New Dynamic Prediction Approach for the Reliability Indexes of Bridge Members Based on SHM Data

Abstract

The last several decades have witnessed a bridge performance assessment shift from deterministic methodology to probabilistic methodology. Structural health monitoring (SHM) has been subjected to a rapid development process. SHM has become a predominant emerging technology to challenge and improve the traditional reliability assessment on new and existing bridges. Nevertheless, there is still a strong need for the efficient use of SHM data in the reliability prediction models. In the long-term service periods, the SHM system produces a huge amount of monitoring data, such as extreme stress data. How to properly predict structural dynamic reliability indices with these data is a bottleneck in the development processes of SHM technology. The aim of this paper is twofold: (1) to propose the newly developed combinatorial Bayesian dynamic linear models (BDLMs) with time-variant weighted coefficients based on SHM extreme stress data and (2) to present an approach for effectively incorporating the proposed combinatorial models in the dynamic reliability prediction processes of bridge members. The monitoring extreme stress data of an existing bridge is provided to illustrate the application and feasibility of the proposed models and presented approach, which can provide the theoretical foundation and application method for reliability assessment of the SHM system.