Normalized Optimal Control Modification and Flight Experiments on NASA F/A-18 Aircraft

It is well-known that the use of a large adaptive gain in model-reference adaptive control (MRAC) can result in poor robustness.1 Two well-known robust modification methods, namely; the σ -modification2 and e-modification,3 have been used extensively. The last decade has seen a surge in many new adaptive control methods such as the L1 adaptive control,4 adaptive loop transfer recovery,5 Kalman filter adaptive control,32 derivative-free adaptive control,7 composite model-reference adaptive control,8 optimal control modification,9 and others. The optimal control modification was developed in 2008 based on an optimal control framework.9, 10 For linear uncertain plants, the optimal control modification can be shown to have a linear asymptotic property for fast adaptation.11–13 This allows stability properties of the optimal control modification to be analyzed for asymptotic linear plants in the limit. Using this approach, the Rohrs counterexample has been successfully redesigned with the optimal control modification to provide an analytical stability margin in the limit.11 Maintaining fast adaptation even after the ∗Research Scientist, Intelligent Systems Division, Mail Stop 269-1, AIAA Associate Fellow †Aerospace Engineer, Flight Controls Branch, P.O. Box 273/Mail Stop 4840D, AIAA Member ‡Aerospace Engineer, Flight Controls Branch, P.O. Box 273/Mail Stop 4840D, AIAA Senior Member §Aerospace Engineer, Flight Controls Branch, P.O. Box 273/Mail Stop 4840D, AIAA Member