Modal parameter extraction based on Hilbert transform and complex independent component analysis with reference

●Hilbert Transform is introduced in this study.●Mode shapes, damped frequencies and damping ratios can be extracted directly.●Orders of modal parameters are determined by arranging the reference signals.●A modified version of complex Fast ICA for BSS is presented.

A novel method for extracting modal parameters, including mode shapes, natural frequencies and damping ratios, is developed in this study. By applying Hilbert Transform (HT) and complex Independent Component Analysis (ICA) with reference for Blind Source Separation (BSS), the modal parameters can be directly extracted while other articles focus on the mode shapes only. Moreover, a revised version of fixed-point complex ICA with reference for BSS is derived for a better separating performance with higher separation accuracy by comparison. The simulation and experiment results are both presented.