Research On Nonlinear Cruising Flight Control Techniques Of Hypersonic Flight Vehicle

Owing to the unique advantages of high flight altitude and speed,strong maneuverability and so forth,hypersonic flight vehicle(HFV)has gained huge attention all over the world.During the last decades,all military powers,especially United States and Russia,have carried out a large number of researches relating to the key techniques of hypersonic flight and have achieved fruitful results.Nonetheless,due to the characteristics such as high nonlinearities,strong coupling,strong uncertainties and fast time-varying parameters,the control system design is still facing huge challenges.This thesis mainly focuses on the cruising flight control of hypersonic flight vehicle.A series of in-depth studies have been conducted based on several nonlinear control methods,such as nonlinear dynamic inversion control,back-stepping method,dynamic surface control,adaptive control and sliding mode control.Firstly,based on the Winged-Cone hypersonic flight vehicle,a six-degree-of-freedom mathematical model of the vehicle has been established via Newton-Euler method.For the convenience of cruising flight control system design,the nonlinear longitudinal model of HFV is obtained by reasonable simplification hypothesis.Under the nominal circumstance,the linear model at a specific equilibrium is obtained via small perturbation linearization method and the open-loop characteristics are analyzed.It is found that there is one open-loop pole in the right half complex plane,so the longitudinal model of the vehicle is unstable.Hence,flight control system design for hypersonic flight vehicle is necessary.Secondly,based on the exact feedback linearization technique,the decoupling of the velocity and altitude subsystems of hypersonic flight vehicle is achieved by deriving velocity and altitude three and four times,respectively.So the longitudinal inverse model of the hypersonic flight vehicle is obtained,and the dynamic inversion controller is directly designed.The tracking errors characteristic polynomial is made Hurwitz via pole assignment method,which ensures the stability of the system.A second order extended state observer(ESO)is designed to effectively estimate the parameter uncertainties and external disturbances,which makes the closed-loop system robust to the disturbances through the dynamic compensation in the proposed controller.Thirdly,an ESO-based robust back-stepping controller is proposed as the outer control loop, while the acquired longitudinal inverse model of HFV is the inner loop.Considering the existence of the problem of “differentiation explosion” in the classical back-stepping method,the dynamic surface technique is introduced and an inverse hyperbolic sine function based tracking differentiator(IHSTD),which has stronger noise suppression ability,is designed to obtain the derivation signal of the virtual control quantities.So the complexity of the traditional back-stepping controller is significantly reduced.And a second order linear extended state observer(LESO)with simpler configuration and only one adjustable parameter is designed,which ensures the precise estimation and compensation of the system uncertainties and external disturbances.The proposed method guarantees the stable velocity and altitude commands tracking performance,while the disturbance-rejection ability of the system is enormously improved.Fourthly,an adaptive sliding mode controller based on exponential reaching law is exploited.The system uncertainties with unknown upper bound is effectively overcome by the adaptive gain of switching term,which ensures the stable command tracking performance of hypersonic flight vehicle.Considering the problem of slow convergence speed of traditional sliding mode control,a multiple reaching law based sliding mode controller,which adjusts the sliding surface reaching process through three adjustable exponential terms,is proposed.The system convergence speed is improved significantly.A novel nonlinear disturbance observer(NDO)based on IHSTD is designed,which can estimate the system uncertainties and external disturbances precisely.The simulation results show that the proposed method ensures the system states stay on the sliding surface during almost the entire simulation process comparing with traditional sliding mode controller and double reaching law based sliding mode controller.Furthermore,the disturbances estimation errors of the newly designed NDO can converge to any small neighborhood of the origin within limited time,which massively enhances the robustness and adaptability of the system to the parameters perturbation and external disturbances.At last,concerning the problems of unmatched disturbances and control inputs constraint of hypersonic flight vehicle,the longitudinal inverse model of the vehicle is transformed into the strict parameter feedback form,and an adaptive dynamic surface controller is exploited.The derivatives of virtual control quantities are obtained by using first-order low pass filter(LPF),which reduces the designing complexity of the controller.The precise estimation of unmatched disturbances and filtering errors is realized via the designed adaptive compensation term.By designing auxiliary systems to compensate the actual outputs of the actuators,the problem of control inputs constraint is effectively improved.Simulation results and analysis show that the vehicle can stably track large reference commands under the control of the proposed method while ensuring the robustness to the unmatched disturbances and control inputs constraint.