Research On Parameters Identification And Attitude Control For On-orbit Servicing Spacecraft

On-orbit servicing technologies with its remarkable economic value and potential military significance,has always been an important research focus in the aerospace area.Currently,the autonomous on-orbit service mode is becoming an important development direction of orbital service technology.Two key problems of autonomous on-orbit service,including “on-orbital identification” and attitude control for combined spacecraft,are studied in detail,and the main results achieved in this dissertation are summarized as follows.1 The attitude dynamics model of combined spacecraft during the service operation is derived.According to the characteristics of on-orbit service operation,this paper abstracts the combined spacecraft system as a multi-body system which composed of a main body and several sub-bodies,and the attitude dynamic model of the combined spacecraft is derived by using the Newton-Euler formulation.Then,the stability of the attitude dynamics system is analysis by means of Lyapunov stability theorem.At last,the proposed mathematical model is used to simulate the operation process of on-orbit refueling service,and the influence of tank layout on attitude motion is analyzed.2 An optimal excitation trajectories design approach is proposed to improve the accuracy and efficiency of on-orbit identification of combined spacecraft inertia parameters.The problem of optimal trajectories design is formulated as a kind of optimal control problem involving Mayer formed cost function,dynamic constraint,path constraint and boundary condition,by taking the minimum of condition number of the Normal Matrix as the optimization criterion.In this paper,the input trajectory optimization approach based on Radau pseudospectral method is proposed,which includes the initial trajectories generator and the serial solver,solves the problems of initial trajectories selection and fast calculation.The filter approaches related to inertia parameters estimation are furtherly studied,and a new nonlinear filtering algorithmDual Unscented Kalman Filter(DUKF)is introduced to solve the strong nonlinear filtering model.The simulation results show that,the proposed optimal excitation trajectories enables the convergence time of the DUKF algorithm is reduced by 54% and the relative error of inertia parameters estimates is reduced by 31.3%,when compared to parameter estimation using the sinusoidal trajectories.3 Four novel adaptive attitude controllers for attitude tracking applications without unwinding are developed for the case when the combine spacecraft parameters experiences fast variations.Firstly,a classical adaptive controller is designed based on the certainty-equivalence framework and the nonlinear cancellation technique.Then,a modified adaptive controller is proposed by utilizing the smooth projection scheme to bound the parameters estimates to with a convex set.Finally,the nonlinear damping technique is adopted to modify the proposed control laws,and the two inverse optimal adaptive controllers are designed.The almost global stability of the closed-loop systems are proven by using the Lyapunov stability theorem.Simulation results demonstrate that: when the closed-loop system is asymptotically stable,smooth projection scheme can improve the response speed and control precision;when the closed-loop system is uniformly bounded stable,the inverse optimal adaptive controller enable to attenuate the influence of external disturbance.4 This dissertation also investigates the finite-time attitude tracking problem of the combined spacecraft system.Firstly,two types of 2-SM super-twisting-like structure with adaptive gains are proposed,in which the adaptive gains are updated in such away that they are as small as possible,and yet large enough to sustain a sliding motion.Secondly,a novel reaching law for SM control is proposed,this novel reaching law has the characteristic of global fixed-time convergence,which means the upper bound of convergence time is independent of initial states value,the upper bound of convergence time and the steady-state error are also provided.Thirdly,a non-singular terminal sliding mode manifold with rotation matrix measurement is designed,it has the characteristic of finite-time convergence when the states of attitude motion are restricted on this sliding manifold.Finally,by utilizing the proposed super-twisting-like structure and the novel approach law,adaptive-gain super-twisting-based controller and disturbance observer-based controller are proposed,respectively,and two almost finite-time rigorous convergence proofs for spacecraft attitude tracking system driven by proposed controllers are also given.Simulation results demonstrate that the proposed controllers enable the attitude tracking error converges to zero in finite-time despites of the parameters uncertainty and persistent external disturbance,and also synthesizing a set of smooth and continuous control trajectories,which attenuates the chattering effectively and has better performance against the traditional SM controller.5 Base on the method of integrator backstepping,nonlinear feedback control and extended state observer(ESO),the attitude tracking problem for combined spacecraft under actuator saturations is addressed.Firstly,by utilizing the characteristic that the overall attitude tracking system is a nonlinear cascade interconnection,a virtual control law with rotation matrix feedback is designed to stabilize the kinematics subsystem,the almost global asymptotic stability of the closed-loop control subsystem is proven via the LaSalle invariant principle and the Lie group properties of the SO(3).Then,a more succinct spacecraft dynamics equation is obtained by defining a new state vector,which transforms the original attitude-tracking problem into a new state-stabilized problem.Finally,Four ESO-based controllers base on nonlinearity cancellation or nonlinearity damping are proposed.The ultimately bounded stability of the closed-loop system driven by proposed controllers are proved via Input-State Stability theorem,and the accurate expression of convergence region is also provided.The effectiveness and robustness of the proposed controllers are demonstrated via simulation studies.Contrasting simulation results indicate that the controllers base on nonlinearity damping enable to shorten the period of actuator saturation,and has better attitude tracking accuracy against the nonlinearity cancellation.