Closed-loop-reference-based Adaptive Learning Control for Nonlinear Systems with Parametric and Non-parametric Uncertainties

In this work, we present a novel adaptive learning control (ALC) scheme for output tracking of nonlinear systems. In order to achieve the perfect tracking performance over a given time interval, a closed-loop reference model is developed to provide both the target trajectory and the estimated state information, owing to which the composite energy function (CEF)- based ILC approach is extended from state tracking control to output tracking control of nonlinear systems. The present work considers a generic class of nonlinear systems, which include three class of uncertainties: parametric and non-parametric input uncertainties as well as input distribution disturbances. With the CEF-based convergence analysis method, we will show that both the state estimation error and tracking error will converge to zero asymptotically as the iteration number goes to infinity. The effectiveness of the proposed ALC algorithms will be verified through a numerical example.