Anti-windup attitude control for UAVs based on exponential coordinates

This paper investigates the manoeuvring capabilities of UAVs with a focus on addressing the challenges posed by kinematic representation singularities and actuator saturation. Initially, it introduces a novel approach where the attitude kinematics of UAVs are represented on a configuration manifold, employing exponential coordinates to describe their evolution on the tangent space of this manifold. Subsequently, the paper develops a control model based on the UAV's attitude model with exponential coordinates, followed by the creation of an inverse model. Leveraging this, a state feedback controller using PCH (pseudo control hedging) is devised. The simulation results underscore the significance of this work, demonstrating that the designed controller effectively enables UAVs to sharply and accurately track desired angles while mitigating performance degradation caused by actuator saturation.