Abstract: The effects of uncertainties due to test and model errors on the updated model parameters are not consi- dered in the traditional finite element model updating methods． These methods are difficult to be applied to compli- cated structures． To solve the above problems，a Bayesian finite element model updating method based on ambient vibration test and interactive access of commercial software was proposed． Modal parameters were identified by a fast Bayesian FFT method with the ambient vibration test data． Taking the identified modal parameters as known data，the posterior probability density function of the updated parameters was obtained by Bayesian theorem． By minimizing the negative log-likelihood function with the simplex method，the most probable values and coefficients of variation of the updated parameters were obtained． A data-exchange algorithm was employed to integrate the Bayesian structural model updating method in MATLAB with the finite element analysis results in SAP2000 to up- date the finite element model parameters automatically． Ambient vibration tests and modal parameter identification were conducted on two long-span floor structures． The proposed method is employed to update the values of elastic modulus of concrete floors and steel beams． The modal analysis results from the updated finite element model agree well with the identified ones．

Key words: ambient vibration test, Bayesian theorem, model updating, modal parameter identification, finite element model, long-span floor structure