Explicit model predictive control and L1-navigation strategies for fixed-wing UAV path tracking

A control strategy for fixed-wing Unmanned Aerial Vehicles is proposed and relies on the combination of linear model predictive control laws for the attitude dynamics of the system, along with an implementation of the L₁-navigation logic that provides attitude reference commands to achieve precise path tracking. The employed predictive controllers ensure the performance characteristics of the critical attitude loops, while respecting the actuation limitations of the platform along with safety considerations encoded as state constraints. Being explicitly computed, these strategies are computationally lightweight and allow for seamless integration on the onboard avionics. Once the desired attitude response characteristics are achieved, tuning the cascaded nonlinear L₁-navigation law becomes straightforward as lateral acceleration references can be precisely tracked. A wide set of experiments was conducted in order to evaluate the performance of the proposed strategies. As shown high quality tracking results are achieved.