Robust analysis and synthesis with unstructured model uncertainty in lifted system iterative learning control

This paper discusses robust iterative learning control (ILC) analysis and synthesis problems that account for model uncertainty in the lifted system representation. In the robust analysis, we transform the robust monotonic convergence condition with unstructured uncertainty into an equivalent convex problem. In this framework, for a given learning gain Q, the design of the learning gain L that maximizes the convergence speed is reformulated as a convex optimization problem. We discuss various properties of the proposed robust ILC analysis and design, and analyze the performance of the proposed robust ILC design through numerical simulations.