Design of model predictive controller based on iterative learning control

A model predictive controller based on iterative learning control is proposed. This algorithm which combines real-time control with iterative learning control is developed to address the trajectory tracking for a class of repetitive system with non-repetitive disturbances. First, a generic model which describes the state transition of a time-varying linear repetitive system along batch indices as well as time indices is derived in a temporal state space form. Based on this model, predictive control algorithm that utilizes past data along with real-time measurements is devised. Then iterative learning control law is given. It is shown that, by using this algorithm, perfect tracking can be achieved as the number of iteration grows.