An identification based optimization of fractional-order iterative learning control

In this paper, a practical Identification strategy is applied to the optimal design of fractional-order iterative learning control (FOILC). The initialized fractional-order gray-box system with commensurate order or non-commensurate order is identified by using the fractional-order iterative learning Identification and the least square or instrumental variable method. The optimal D^α-type FOILC is derived from the estimated system and by the assumption that the fractional-order system is repeatable. Many other discussions on FOILC can be included as special cases of this paper. A number of illustrated examples are shown to validate the concepts.