Abstract: An iterative learning controller based on model prediction is proposed to address the trajectory tracking for a class of repetitive system with nonrepetitive disturbances.The iterative learning controller acts on repetitive process as feed-forward.Based on the temporal model of tracking error,the model predictive control is applied to improve the tracking performance along the time index for controlling the current stochastic disturbances.In order to restrain the repetitive disturbances,the P-type iterative learning controller is introduced along the batch index.Finally,the effectiveness of the proposed method is demonstrated by simulation results.