Statistical Analysis of Vibration Modes of a Suspension Bridge Using Spatially Dense Wireless Sensor Network

Abstract

A spatially dense wireless sensor network was designed, developed and installed on a long-span suspension bridge for a 3-month deployment to record ambient acceleration. A total 174 sets of data (1.3 GB) were collected from 64 sensor nodes on the main span and south tower of the Golden Gate Bridge. Analysis of the vibration data using power spectral densities and peak picking provide approximate estimates of vibration modes with minimal computation. For more detailed analysis of the data, autoregressive with moving average models (ARMA) give parametric estimates of vibration modes for frequencies up to 5 Hz. Statistical analysis of the multiple realizations give the distributions of the vibration frequencies, damping ratios, and mode shapes and 95% confidence intervals. The statistical results are compared with vibration properties using the peak picking method and previous studies of the bridge using measured data and a finite-element model. Analysis of the ambient vibration data and system identification results demonstrate that high spatial and temporal sensing using the wireless sensor network give a high resolution and confidence in the identified vibration modes. The estimation errors for the identified vibration properties are generally low, with frequencies being the most accurate and damping ratios the least accurate.