Neural control of hypersonic flight dynamics with actuator fault and constraint

This paper deals with the control problem of actuator fault and saturation for hypersonic flightvehicles. Different from previous back-stepping design, the scheme is on transforming the dynamics into the“prediction function”. The controller is constructed with high gain observer, while the effect of fault andsaturation is compensated by neural networks. For the input saturation, the auxiliary dynamics is included todesign the adaptive learning law. The neural weights and filtered tracking error are guaranteed to be boundedvia Lyapunov approach. The effectiveness of the proposed method is verified by simulation of winged-conemodel.