Automated identification of modal properties in a steel bridge instrumented with a dense wireless sensor network

Wireless sensing technology has made it possible to instrument large civil structures, such as bridges, with dense networks of wireless sensors. Because many wireless sensing platforms integrate data processing capabilities with high-precision analog-to-digital converters, it is possible to autonomously process sensor data without the need for a central data server. By employing parallel processing techniques on typically serial modal analysis methods, a network of sensors can be used to autonomously determine modal properties (modal frequencies, damping ratios, and mode shapes). In this study, distributed versions of three output-only modal identification techniques (peak picking, random decrement, and frequency domain decomposition) are modified for implementation within a wireless sensing platform designed for bridge monitoring. A network of wireless sensors are deployed on 30m long pedestrian bridge; this wireless monitoring system is used to autonomously determine the bridge’s modal properties.