Application of Independent Component Analysis and Blind Source Separation Techniques to Operational Modal Analysis

Independent Component Analysis (ICA) / Blind Source Separation (BSS) is an emerging research area in the field of signal processing. The goal of ICA / BSS techniques is to identify statistically independent and non-Gaussian sources from a linear mixture of such sources. It also extracts the unknown mixture matrix in the process. The purpose of this paper is to explore the possibilities of utilizing this concept of ICA for the purpose of Operational Modal Analysis in which the system is identified only on the basis of output responses which are a function of unknown input forces and underlying system characteristics. Four different ICA algorithms are used in this study by applying them on a 15 degrees-of-freedom analytical system. These algorithms differ from each other on the basis of the optimizing techniques implemented and by utilization of second or higher order statistics. The modal parameters estimated using these methods are compared with true system parameters and also with one of the Operational Modal Analysis techniques.