Wavelet De-noising of GNSS Based Bridge Health Monitoring Data

GNSS signal multipath occurs when the GNSS signal reflects off objects in the antenna environment and arrives at the antenna via multiple paths. A bridge environment is one that is prone to multipath with the bridge structure, as well as passing vehicles providing static and dynamic sources of multipath. In this paper, the Wavelet Transform (WT) is applied to bridge data collected on the Machang cable stayed bridge in Korea. The WT algorithm was applied to the GNSS derived bridge detection data at the mid-span. Up to 41% improvement in RMS was observed after wavelet shrinkage de-noising was applied. Application of this algorithm to the torsion data showed significant improvement with the residual average and RMS decreased by 40% and 45% respectively. This method enabled the generation of more accurate information for bridge health monitoring systems in terms of the analysis of frequency, mode shape and three dimensional deflections