小波变换具有特别适合于对包括多个频率分量的非平稳序列的变形监测数据进行

Wavelet transform has the advantages of being particularly suitable for noise removal processing on deformation monitoring data of non-stationary sequences including multiple frequency components, multi-decomposition analysis and time-frequency positioning for extracting deformation information. In the case of using wavelet transform to perform noise removal processing on deformation monitoring data, different deformation monitoring data have different characteristics such as sampling rate, noise pollution degree, etc., including the selection of different noise removal parameters based on microwave selection. This issue has always been the focus of research. At present, many experts and scholars have conducted a lot of research on the selection of the optimal decomposition layer and the value of A and B, but there is no system standard for selecting the best microwave base. On the other hand, the deformation information of the engineering body often appears in the changes of the frequency components of the monitoring data. Based on this fact, the extraction of deformation information in the deformation monitoring data is where the frequency of the acquired data signal generates specificity. In the process of using the single-band reconstruction algorithm to analyze the data signal, because the inherent frequencies in the Mallat algorithm are mixed, the wrong feature information is often extracted, and the obtained reconstruction data also changes in length along with the filter and convolution Then the problem of boundary effect arises.

Wavelet transformation is particularly suitable for noise removal of deformation monitoring data including non-smooth sequences with multiple frequency components, and the advantages of multi-decomposition analysis and time frequency positioning of deformation information. In the case of noise removal treatment of deformation monitoring data using wavelet transformation, different deformation monitoring data have different characteristics such as sampling rate and noise pollution degree, including different noise removal parameters must be selected based on microwave selection. This problem has always been the focus of research. At present, many experts and scholars have done a lot of research on the optimal number of decomposition layers and the selection of A-B values, but there is no standard for selecting the best microwave base. On the other hand, the deformation information of the engineering body often appears in monitoring the change of the frequency component of the data. Based on this fact, the extraction of deformation information in deformation monitoring data is a specific place where the frequency of data signal is obtained. In the process of analyzing data signals using single-band reconstruction algorithms, because the frequencies inherent in Mallat algorithms are mixed, incorrect feature information is often extracted, and the resulting reconstructive data changes length with filters and convolution, resulting in boundary effects.

Wavelet transform is especially suitable for noise removal, multi decomposition analysis and time-frequency location of deformation monitoring data of non-stationary series including multiple frequency components. In the case of using wavelet transform to remove the noise of deformation monitoring data, different deformation monitoring data have different characteristics, such as sampling rate, noise pollution degree and so on. This problem has always been the focus of research. At present, many experts and scholars have done a lot of research on the selection of the optimal number of decomposition layers and a / b value, but there is no systematic standard to select the optimal microwave base. On the other hand, the deformation information of engineering body often appears in the change of frequency component of monitoring data. Based on this fact, the extraction of deformation information in deformation monitoring data is the place where the frequency of data signal is obtained to produce specificity. In the process of using single band reconstruction algorithm to analyze the data signal, because the inherent frequency of Mallat algorithm is mixed, the wrong feature information is often extracted, and the reconstructed data also changes the length with the filter and convolution, resulting in the problem of boundary effect.<br>