Study On Structural Vibration Frequencies Of Bridges Based On GPS And Accelerometer

It is a key task to monitor dynamic deformations of bridges in structural health monitoring, and also the necessary measures to ensure the safety of bridges. Global positioning system is the main technology in structural health monitoring, playing a more important role in dynamic deformation monitoring as a result of unique advantages. Based on GPS and accelerometer in structural vibration monitoring of bridges, the present studies focus on multipath extraction and correction model, vibration frequencies extraction of bridges based on high sampling GPS data, deformation monitoring models based on GPS and accelerometer, multi-rate Kalman filtering and data filtering technique. The major contributions are provided as follows:(1) By distinguishing the error sources and characteristics in measurements of GPS dynamic deformation monitoring technique, the means were proposed to mitigate the corresponding errors and improve the monitoring accuracy. The errors caused by ephemeris can be eliminated by carrier phase relative positioning technique. Caused by the ionosphere and troposphere delays, the errors in GPS measurements can be mitigated using the error correction model or the observation difference method. The errors caused by antenna phase center can be corrected using the observation difference method or antenna correction model. Other errors in GPS measurements due to the earth’s rotation and tides can be reduced by the correction model. The errors caused by relativistic effects can be eliminated by adjusting the clock frequencies of satellites.(2) Methods of extracting and correcting multipath errors were developed. Multipath effects are the major error sources for the short-baseline and high-precision dynamic deformation monitoring. Considering the repeatability characteristics in multipath effects, multipath extraction models based on EMD, WT and Vondrak filter were proposed and dynamic data between the two successive days were processed. The research demonstrates that the multipath effects have strong correlation between two successive days and the stability and the accuracy are markedly improved after multipath effects being removed.(3) A scheme for vibration frequencies extraction of bridges from high sampling GPS data were proposed. A Bi-linear Chebyshev high-pass filter is designed to isolate the quasi-static movements, the FFT algorithm and peak-picking approach are applied to extract the vibration frequencies, and a GPS data accumulation counter is suggested for real time monitoring applications. To understand the change in the structural characteristics under different loadings, the deformation results from three different loading conditions are presented. The results show that GPS can not only capture absolute 3 D deflections reliably, but also can be used to extract the frequency response accurately. The strategy employed in this chapter can be used for the future safety alarm of the bridge in real time.(4) A method integrating GPS and accelerometer sensors to monitor structural dynamic displacements of bridges was developed. Considering the multipath and random errors, a noise filter model based on the Empirical Mode Decomposition was built to extract the low-frequency component of GPS measurements. Frequency domain integration was used to reconstruct displacement using acceleration measurements. Then the high-frequency component from acceleration measurements and the low-frequency component from GPS measurements were integrated to extract the whole dynamic response and deformation information of bridges.