Adaptive Iterative Learning Control of Switched Nonlinear Discrete-Time Systems With Unmodeled Dynamics

This paper is concerned with the issue of adaptive iterative learning control (AILC) for single-input, single-output (SISO), switched nonlinear discrete-time systems with unmodeled dynamics and stochastic measurement noise. Under the random switching rule, the nonlinear subsystem is estimated by a linear model which is constructed by minimizing the discrepancy between the real system output and the estimated system output with gradient-type technique. Meanwhile, the adaptive switched control law is designed by the updated subsystem matrix estimation. By taking advantages of norm theory, the boundedness of the estimation error is derived. Further, by the manner of the statistical technique, the results are conducted that the expectation of the tracking error is monotonically convergent to zero and the covariance matrix of the tracking error is bounded. Finally, simulation results are given to validate the proposed scheme is effective.