Hybrid Data-Driven and Physics-Based Flight Trajectory Prediction in Terminal Airspace

With the growing demand of air traffic, it becomes more important and critical than ever to develop advanced techniques to control and monitor air traffic in terms of safety and efficiency. Especially, trajectory prediction can play a significant role on the improvement of the safety and efficiency because predicted trajectory information is used for air traffic management such as conflict detection and resolution, sequencing and scheduling. Recently, there have been extensive efforts for the development of trajectory prediction algorithms, which can be categorized into two approaches: (i) in physics-based approaches, a model describing the behaviors of an aircraft is developed based on the aircraft dynamics or governing physics, which is then used in Kalman filtering or its variants, and (ii) in data-driven approaches, collected flight data is used to learn a data-driven model that can be used to predict the future trajectory of an aircraft. Both of the approaches, however, have limitations: Without assistance of dataset, the physics-based approaches use only the aircraft dynamics for its trajectory prediction without correction by the measurement since no measurements will be available in the future; on the other hand, the data-driven methods do not explicitly use the aircraft dynamics at the current time (e.g., an aircraft is performing a coordinated turn). In this work, we propose a new framework that can overcome these limitations by integrating the two methods, called hybrid data-driven and physics-based trajectory prediction. The proposed algorithm is applied to real air traffic surveillance data to demonstrate its performance. Results show that our new algorithm has a higher trajectory predicting accuracy than the two baseline methods, which could help enhance the safety and efficiency of air traffic operations