Traffic Flow Management (TFM) implements traffic management initiatives (TMIs) such as Ground Delay Programs (GDPs) and TFM reroutes in order to manage airport and airspace-sector traffic demand. TMIs are not necessarily optimal in terms of minimizing overall delays because they are often designed with a degree of conservativeness to account for the uncertainty in traffic demand and capacity forecasts. Moreover, in the current manual TMI design process each TMI is typically designed to address only one congestion event independent of other disparate congestion events that might be affecting the same set of flights. Researchers have developed TFM optimization algorithms to minimize National Airspace System (NAS)-wide TFM delays while ensuring that all NAS-wide congestion events are addressed. The Bertsimas Stock-Patterson (BSP) algorithm is one example of such an approach. Developing methods for adapting research TFM algorithms into practical TFM decision support aids is the topic of this research paper. Mainly, we focus on two modifications to the BSP algorithm to alleviate some of its undesirable (from the perspective of real-world implementation) characteristics and to make its controls practically implementable in a form similar to current-day TMIs. First, we modify the original BSP formulation to account explicitly for uncertainty within the BSP optimization. We call this the Probabilistic-BSP (P-BSP) formulation. P-BSP alleviates the original BSP's sensitivity to traffic-demand/capacity forecasts and provides the robustness necessary in real-time applications. Second, we modify the BSP formulation to enhance the coordination between strategic TFM airborne and ground delays. This modification reduces unnecessary airborne delays caused by the lack of coordination in current-day TMIs. We call this formulation the Coordinated BSP (C-BSP). We present results from simulations showing the benefits of the P-BSP formulation as compared to the original BSP using a real-day, NAS-wide traffic and capacity scenario. We are currently investigating the benefits of the C-BSP formulation.
展开▼