A new approach to feedback feed-forward iterative learning control with random packet dropouts

This paper is devoted to the feedback feed-forward iterative learning control (FFILC) with random packet dropouts. Resorting to a 2D Roesser system manner, the resultant FFILC falls in the robust framework. With the help of Lyapunov method and stochastic analysis technique, a new condition is developed to guarantee the desired mean-square stability with the prescribed H_∞ performance. To co-design controller and learning parameters by means of linear matrix inequalities, an effective strategy is employed to overcome the coupling nonlinearities induced by these parameters and Lypunov variable. An injection modeling process is adopted to show the effectiveness of the proposed method.