Robust Adaptive Optimal Control Of Near Space Vehicle With Multiply Indexes And Multi-constriaints

Near space vehicle(NSV)is a kind of aircraft which can continuously work in the near space and complete the specified flight missions.Compared to the traditional aircraft,NSV has some advantages such as high speed,wide envelop,high mobility and multitasking ability,which makes it to have great application prospects in both military and civil fields.The core research direction for NSV is to design a suitable flight control system to guarantee the flight security and quality.However,some multiple constraints such as unmodeling error,external disturbance and input saturation should be taken into account during the flight control design of NSV due to its complex structural characteristic and flight environment,and the robust and anti-windup ability of controller can be enhanced.Moreover,it is generally expected that the controller not only can guarantee the stable flight of NSV,but also can meet certain performance index requirements while considering the optimal flight control.Therefore,in this thesis,a robust adaptive optimal flight control scheme will be researched for NSV system with multiple constraints including unmodeling error,external disturbance and input saturation.The main content of the thesis is shown as follows:(1)Based on the existing NSV model,the influence of external disturbance is considered.A robust optimal control scheme which combine the nonlinear disturbance observer(NDO)and adaptive dynamic programming(ADP)methods is proposed for the NSV attitude system in the presence of moment disturbance.Firstly,NDO method is used to estimate the external disturbance generated by an exogenous system,and a steady state control input combine the disturbance estimation value,known system dynamic and reference signal is designed to remove the effect of external disturbance.Then a single critic network based ADP method is applied to design the optimal feedback controller to ensure the optimal tracking performance.Finally,Lyapunov method is used to show the stability of closedloop system and simulation tests on NSV attitude system is given to show the effectiveness of proposed robust optimal control scheme.(2)The effect of system unmodeling error will be further considered into the attitude flight control of NSV since it is difficult to obtain its accurate nonlinear model.Under the backstepping control frame,a robust adaptive optimal control scheme combing the NDO,radial basis function neural network(RBFNN)and ADP method is presented for NSV attitude system in the presence of system unmodeling error and external disturbance.Firstly,RBFNN technique is used to approximate the system unmodeling error,and NDO method is applied to estimate the composite disturbance which is consisted of RBFNN estimation error and external disturbance.And the effects of system unmodeling error and external disturbance is eliminated by designing a feedforward control input,and the tracking control problem of original system can be transformed into a stabilization problem of a nominal error system.Then,a single critic network based ADP method is applied to design the optimal feedback controller to guarantee the optimal tracking performance.Finally,Lyapunov method is used to show the stability of closedloop system and simulation tests on NSV attitude system with system unmodeling error and external disturbance are given to show the effectiveness of proposed robust adaptive optimal control scheme.(3)Since the actuator is prone to rendering saturation,the multiple constraints including system unmodeling error,external disturbance and input saturation are considered simultaneously into the optimal attitude control of NSV.A robust optimal control scheme based on integral sliding mode control(ISMC)and ADP method is proposed.Firstly,an auxiliary system is constructed to deal with the input saturation problem,and RBFNN method is applied to approximate the system unmodeling error.And an adaptive ISMC is designed to remove the effects of system unmodeling error,external disturbance and input saturation by guaranteeing the stable of sliding mode surface.Then,an ADP based optimal tracking controller is designed for inner dynamic of sliding mode surface to guaranteed optimal tracking performance.The stability analysis of whole close-loop system is given and simulation tests for NSV attitude system with multiple constraints are listed to show the effectiveness of proposed robust optimal control scheme.(4)For the same optimal attitude control of NSV with multiple constraints such as system unmodeling error,external disturbance and input saturation,a differential game-based finite-horizon robust adaptive optimal control strategy is presented.Firstly,RBFNN approximation technique and auxiliary system control method are applied to deal with system unmodeling error and input saturation,respectively.And an adaptive feedback control input is designed such the effects of unmodeling error and input saturation are removed,and the tracking control problem of original system with multiple constraints can be transformed into a H∞ optimal problem of an affine sysem.Then,a ADP based differential game strategy is used to designed the finite-horizon H∞ optimal feedback controller,and the robust optimal tracking performance of system is ensured during a finite horizon.Finally,the goal of finitehorizon robust adaptive optimal attitude control with multiple constraints is realised.And the stability analysis of closed-loop system and simulation tests are given to verify the effectiveness of proposed control scheme.(5)Based on the research works about optimal attitude control in previous chapters,the trajectory optimal control problem will be further considered for NSV longitudinal model with multiple constraints including system unmodeling error,external disturbance and input saturation,and a finitehorizon robust adaptive optimal trajectory control scheme will be designed.Firstly,for the out-loop trajectory subsystem without considering the above constraints,feedback linearization methos is applied to design the height and velocity controllers,respectively,and Newton iteration method is used to calculate the reference attitude angle signals for inner-loop attitude subsystem.Then,a finite-horizon robust adaptive optimal tracking control scheme combing NDO,RBFNN and auxiliary system control methods is proposed,and the reference attitude angle signals can be optimally tracked under multiple constraints.Finally,the finite-horizon robust adaptive optimal trajectory control of NSV can be realised,and the stability analysis of closed-loop system and simulation tests for NSV longitudinal model are given to verify the effectiveness of proposed control scheme.