Two-mode ILC with pseudo-downsampled learning in high frequency range

In this paper, an iterative learning control scheme that combines the designs in the frequency and time domains is proposed to improve the tracking accuracy. The scheme separates the error signals into low and high frequency band error components and employs different learning mechanisms to deal with these error components, respectively. On the low frequency band, a conventional iterative learning control is used. On the high frequency band, a downsampled learning is implemented to suppress more error components to generate a high tracking accuracy. We will address in detail the issues of signal separation, learning controller design, convergence analysis, and experimental verification of the proposed scheme