Showing papers by "Petar V. Kokotovic published in 2002"

On Identifiability of Linear Time-Delay Systems

Abstract: Identifiability analysis is developed for linear time-delay sys- tems with delayed states, control inputs, and measured outputs, all with a finite number of lumped delays. These systems are governed by linear func- tional differential equations with uncertain time-invariant parameters and delays. It is shown that the transfer function of such a system admits the online identification if a sufficiently nonsmooth input signal is applied to the system. Sufficiently nonsmooth signals are constructively defined by im- posing different smoothness properties on the control input and the state of the system. The required nonsmothness property is verified independently of any underlying time-delay system.

Nonlinear maneuvering with gradient optimization

Abstract: The maneuvering problem involves two tasks. The first one, called the geometric task, is to force the system states to converge to a desired parametrized path. The second task, called the dynamic task, is to satisfy a desired dynamic behavior along the path. The desired geometric path /spl xi/ is viewed as a target set /spl Xi/ which is parametrized by a scalar variable /spl theta/. The proposed dynamic controller consists of a stabilization algorithm that drives the state x(t) to the point /spl xi/(/spl theta/(t)), and a smooth dynamic optimization algorithm that selects the point /spl xi/(/spl theta/) in the set /spl Xi/ that minimizes the weighted distance between x and /spl xi/. Choosing a gain /spl mu/ large in the optimization algorithm, induces a two-time scale behavior or a closed-loop plant. In the fast time-scale /spl theta/(t) rapidly converges to the minimizer, and in the slow time-scale x(t) converges to /spl Xi/. Two motivational examples illustrate the design and the achieved performance of the closed-loop.