Monotonically convergent feedback-forward iterative learning control for linear discrete-time systems

For a class of linear discrete-time systems, a feedback feed-forward iterative learning control (ILC) scheme is proposed, which can guarantee the system is monotonically convergent along the iteration direction and is Bounded-Input Bounded-Output (BIBO) stable along the time direction. First, the feedback feed-forward iterative learning system is presented by a two-dimensional (2-D) Roesser model system. Then, the monotonic convergence problem along the iteration direction is converted to a H_∞ disturbance attenuation problem of a one-dimensional system. Third, the sufficient condition of monotonic convergence along the iteration direction is given by LMIs which can also guarantee the system is BIBO stable along the time direction. Furthermore, the LMI condition can determine the gain matrix of the feedback feed-forward iterative learning controller. Finally, the simulation results are presented to demonstrate the effectiveness of the proposed scheme.