Analysis of alternative flat representations of a UAV for trajectory generation and tracking

Abstract

Motion planning problems benefit greatly from the properties of differential flatness, which are widely employed for trajectory generation and controller design. This paper aims to highlight the fact that various flat representations of a system can have different implications in the trajectory generation and tracking objectives. In particular, we consider a fixedwing UAV (Unmanned Aerial Vehicle) and analyze various flat representations. We reformulate the trajectory generation and tracking problem in terms of different flat outputs and analyze the optimal cost, constraints satisfaction, tracking error and computational complexity. Insights on the future work complete the analysis. The research shows promising directions, especially in the area of disturbance rejection and robust control.