Convergence analysis of ILC process for networked system with system noise

The paper is devoted to the convergence analysis of iterative learning control process for discrete-time networked system, where the system dynamics are assumed to be unknown. We use the iterative learning scheme developed in [2], [12] to obtain the required system information. By using λ-norm technique, we demonstrate the effect of system noise on the learning process for networked system. The results show that the tracking performance is highly dependent on the statistical characteristics of system noise, independent of output packet loss rate and decreases as the input packet loss rate increases. In addition, a numerical example is presented to show the findings.