Data preprocessing based on wavelet and its application in settlement monitoring for urban subway

Urban subway is an important underground transportation facility, settlement monitoring and settlement analysis as well as settlement prediction are usually carried out in the construction process. In order to obtain a better settlement analysis and effect on predictions, the reliability of monitoring data should be first tested and analyzed. For the problem of data preprocessing in settlement monitoring, according to the basic principles of wavelet, detection method of settlement anomaly is studied and wavelet denoising is further researched. Combining with the settlement monitoring data in Nanjing subway, db3 wavelet is applied to three-layer wavelet decomposition based on sequence of the original settlement, through the module maximum value of high-frequency part, the settlement anomaly is analyzed and judged, and further the noise of high-frequency part is reduct,the new signal is reconstruct, the new sequence of settlement for settlement analysis and prediction is obtained. The method this paper researched is a pre-testing method of data reliability testing,and also has some reference value in other measured data preprocessing.