A probabilistic trajectory constraint modeler

Abstract

Trajectory predictors are core components of many air traffic applications. This includes cockpit, dispatch, flight planning, strategic fleet planning, air traffic control, traffic flow management, and aviation research applications. In all these different approaches, aircraft performance models are often required such as, for instance, drag polar aerodynamic data and maximum thrust propulsion data. For many applications these performance data may be required for a large number of different airframes and propulsion models. The emergence of new and non-traditional vehicle types, such as unmanned aerospace vehicles, further adds to and complicates this effort. Furthermore, the trajectory predictor software itself often is highly complicated. It has its own development and maintenance costs, and inevitably raises issues of portability, standardization and modeling consistency. Finally, two common drawbacks in typical trajectory predictors are that (i) they are deterministic and do not model the many uncertainties that impact flights in the national airspace system, and (ii) they often do not account for air traffic control constraints which flights are subjected to. Here I describe a Probabilistic Trajectory Constraint Modeler (PTM) that addresses these many issues, and stochastically models flight constraints and flight times. These outputs can be used by existing trajectory models or other decision support tools.