Limited Authority Adaptive Flight Control for Reusable Launch Vehicles

In the application of adaptive flight control, significant issues arise due to limitations in the plant inputs, such as: actuator displacement limits, actuator rate limits, linear input dynamics, and time delay. A method is introduced that allows an adaptive law to be designed for the system without these input characteristics, and then to be applied to the system with these characteristics, without affecting adaptation. This includes allowing correct adaptation while the plant input is saturated, and allows the adaptation law to function when not actually in control of the plant. To apply the method, estimates of actuator positions must be found. However, the adaptation law can correct for errors in these estimates. Proof of boundedness of system signals is provided for a single hidden layer perceptron neural network adaptive law. Simulation results utilizing the methods introduced here for neural network adaptive control of a reusable launch vehicle are presented for nominal flight and under failure cases that require considerable adaptation.