Bridge Monitoring with a Density-Based Clustering Automated Operational Modal Analysis Methodology
摘要
Operational Modal Analysis (OMA) is a technique used in structural dynamics to determine a structure’s modal parameters (natural frequencies, damping ratios, and mode shapes) under operational conditions. Automating OMA enables continuous data processing from long-term monitoring activities to be performed with repeatability and without user-interaction. This work presents an Automated Operational Modal Analysis (AOMA) methodology that employs the DBSCAN clustering technique, integrated with verification steps based on expert modal analysis judgment. The developed AOMA methodology was applied to monitor two bridges: the Infante Dom Henrique bridge, monitored over one month, and the Z24 bridge, monitored over one year. A modal tracking procedure was implemented to consistently correlate a reference set of modes for each bridge throughout the extended measurement campaigns. The application of AOMA demonstrated high accuracy and effectiveness in long-term monitoring, confirming its adaptability and robustness for different types of bridges. Additionally, this study considers factors that may affect the AOMA process, such as sensor noise levels and low excitation. The dependency of modal parameters on environmental conditions is also explored.