Discrete-Time Design of Model Reference Learning Control System

This paper presents the design of discrete-time model reference learning control of linear system for tracking aperiodic signal while rejecting periodic disturbance. The design integrates Model Reference Control (MRC) and Iterative Learning Control (ILC) in order to achieve the desired objective. The use of MRC makes tracking any reference signal is possible, while ILC is used to eliminate the periodic disturbance. The proposed design is simulated on LTI system with repetitive output disturbance. Model reference learning control with several ILC learning gains are simulated. It is shown that ILC learning gain can be tuned to determine the convergence rate and stability of the whole system. Simulation results indicate that the accurate tracking of aperiodic reference can be achieved as long as the plant and disturbance models are accurately known.