Finite-dimensional adaptive observer design for uncertain nonlinear systems with time delays and sampled outputs

This paper addresses the problem of observer design for a class of uncertain nonlinear time-delay systems. A finite-dimensional adaptive observer is proposed based on Lyapunov-Krasovskii functional method and neural-networks-based approximation technique. The designed observer guarantees that all signals in the closed-loop system are bounded and the estimation error decreases to zero asymptotically. In order to prevent drift of adaptive parameters, projection modification algorithm is employed. Theoretical analysis and simulation both verify the effectiveness of the proposed observer design method.