Adaptive control in nonlinear dynamics (Q916618)

This paper studies the extension of an adaptive control algorithm recently suggested in the literature, to multivariable nonlinear systems with more than one parameter. Several systems with fixed point equilibrium regime and with periodical dynamic behavior, such as limit cycles, are explored. The results confirm that the conclusions in the scalar case are valid for multivariable nonlinear systems, too, the most important fact being the inverse proportionality relation between the recovery time and control stiffness (for small stiffness). The authors also studied the relation between control stiffness and the maximum strength of a shock beyond which the system fails to recover. In the case of a two-dimensional discrete system, they observed that there is a control stiffness value beyond which the system cannot recover from all shocks, however small in magnitude. The present study confirms that adaptive control provides a powerful and robust approach for regulating nonlinear multidimensional systems.