Accuracy evaluation of a new generic Trajectory Prediction model for Unmanned Aerial Vehicles

Abstract Unmanned Aerial Vehicles (UAVs) attract much attention, and they require trajectory information for planning and tactical operations. Based on the current trajectory data obtained by navigation systems, appropriate Trajectory Prediction (TP) models are required to determine future UAV trajectories. Regardless of the disparity of the models, the knowledge of TP accuracy and its performance evaluation are of paramount importance. After the review of extensive literature on current TP models, this paper develops a more accurate TP model by creating valid methodologies for error budgeting and mitigation. This is followed by the specification for quantifying TP accuracy and sub-functional testing employed in the process of model development. The new model is tested and validated by three credible case studies in relation to the most stringent UAV applications. The validation evidence demonstrates that the new model improves TP accuracy based on previous models.