In a full Bayesian probabilistic framework for “robust” system identification, structural response predictions and performance reliability are updated using test data 𝒟 by considering the of whole set possible models that weighted their probability. This involves integrating h(θ)p(θ|𝒟) over parameter space, where θ is vector defining each model within structure, h(θ) prediction quantity interest, p(θ|𝒟) probability density θ, which provides measure how plausible given 𝒟. The evaluation this...

A Bayesian probabilistic methodology for structural health monitoring is presented. The method uses a sequence of identified modal parameter data sets to compute the probability that continually updated model stiffness parameters are less than specified fraction corresponding initial parameters. In this approach, high likelihood reduction in at location taken as proxy damage location. concept extends idea using indicators changes from when structure initially an undamaged state and then...

A method is presented for efficiently computing small failure probabilities encountered in seismic risk problems involving dynamic analysis. It based on a procedure recently developed by the writers called Subset Simulation which central idea that probability can be expressed as product of larger conditional probabilities, thereby turning problem simulating rare event into several involve simulation more frequent events. Markov chain Monte Carlo used to generate samples, otherwise nontrivial...

This paper develops a Monte Carlo simulation (MCS)-based reliability analysis approach for slope stability problems and utilizes an advanced MCS method called “subset simulation” improving efficiency resolution of the at relatively small probability levels. Reliability is operationally decoupled from deterministic implemented using commonly available spreadsheet software, Microsoft Excel. The package validated through comparison with other methods commercial software. then used to explore...

A statistical methodology is presented for optimally locating the sensors in a structure purpose of extracting from measured data most information about parameters model used to represent structural behavior. The can be updating and damage detection localization applications. It properly handles unavoidable uncertainties as well uncertainties. optimality criterion sensor locations based on entropy, which unique measure uncertainty parameters. these computed by Bayesian methodology, then...

Previously a Bayesian theory for modal identification using the fast Fourier transform (FFT) of ambient data was formulated. That method provides rigorous way obtaining properties as well their uncertainties by operating in frequency domain. This allows natural partition information according to frequencies so that well-separated modes can be identified independently. Determining posterior most probable parameters and covariance matrix, however, requires solving numerical optimization...

A Bayesian probabilistic methodology for structural health monitoring is presented. The method uses a sequence of identified modal parameter data sets to continually compute the probability damage. In this approach, high likelihood reduction in model stiffness at location taken as proxy damage corresponding location. concept extends idea using indicators changes parameters linear finite‐element and from structure undamaged possibly damaged states. This extension needed because uncertainties...

This paper presents a two-stage structural health monitoring methodology and applies it to the Phase I benchmark study sponsored by IASC-ASCE Task Group on Structural Health Monitoring. In first stage, modal parameters are identified using measured response from undamaged system then (possibly) damaged system. second these data used update parametrized model of Bayesian identification. The approach allows one obtain not only estimates stiffness but also probability that damage in any...