Reliability Assessment Of Inservice Reinforced Concrete Bridge Based On Structure Inspection Information

Due to environmantal factor,working load and artificial factor ,the performance of materials may be degration, and cause structure function attenuation during structure service life,thus the safety of structure may be influenced.And that , the law of resisting force degration of materials and function regression of structure is the premise of structure reliability analysis .Based on the view of attenuation law of inservice reinforced concrete bridge can be reflected by structure inspection information ,this article think there must be a change in dealing with structure function regression problems,that is ,a shift from static to dynamic, namely combine strong personality of the characteristics of engineering structures, uitilize the inspection information of structure continuous updating the law of structure performance and make the rusults continuous approach the law of structure practical attenuation. Based on these,this article do profound investigation on Bayesian continuous updating model and neural network self-adapting model, combine the experical attenuation law of concrete and reinforcing steel and inspection information of engineering to establish Bayesian continued updating model and neural network self-adapting model,thus relize updating of the law of structure riliability attenuation. And through examing and certifying of example,we obtain that, on condition of small sample structual inspection ,the information updating efficiency of BP neural network self-adapting model isn't perfect,but Bayes continued updating model inherit the maturity of expirical regression formulas and the veracity of natural information and make the predicted results more accurate.Finally, by using Bayesian continued updating model of resisting force and current research of bidge time-dependent reliability, through a practical project,a complete process of inservice bridge time-dependent continued updating model based on Bayes is demonstrated.Thus a economical,precise and dynamic method can be proposed in structural reliability analysis based on research results.