ABSTRACT
Structural health monitoring (SHM) integrated reliability analysis, which is extensively investigated by the researchers during the past few decades, provides a rather convenient framework for performance assessment of structures. Classical reliability analysis takes into account the uncertainty of loading (induced by dead, live, wind, seismic load, etc.) as well as the resistance based on assumed analytical models. In this context, SHM makes it more practical and credible to consider the parametric uncertainties in analytical models during the reliability analysis. Therefore, the implemented SHM methodology has a critical role in the uncertainty quantification of the considered analytical model of the structure. In this study, a recently developed two-stage Bayesian finite element model updating methodology is adapted to the classical reliability problem. To constitute a more robust methodology, first, the probability distribution function for model parameters is estimated by using the two-stage methodology. Then, the resulting probability density function (pdf) is substituted into the reliability integral. By making use of this, time-costly methods such as Markov Chain Monte Carlo simulation can be eliminated, and reasonable results can be achieved through the classical Monte Carlo simulation.
