Investigation of Operational Modal Identification of a Cable-Stayed Bridge Based on Bayesian Estimation Considering Stochastic Uncertainty

Abstract

A previous study reported that results of operational modal analysis applied to a cable stayed bridge was drasticly changed depending on wind speed. Aiming at investigation of an effective bridge health monitoring technology, it is desired to identify the mechanism of the modal identifiability problem. This study presented a Bayesian estimation method based on frequency domain of ambient vibration data. Modal properties identified from SSI are compared to ones identifed from the Bayesian method. Coefficient of variances of modal parameters and S/N ratio of modal responces indicated uncertainty of modal estimation directry. This observation shows Bayesian estimation can be helpful approach to interpret quarity of the estimation itself.