Accuracy considerations of a simple aircraft trajectory prediction model for idle thrust descents

Abstract

This paper addresses the suitability of simplified models for fast (real-time) trajectory prediction during idle-thrust descents and discusses a technique for quantifying the state uncertainty using parametric models for the aircraft initial position and forecast weather uncertainties. Results suggest that during the typical idle-thrust descent, model simplifications are only useful for short look-ahead projections, as the bias error accumulation quickly exceeds conservative error bounds, necessitating the use of higher fidelity models. Additionally, uncertainty-modelling results indicate that the prediction spreads can be quantified with sufficient accuracy using knowledge of the expected input parameters accuracy.