PI control of a tailless fighter aircraft with dynamic inversion and neural networks

This paper presents an application of dynamic inversion control with adaptive neural networks to a tailless fighter aircraft. We propose an extension of the dynamic inversion control using proportional-integral desired dynamics. The performance of the PI adaptive dynamic inversion controller is compared to an adaptive dynamic inversion controller using PD desired dynamics. The controller was tested in a 6-DOF simulation of the ICE tailless fighter aircraft. Simulation results are given for simultaneous longitudinal and lateral manoeuvres performed under nominal conditions and also with the simulated failure of a major control surface. The PI dynamic inversion controller with neural network adaptation outperforms both the baseline PI controller without adaptation and also a PD dynamic inversion controller with neural network adaptation.