Convergence of MFAC based feedback-feedforward ILC systems

Using the repeatability of controlled system, a new feedback-feedforward configuration for the iterative learning control (ILC) based on model free adaptive control (MFAC), which consists of a feedback and a feedforward component, is presented for a general discrete-time nonlinear system and rigorous convergence is proved. The effectiveness and advantage of the proposed learning control method is demonstrated through simulations. The main features of this method are that, the controller doesn’t include any information of the controlled system, and, it is a modularized control system design method. The ILC control loop is added on to the existing MFAC feedback control loop, and no reconfiguration and no re-design are needed, and the ILC is a complimentary part to the existing MFAC controller.