Improving trajectory prediction through understanding uncertainty

ATM is gradually moving towards the notion of allowing aircraft to fly their preferred trajectory, otherwise known as trajectory-based operations.  One of the challenges related to the implementation of TBO is the ability to identify, model and manage the uncertainty associated to a trajectory.

The integration of the uncertainty models in the planning systems improves the trajectory predictions and supports the assessment of the feasibility of integrating the models into existing demand and capacity balancing (DCB) tools. The COPTRA project researched three areas related to uncertainty modelling:

• Defining and assessing probabilistic trajectories in a TBO environment;
• Combining probabilistic trajectories to build probabilistic traffic prediction;
• Applying probabilistic traffic prediction to air traffic control planning.

COPTRA showed that in addition to quantifying uncertainty through data analytics, it is possible to limit it through model-driven state estimation techniques. This enables not only to include flight intent or initial condition uncertainties but also to take into account model uncertainties.

COPTRA's models provide us with a clear quantitative understanding of delay propagation dynamics in space and time. The project results provide insight into how to achieve more efficient ATM operations in the future.

Benefits

Visualisation of uncertainty

More accurate and stable demand prediction

Quantitative understanding of delay propagation dynamics in space and time