Abstract:

For a class of timevarying nonlinear systems with nonparametric uncertainties, a new iterative learning control (ILC) algorithm is presented. The convergence is proven by the Lyapunovlike approach. In order to improve its convergence rate， an equivalent system is built for the original nonlinear system. The equivalent control of the original system is given by the linear extended state observer (LESO) of the equivalent system. The equivalent control is an approximation of the desired control. In order to eliminate the peaking phenomenon of LESO at initial time, a solving method of the equivalent control is presented， which combines clockwise and counterclockwise estimation. The implementation flow diagram of the ILC based on equivalent control is given. Simulation results verify the effectiveness of the two proposed methods and the ILC based on equivalent control can get satisfactory tracking performance almost after one iteration.