Reliability-based condition assessment of in-service bridges using mixture distribution models

摘要: Integrating structural health monitoring (SHM) data with reliability analysis procedures provides a novel approach for bridge condition assessment since is an important performance measure of and reliability-based have the capability accommodating uncertainties in measurement data. Because strain response acquired from under in-service environment usually result multi-load effect such as traffic (highway, railway, or both them) wind (monsoon typhoon), it cannot be characterized by standard probability distribution model adequately. In present study, to using mixture models proposed. With Weibull distributions being component density functions, expectation maximization (EM) algorithm conjunction Akaike information criterion (AIC) implemented iterative solution optimal number components parameters finite modeling peak stresses which are derived long-term models, proposed method capable handling any structure accurately evaluating indices monitored components. case applied assess deck trusses instrumented Tsing Ma Bridge (TMB) various load combinations monsoon, typhoon, without railway traffic; efficiency characterizing statistical properties peak-stress multiple engendering effects convergence EM-based iteration estimation validated.