Bridge condition assessment from long-term monitoring by means of Bayesian hypothesis test

Fifth International Symposium on Life-Cycle Civil Engineering (IALCCE 2016), 16-19 October 2016, Delft, The Netherlands.This study presents an approach to reduce effects of environmental and operational factors on long-term monitoring data of bridges. The Bayesian approach comprising both Bayesian regression and Bayesian hypothesis test is applied to investigate monitoring data of an in-service seven-span plate-Gerber bridge. This study considers time-varying temperature and vehicle weights as environmental and operational factors respectively. Vehicle weights were measured utilizing a bridge weigh-in-motion (BWIM) system installed on the bridge. All data was taken from a healthy bridge, since no damage and deterioration was reported during the monitoring period. Observations through the study demonstrated that considering both temperature and vehicle weight as environmental and operational factors in Bayesian regression led to improved regression results than that considering only temperature. It also showed that monitoring the data observed at a specific time could reduce influence of traffic in long-term monitoring. In the Bayesian hypothesis testing utilizing data from the healthy bridge, the bridge was judged as healthy