Probabilistic damage detection and identification of coupled structural parameters using Bayesian model updating with added mass

Damage detection inevitably involves uncertainties originated from measurement noise and modeling error. It may cause incorrect damage results if not appropriately treating uncertainties. To this end, vibration-based Bayesian model updating (VBMU) is developed to utilize vibration responses or modal parameters estimate structural the associated of those estimates. However, traditional VBMU often assumes that mass well known invariant because simultaneous identification stiffness yield an unidentifiable problem due coupling effect stiffness. In addition, posterior PDF in usually approximated by single Markov Chain Monte Carlo (MCMC), leading a low acceptance rate limited capability for complex structures. This paper proposed novel address identify adding mass. Two data sets are acquired original modified systems with added mass, giving new characteristic equations. Then, reformulated measured predicted counterparts For efficiently approximating PDF, Differential Evolutionary Adaptive Metropolis (DREAM) algorithm adopted draw samples running multiple chains parallelly enhance sufficiently explore possible solutions. Finally, numerical example ten-story shear building laboratory-scale three-story frame structure utilized demonstrate efficacy framework. The show method can successfully both stiffness, their Reliable probabilistic also be achieved.