Precise aircraft trajectory prediction is a critical issue in civil aviation. Traditional dynamics analysis using the classical point mass model often encounters deviations due to inaccuracies in aircraft mass estimation, which can ultimately undermine the validity of the energy conservation equation. This study introduces a novel approach to enhance trajectory prediction accuracy by integrating the point mass model with a closed control loop and artificial intelligence techniques, aimed at reducing energy deviations. An ensemble learning method was employed and fitted to real‐world flight data, allowing for dynamic estimation of aircraft mass based on current flight information. The proposed method was rigorously evaluated with data from eight common aircraft types, achieving accuracy improvements of 56% for mass, 44% for altitude, and 11% for speed.
Zhao et al. (Thu,) studied this question.