An improved learning variable structure control method for multi-periodic disturbances rejection

Traditional learning variable structure control (LVSC) synthesizes variable structure control (VSC) as the robust part and learning control (LC) as the intelligent part to improve tracking performance for repeatable tracking control tasks. However, it can only deal with the cases with one periodic disturbance. In this paper, an improved learning variable structure control (ILVSC) method is proposed, aiming at rejecting multi-periodic disturbances with uncorrelated frequencies. In particular, the learning law is redesigned to be able to separate and approximate any of the multi-periodic disturbances in an efficient way. The stability analysis of the control system is provided. The simulations of the algorithm are presented to validate its effectiveness.