Nonlinear dynamic inversion (NDI) is a methodology that obtains a nonlinear control law that globally reduces system dynamics with respect to selected control variables to a chain of integrators. However, this approach is sensitive to model uncertainties. In this paper, a cascaded NDI based load factor controller is designed for the longitudinal dynamics of a scaled flight test aircraft and its performance is compared to an established gain scheduled controller. Due to the fast inner control loop, unmodeled actuator dynamics and time delays negatively affect the system. The effects of unmodeled actuator dynamics and time delays on NDI are therefore investigated analytically and in simulation. We propose measures to improve performance in the presence of unmodeled actuator dynamics and time delays. Furthermore, the performance of the proposed NDI controller is compared to the performance of an established gain-scheduled controller.
