Mixed H2/H∞ Control for Autonomous Quadrotor Landing via Linear Matrix Inequality Optimization

Landing a quadrotor on a mobile platform with unknown trajectories poses significant challenges, particularly when the platform is noncooperative. This study presents a vision-based autonomous landing architecture for a quadrotor and an unmanned ground vehicle, integrating a mixed [Formula: see text] control framework with a linear matrix inequality approach. To achieve the proposed controller, first, experimental data are used to encapsulate multiplicative uncertainty, which represents the difference between the experimental response of the real quadcopter and the response of the modeled system. The proposed control methodology combines [Formula: see text] and [Formula: see text] control to achieve precise landings and robust disturbance rejection, using linear matrix inequalities for efficient controller synthesis. To validate the algorithm, several software-in-the-loop simulations were implemented to demonstrate tracking and vertical landing performance under different disturbance conditions while considering actuator dynamics. The proposed mixed [Formula: see text] controller significantly outperforms the traditional Proportional–integral–derivative (PID) controller in both landing time and accuracy across various disturbance conditions, achieving up to 82.89% improvement in landing time under a [Formula: see text] wind disturbance and 56.6% improvement in landing accuracy under undisturbed conditions.