Asymptotic Behaviro of Adaptive Systems: Convergence Analysis Without the Barbalat's Lemma

Convergence of the state error e to zero in adaptive systems is shown using the uniqueness of solutions and the existence of a Lyapunov function in which the adaptation laws are constructed. Results in the paper are general, and therefore applicable to any adaptive control of a linear/nonlinear, time-varying or distributed-parameter system. Since the approach taken in the paper does not require the boundedness of the derivative of the state error e for all t .geq. 0, it is particularly useful in the adaptive control of infinite dimensional systems.