Multi-Source Data Interpretation for Smart Structural Monitoring

Structural health monitoring (SHM) systems have been gradually equipped for important structures to trace structural health status through various sensors. The data recorded by sensors reflect the variation of structural responses and the environment, which is valuable for evaluating structural performance. Developing the prediction methods that can quantify the uncertainties is critical for smart structural monitoring. Bayesian approaches provide a powerful tool to characterize the uncertainty involved in different civil engineering problems. This study aims to develop prediction approaches that identify the abnormal data and structural behaviour as well as reconstruct the missing data within the Bayesian framework.