Robustness, Convergence, and Lyapunov Stability of a Nonlinear Iterative Learning Control Applied at a Wafer Scanner

The robustness, convergence, and Lyapunov stability properties of a nonlinear iterative learning control with applications to an industrial wafer scanner are studied. Both measured and simulated, it is demonstrated that the nonlinear design balances fast convergence under ´large-gain´ feedback during the first trials with sufficient robustness against model uncertainty under ´small-gain´ feedback during the subsequent trials. The nonlinear design is demonstrated to outperform underlying linear designs that either become unstable or have to reside to lower convergence rates.