Design method of PID-type model predictive iterative learning control based on the two-dimensional generalized predictive control scheme

For the repetitive, cyclic or batch processes, iterative learning control is essentially a two-dimensional feedback control. By transforming the iterative learning control into a two-dimensional generalized predictive control, a new model predictive iterative leaning control scheme is proposed in this paper for the repetitive, cyclic or batch processes with both time-wise and cycle-wise dynamics. As the proportional, integral and derivative of the prediction errors are considered in the cost function, the proposed ILC scheme can be referred to as a PID-type model predictive iterative learning control (PID-MPILC). Compared with the two-dimensional model predictive iterative learning control (2D-MPILC) proposed in the previous works, the proposed PID-MPILC can provide much better control performances not only along cycle but also along time, which is illustrated by the numerical simulation.