In this paper, the extension of the all-coefficient adaptive control method to nonlinear time-varying systems is studied. A novel discretizing method is first proposed to derive the discrete-time model for a class of nonlinear time-varying systems. The characteristics of the coefficients of the discrete-time model are derived by this method, based on which the all-coefficient adaptive control method is given for the class of nonlinear time-varying systems. Sufficient conditions on the closed-loop stability are given. Simulations show the purposed adaptive control method can achieve satisfying performance. Finally further discussion on applying low-order all-coefficient adaptive control method to high order nonlinear time-varying system is given.
In this paper, a characteristic model based longitudinal control design for the trans-aerosphere vehicle X-34 in its transonic and hypersonic climbing phase is proposed. The design is based on the dynamic characteristics of the vehicle and the curves it is to track in this climbing phase. Through a detailed analysis of the aerodynamics and vehicle dynamics during this climbing phase, an explicit description of the tracking curve for the flight path angle is derived. On the basis of this tracking curve, the tracking curves for the two short-period variables, the angle of attack and the pitch rate, are designed. An all-coefficient adaptive controller is then designed, based on the characteristic modeling, to cause these two short-period variables to follow their respective tracking curves. The proposed design does not require multiple working points, making the design procedure simple. Numerical simulation is performed to validate the performance of the controller. The simulation results indicate that the resulting control law ensures that the vehicle climbs up successfully under the restrictions on the pitch angle and overloading.
The characteristic modeling problem of flight vehicles'attitude dynamics is considered in this paper.In terms of the affine nonlinear system with triangle form of flight vehicles'attitude dynamics,a general method is presented to compress the dynamics into the characteristic model parameters,by introducing the time scale of nonlinear systems and a class of system states related compress functions.The parameter region and limit of the characteristic model are also given.From the given parameter region it is seen that the bound of the characteristic model parameters is dependent on the sampling period,the modeling error,the system order and the system change rate.The modeling error of the established characteristic model can be arbitrarily small according to the control precision,showing the difference between the characteristic model and other model reduction methods,that is,no system information is lost using this approach.On the basis of this modeling approach,the characteristic model of the flexible satellite attitude is established,as well as the bound and limit of the parameters,which sets a theoretical foundation for characteristic model based control design of flight vehicles.
MENG Bin &WU HongXin National Laboratory of Space Intelligent Control,Beijing Institute of Control Engineering,Beijing 100190,China
