Multi-objective iterative learning control using convex optimization

This paper presents a multi-objective iterative learning control (ILC) design approach that realizes an optimal trade-off between robust convergence, converged tracking performance, convergence speed, and input constraints. Linear time-invariant single-input single-output systems which are represented by both parametric and nonparametric models are considered. The noncausal filter Q(q) and learning function L(q) are simultaneously optimized by solving a convex optimization problem. The proposed method is applied to a non-minimal phase system and compared with a model-inversion based ILC design. Using the developed ILC design the underlying trade-off between tracking performance and convergence speed is thoroughly/quantitatively analyzed.