Development of an Integrated Method for Probabilistic Bridge-Deterioration Modeling.

Probabilistic deterioration models such as state-based and time-based models are only capable of predicting future bridge-condition ratings when a sufficient amount of condition data and reasonable data distribution are available. However, such are usually difficult to acquire from limited bridge-inspection records. As a result, these probabilistic models cannot guarantee reliable long-term prediction for each of the bridge elements concerned. To minimize this shortcoming, this paper proposes an advanced integrated method to construct workable transition probabilities for predicting long-term bridge performance. A selection process within this method automatically chooses a suitable prediction procedure for a given situation in terms of available inspection data. The backward prediction model (BPM) is also incorporated to effectively predict the bridge performance when sufficient inspection data are unavailable. Four different situations in regard to the available inspection data are predefined in this study to demonstrate the capabilities of the proposed integrated method. The outcomes show that the method can help develop an effective prediction model for various situations in terms of the quantity and distribution of available condition-rating data. [ABSTRACT FROM AUTHOR]