A separation principle for the prescribed-time stabilization of a class of nonlinear systems

Abstract

Despite the recent development of prescribed-time control theory, the highly desirable separation principle remains unavailable for nonlinear systems with only the output being measurable. In this paper, for the first time we establish such separation principle for a class of nonlinear systems, such that the prescribed-time observer and prescribed-time controller can be designed independently, and the parameter designs do not affect each other. Our method makes use of two parametric Lyapunov equations (PLEs) to generate two symmetric positive-definite matrices, aiming to avoid conservative treatments of nonlinear functions commonly associated with high-gain methods during the design process. Our work provides a stronger version of the matrix pencil formulation that is applicable when nonlinearities satisfy the so-called linear growth condition, even if the growth rate is unknown. In our method the selection of design parameters is straightforward as it involves only three parameters: one for the prescribed convergence time $t_f$, and the other two are for the controller and the observer respectively, and the choice of the latter two parameters does not affect each other. Once the system order is determined, one can directly obtain reasonable ranges for these two parameters. Numerical simulations verify the effectiveness of the proposed method.