American Institute of Aeronautics and Astronautics 1 Air Traffic Estimation and Decision Support for Stochastic Flow Management Prasenjit Sengupta * , Monish D. Tandale * , Victor H. L. Cheng † , and P. K. Menon ‡ Optimal Synthesis Inc., Los Altos, CA 94022-2777 Development of a decision support system that uses real-time track data to estimate statistical parameters describing the stochastic traffic flow is described. Modern statistical decision theory is applied to optimize traffic flow. An advanced estimation algorithm provides the parameter estimates based on queuing network models of traffic flow. A hypothesis testing approach is developed for triggering traffic flow management initiatives in the terminal area, and a stochastic quadratic programming methodology is advanced to achieve flow control objectives such as runway load balancing. The use of this methodology is demonstrated using multi-day track data in the San Francisco terminal area. It is shown that the methodology can correctly identify the need for restricting the traffic flow into the terminal area, and provide decision support to balance the traffic flow at the runways under uncertain traffic flow conditions. The present approach can be extended to the creation of decision support tools for a wide variety of stochastic air traffic flow control situations. I. Introduction everal research efforts are underway at NASA on air traffic flow management (TFM) using advanced iterative numerical algorithms 1-8 . These algorithms are being considered not only for strategic TFM in the National Airspace System (NAS), but also in managing surface traffic flows. While these algorithms can provide precise solutions to the traffic management problem, they are more suitable for predictive control based on deterministic data. Traffic flow control in the presence of uncertainties inherent in the air traffic management system requires the integration of estimation algorithms to derive the stochastic description of traffic flow, followed by the application Statistical Decision Theory 9 to either trigger iterative numerical algorithms or to create stochastic flow control decisions. The focus of the present research is on stochastic traffic flow control and resource management in the terminal area and the runways. Under normal circumstances, the aircraft are allowed to operate according to their intent, and no major traffic flow management initiatives are necessary to ensure smooth operation. However, various perturbing influences such as adverse weather require traffic flow management initiatives to be brought into effect to match the available capacity with the demand. The objective of TFM algorithms on the surface as well as in the air is to mitigate the impact of these perturbing factors on traffic flow before they actually become disruptive to the NAS operations. TFM is generally initiated upon the detection of adverse events in the air traffic environment. For instance, low visibility conditions may restrict simultaneous operations on closely-spaced parallel runways at airports such as San Francisco. Once adverse traffic flow conditions are predicted to occur, the TFM algorithm can be initiated to mitigate them. Since the predictions are generally based on uncertain data, it is important to rule-out traffic flow variations cased by minor flow transients or due to the naturally occurring variability in the traffic. Consequently, decisions to initiate traffic flow management must explicitly recognize the stochastic nature of traffic flow. Since the uncertainties in the system cannot be precisely predicted, the traffic flow parameters must be estimated from actual measurements, followed by the application of methods from Statistical Decision Theory 9 to create actionable decisions. * Research Scientist, 95 First Street Suite 240, Senior Member AIAA. † Vice President, 95 First Street Suite 240, Associate Fellow AIAA. ‡ Chief Scientist and President, 95 First Street Suite 240, Fellow AIAA. S AIAA Guidance, Navigation, and Control Conference 08 - 11 August 2011, Portland, Oregon AIAA 2011-6515 Copyright © 2011 by Optimal Synthesis Inc. Published by the American Institute of Aeronautics and Astronautics, Inc., with permission.