectory management. This function enables plan-based operation as opposed to the state-based approach of the present system. Plan-based operation enables the optimization of traffic flows by generating 4D trajectories. A part of the traffic flow generation process is scheduling.
The research presented in this thesis focuses on these scheduling opportunities. In this research the scheduling opportunities for departure traffic at a runway are investigated. A study of the existing literature showed that the most common scheduling algorithms currently available can be divided into four categories: first come first served, brand-and-bound, greedy search and genetic algorithms. A simulation environment is designed for evaluation of the departure scheduling algorithms using various input parameters like traffic situation, airport map and algorithm. The four algorithm categories are evaluated on output aspects like delay and robustness of the schedule and are compared with the current method of traffic scheduling.
The evaluation of the scheduling algorithms shows that the performance of the current method of scheduling departure traffic performs well in comparison with the tested algorithms. In case of no disturbances the genetic algorithm performs slightly better than the current method, but the other algorithms do not have a better performance. When disturbances are taken into account, a bigger performance increase can be obtained by using scheduling algorithms.