Nonlinear Output Feedback Control Of A Hypersonic Vehicle

Due to the complex flight environment, scramjet hypersonic vehicles adopt the engine-airframe integration technology and other advanced technologies. It is not only lead to a strong inherent coupling among the elastic airframe, propulsion system, and structural dynamics, but also bring a highly nonlinear dynamic model; meantime, as a result of the wide range of flight altitude and Mach, the aerothermoelastic effects and the aerodynamic characteristic have a dramatic change in the process of flight. On the other hand the dynamic model has a large number of uncertain aerodynamic parameters and unknown disturbances. All these facts greatly affect the flight stability and reliability and bring a great challenge to the hypersonic vehicle control design.In this thesis, the longitudinal dynamics of hypersonic vehicles is utilized for control design. First, the longitudinal mode motion and dynamic equations are analyzed. Then feedback linearization techniques are utilized to make the complex longitudinal model of hypersonic vehicle be completely input/output linearized. Further more, by a simple mathematical transformation to obtain the transferred model which satisfies the design requirements.A nonlinear output feedback control method is designed based on the transferred longitudinal model. Under the restriction that only the hypersonic vehicle's velocity and altitude are measurable, high gain observers (HGO) are utilized to provide estimation signals for unmeasurable derivatives of the vehicle's velocity and altitude. To compensate for the unknown parts of the vehicle's dynamic model, a radial basis neural network (RBNN) is designed as a feedforward component to compensate for the system's uncertainty. For the proposed control law, it only requires less knowledge about the hypersonic vehicle's dynamic model. A comprehensive Lyapunov based stability analysis is adopted to show that the proposed control law yields semiglobal uniformly ultimately bounded tracking while keeps all the closed loop signals bounded. Finally, numerical simulation results are presented to validate the proposed control design has a good tracking performance for vehicle's velocity and altitude, and achieve good robustness to model uncertainties.