V-Tigerに基づく安定判別と周波数領域近似モデルマッチング

抄録

FRIT and VRFT are widely applied as representatives of Data-driven control, but the actual closed-loop response cannot be predicted. Recently, Data-driven prediction methods including V-Tiger that predict the actual closed-loop response have been proposed. V-Tiger predicts the virtual frequency response and virtual time response of a closed-loop system from one-shot experimental data to read the settling time and Nyquist locus, but these readings become difficult when the noise is large. Therefore, we propose (1) a stability criterion based on the virtual time response, (2) a noise reduction method that averages multiple virtual time responses using only one-shot experimental data, and (3) a control design method to minimize cost function in frequency domain to achieve robustness against noise.