Iterative Learning Control for Quasi-Static MEMS Mirror with Switching Operation

This paper reports an iterative learning control (ILC) to compensate for the errors by the switching operation and the modeling inaccuracies for a quasi-static (QS) MEMS mirror. The modeling errors and uncertainties in dynamics with the switching operation between electrodes result in undesirable oscillations in beam positioning. A wideband frequency-domain ILC is proposed for a QS MEMS mirror with a flatness-based feedforward control. The improvement of the residual oscillations is demonstrated by reduced root mean square (RMS) errors for a 2 Hz and a 2-degree-amplitude sawtooth reference with a factor 69.9.