ABSTRACT
ABSTRACT: Deterioration modeling of reinforced concrete structures is the heart of almost any Bridge Management Systems. There are many models for prediction of concrete deteriorations depth, but almost all of them can only predict the output without any sense for uncertainty. Concrete deterioration models should be updated after inclusion of new data to existing databases. Since the collected data is uncertain one of the best modeling methods can be based on the Bayesian Decision Theory. This theory introduces an attractive method through combination of old even poor quality data or expert information based on experience with the new high quality data. In this paper a Bayesian regression model is introduced for concrete carbonation depth prediction as an example. It is found that this method of modeling with its framework can be a good candidate as an appropriate predictor for better decision making, maintenance and repair programming in Japanese Bridge Management System.
