Iterative learning control for systems with nonparametric uncertainties under alignment condition.