Pre-Tactical Planning of Runway Utilization Under Uncertainty: Optimization and Validation

Abstract

Efficient planning of runway utilization is one of the main challenges in Air Traffic Management (ATM). In a previous paper, we developed a specific optimization approach for the pre-tactical planning phase that reduces complexity by omitting unnecessary information. Further, we investigated the impact of disturbances on our solutions, since in reality uncertainty and inaccuracy almost always lead to deviations from actual plans. In this paper, we now present approaches to incorporate uncertainty directly in our model in order to achieve a stabilization with respect to changes in the data. Namely, we use techniques from robust optimization and stochastic optimization. Further, we analyze real-world data from a large German airport to obtain realistic delay distributions, which turn out to be two-parametric Γ-distributions. We describe a simulation environment and test our new solution methods against standard algorithms (e.g., First-Come-First-Serve). The encouraging results show that our approaches significantly reduce the number of necessary replan- nings.