En-route arrival time prediction via locally weighted linear regression and interpolation

Abstract

In this paper, a data-driven method for predicting the arrival time of en route aircraft is proposed. The proposed model predicts the arrival time of aircraft to a target waypoint based on the current state values of its longitude, latitude, speed, and heading. A combined method using regression analysis and Barycentric interpolation was applied in order to establish the model. As a case study, the proposed model was applied to the en-route trajectories passing over Incheon International Airport, and the prediction accuracy for different look-ahead times was studied.