Output-Feedback Adaptive Stabilization for Nonlinear Systems with Unknown Direction Control Coefficients
Yun-Gang LIU,Shuzhi Sam GE
DOI: https://doi.org/10.1109/ACC.2005.1470737
2005-01-01
Abstract:This paper investigates the problem of output- feedback adaptive stabilization control design for a class of nonlinear systems with unknown direction control coefficients. By introducing novel input scaling transformations, the un- known control coefficients can be lumped together such that the original system becomes a new system and make the output feedback control design possible. After the introduction of an observer and estimator for state and parameter estimates, respectively, a constructive design procedure is given for the output-feedback adaptive stabilization control based on inte- grator backstepping. It is shown that, under some conditions, the controller designed ensures the state and the estimation error of the closed-loop system asymptotically converge to zero, while other closed-loop signals are uniformly bounded. considered completely in (11), (15), (16). The objective of this paper is to investigate the output-feedback adaptive stabilization control of this nonlinear systems. First, we introduce a series of novel input scaling transformations, through which the unknown control coefficients distributing in the each subsystem can be lumped, thus leading to a new nonlinear system with only one unknown direction control coefficient in the first subsystem. Then, starting with this new nonlinear system, an observer and an estimator can be designed for state and parameter estimates, respectively. Thus, by using the integrator backstepping with tuning functions approach and the observer and parameter esti- mator designed, a constructive design procedure for output- feedback adaptive stabilization control is given. Our main result shows that, under some conditions, the controller designed ensures the state and the estimation error of the closed-loop system asymptotically converge to zero while all other signals are uniformly bounded.