Resumen de A data-driven approach for dynamic and adaptative aircraft trajectory prediction
Álvaro Rodríguez Sanz, José Manuel Cordero García, Icíar García Ovies-Carro, Enrique Iglesias
Traffic Prediction (TP) is a key element in Air Traffic Management (ATM), as it plays a fundamental role in adjusting capacity and available resources to current demand, as well as in helping detect and solve potential conflicts. Moreover, the future implementation of the Trajectory Based Operations (TBO) concept will impose on aircraft the compliance of very accurately arrival times over designated points. In this sense, an improvement in TP aims at enabling an efficient management of the expected increase in air traffic strategically, with tactical interventions only as a last resort. To achieve this objective, the ATM system needs tools to support traffic and trajectory management functions, such as strategic planning, trajectory negotiation and collaborative de-confliction. In all of these tasks, trajectory and traffic prediction represents a cornerstone. The problem of achieving an accurate and reliable trajectory and traffic prediction has been tackled through different methodologies, with different levels of complexity. There are two main aspects to be considered when assessing the most appropriate forecasting methodology: (a) time- horizon: depending on the timescale (anticipation before the day of operations), the level of uncertainty associated to the prediction will be different; and (b) input data: both the source and the quality of the input data (completeness, validity, accuracy, consistency, availability and timeliness) are key characteristics when assessing the viability of the prediction. This study develops a methodology for TP and traffic forecasting in a pre-tactical phase (one day to six days before the day of operations), when few or no flight plans are available.
This should be adjusted to different time scales (planning horizons), taking into account the level of predictability of each of them. We propose a data-driven, dynamic and adaptive TP framework, which can be accommodated to different Airspace Users’ characteristics and strategies.
