A data-driven optimal design of point-to-point ILC

This paper presents a new data-driven optimal design framework of point-to-point iterative learning control (PTP-ILC), where the control signal is directly updated from the errors of given multiple intermediate pass points. A major contribution is that the presented optimal PTP-ILC mechanism only uses the real-time measured I/O data without any model information of the plant for the controller design, convergence analysis, and conduction, from which the distinctive `data-driven' feature of the presented approaches is obvious and intuitional. Rigorous mathematical analysis is developed to illustrate the efficiency of the proposed approach.