小波函数在沉降监测中信噪分离效果对比

Comparison of Wavelet Function in Signal-noise Separation Effect for Settlement Monitoring

变形监测的数据中通常不可避免地存在着各种不易消除的噪声,消除这些噪声,提取真实的形变信息是变形分析中一项重要工作。小波变换是一种时频联合分析方法,广泛应用于信号去噪。基于MATLAB进行了叠加高斯白噪声、突变信号的信号仿真,采用不同小波函数对仿真信号进行去噪试验,以均方根误差、估值偏差、信噪比作为衡量指标。对去噪效果进行了对比分析。最后基于仿真试验所得结论,选择最优小波函数对南京地铁春江新城站沉降观测数据进行了去噪处理,得到了较好的去噪效果。

Inevitably, there are various noises which are not easy to be eliminated in the deformation mon toring data, so an important work of the deformation analysis is to eliminate these noises and extract the real deformation information. The wavelet transform is a time frequency conjoint analysis method, which s widely used in signal denoising. This paper carries out the signal simulation of the overlaid Gaussian white noise and transient signal based on MATI.AB, and uses the different wavelet functions to conduct the denoising test on the simulation signal by taking the root mean square error, valuation deviation and signal to noise ratio as the measurement index. The optimal wavelet function is selected to denoise the denoising effect is compared and analyzed. Finally, the settlement observation data of the Nanjing Chunjiang New Town Metro Station based on the conclusions obtained from the simulation test, and the better de noising effect is obtained