ABSTRACT

OMA algorithms are divided into three categories namely: (i) time domain based methods, (ii) frequency domain based methods and (iii) timefrequency domain based methods (Zhang et al. 2005). In this paper, the performance of a time domain based method known as stochastic subspace identification (SSI) in estimation of modal parameters (natural frequencies and damping ratios) from ambient vibration measurements of Roode Elsberg dam. The three SSI algorithms evaluated here are unweighted principal component (UPC), principal component (PC) and canonical variate analysis (CVA). The measured natural frequencies are compared with those from the Finite element model.