Iterative Learning Control of Discrete Systems with a Switching Reference Trajectory and Saturating Inputs.