A dynamic rerouting model for air traffic flow management

• Published in: Transportation research. Part B: methodological Vol. 43; no. 1; pp. 159 - 171
• Main Authors: Mukherjee, Avijit, Hansen, Mark
• Format: Journal Article
• Language: English
• Published: Kidlington Elsevier Ltd 2009; Elsevier
• Series: Transportation Research Part B: Methodological
• Subjects: Air traffic management; Air traffic management Stochastic optimization Scenario tree Collaborative decision making Ground holding Rerouting Convective weather Integer programming Capacity constraints Aircraft routing Traffic flow management Airspace flow program; Air transportation and traffic; Aircraft routing; Airspace flow program; Applied sciences; Capacity constraints; Collaborative decision making; Convective weather; Exact sciences and technology; Ground holding; Ground, air and sea transportation, marine construction; Integer programming; Rerouting; Scenario tree; Stochastic optimization; Traffic flow management; Transportation planning, management and economics; Collaborative decision making; Ground holding; Integer programming; Aircraft routing; Scenario tree; Convective weather; Traffic flow management; Airspace flow program; Air traffic management; Rerouting; Stochastic optimization; Capacity constraints; Performance evaluation; Modeling; Optimization; Traffic flow; Forwarding; Dynamic model; Script; Decision making; Routing; Weather; Stochastic analysis; Traffic management; Air traffic; Air transportation; Aircraft; Comparative study
• ISSN: 0191-2615; 1879-2367
• DOI: 10.1016/j.trb.2008.05.011

In this paper, we present a stochastic integer programming model for managing air traffic inbound to an airport when both the airport itself and its approach routes are subject to adverse weather. In the model, ground delay decisions are static, while those on rerouting are dynamic. The decision variables in the model are aggregate number of flights planned to arrive at various capacity constrained resources. The model does not directly assign arrival times to individual flights. Therefore, in context of Collaborative Decision Making, which is the governing philosophy of the air traffic management system of the United States, the solutions from the dynamic rerouting model can be directly fed to some resource allocation algorithm that assigns routes and release times to individual flights or to the airlines who operate them. When adverse weather blocks or severely limits capacity of terminal approach routes, rerouting flights onto other approaches yields substantial benefits by alleviating high ground delays. Our experimental results indicate that making rerouting decisions dynamically results in 10–15% delay cost reduction compared to static rerouting, and about 50% less delay cost compared to a “pure” ground holding strategy (i.e., no rerouting). In contrast to static rerouting, the dynamic rerouting capability results in making rerouting decisions that are better matched to realized weather conditions. Lower total expected delay cost is achieved by delaying the rerouting decisions for flights until they reach the divergence point between alternative routes, and hence exploiting updated information on weather while making those decisions. In cases where the airport is the main, but not the only, bottleneck, the dynamic rerouting model may assign higher ground delays so that the rerouting decisions can be deferred until more information on en route weather becomes available.