Performance and Robustness Assessment for a Port-Hamiltonian Input-to-State Stabilizing Flight Controller

This paper presents a flight control law that leverages the structure of fixed-wing aircraft flight dynamics to obtain input-to-state stability, impose a port-Hamiltonian structure on the closed-loop system, and ensure that the closed-loop system is robust to disturbances and model uncertainty. The control approach is compared with two alternatives -- linear-quadratic optimal control and nonlinear dynamic inversion -- to illustrate the practical advantages implied by the theoretical guarantees and some limitations. The three controllers are compared using individual simulations of a nonlinear model for an uncrewed aircraft, including exogenous disturbances and model uncertainty. Further, two simulation studies were performed to analyze the specific impact of flight-based, realistic stochastic disturbances. The new flight control law performed as expected, and usually better than both alternatives. The simulations demonstrate an expected shortcoming of the nonlinear design: sensitivity to disturbances outside of the design uncertainty class. Future work includes investigating alternative formulations of the robust port-Hamiltonian controller and a transition to flight test.