Adaptive Robust Autopilot Design for Bank-to-Turn Aircraft

Based on nonlinear geometric theory, in this paper we propose an adaptive robust autopilot design to achieve the satisfactory tracking performance. The non-minimum phase phenomenon is a typical difficulty encountered in handing the aircraft dynamics. To cope with or relieve this undesirable intrinsic property, the output redefinition method is adopted. Here, we do not ask the underlying system to be minimum phase, but what we need is that the nominal system, which contains the nominal part of the aircraft dynamics, is minimum phase. For promoting the maneuverability of the aircraft, we introduce the concept of output switching: according to the profile of the desired trajectory, the control outputs are suitablely switched to different sets. In general, only two sets of outputs will be enough. Here, for simplicity, the dynamics of the actuators are modelled by first-order systems so that the redefined systems will be square affine nonlinear systems with stable zero-dynamics incorporated with mismatched uncertainties. At the end of the theoretical derivation, we present an adaptive robust autopilot design to deal with the uncertainties efficiently without prior knowledge of the bounds on the uncertainties. The stability analysis shows that the states of the overall system and the tracking errors are uniformly bounded. Moreover, the magnitude of the tracking error is directly proportional to the magnitude of the uncertainties and of some design constant. Finally, the simulation for the practical autopilot design given different desired trajectories demonstrates the effectiveness of the present work.