Modeling Uncertainties in Intents, Guidance, and Pilot Actions for Reliable Trajectory Synthesis

Abstract

This paper presents a systematic framework for modeling intents, guidance strategies, and pilot actions, as well as their uncertainties for advanced trajectory synthesis. At a given time, three independent tactical intents are needed to completely specify flight objectives in the longitudinal, vertical, and lateral directions. In this paper, likely intent variables are classified into six groups. A complete set of tactical intents can be uniquely defined through group specification, variable selection, and value definition. To define strategic intents, a trajectory window is divided into a series of tactical intent segments that are connected through appropriate capture conditions. In comparison, guidance strategies are defined that assign trajectory controls to tactical intents at each segment. Sources of uncertainties in intents and guidance strategies are discussed. Next, pilot actions are described using manual control and feedback linearization, and variations of pilot actions among a pilot population are examined. Monte Carlo simulations of aircraft flights are used to evaluate the impacts of uncertainties in pilot actions, intent variable types, guidance strategies, and intent variable values on both nominal trajectories and variations of predicted trajectories in the presence of uncertainties. In particular, effects of different guidance strategies on the tradeoffs between dispersions of time of arrival and fuel burn at a specified position are studied.