组合IQR和小波分解的基坑监测数据去噪方法研究

Research on Denoising Method of Foundation Pit MonitoringData Based on IQR and Wavelet Decomposition

在国家基础设施的建设过程中,大型基坑的变形监测尤为重要。然而目前基坑监测数据处理存在一定的主观性,且异常点通常难以完全消除。为加强数据处理的客观性及鲁棒性,提出一种组合四分位距(IQR)与小波分解的去噪方法。首先利用IQR和小波分析对原始数据进行去噪,在此基础上,以残差最小为目的建立非线性规划组合去噪模型,利用样本数据求解模型系数,进而对实验数据进行精去噪处理。结果表明,该组合方法能有效消除噪声影响,提高形变数据去噪的精准度。

In the process of national infrastructure construction,the deformation monitoring of large foundation pit is particularly important.However,at present,the data processing of foundation pit monitoring is subjective,and the abnormal points are usually difficult to completely eliminate.In order to strengthen the objectivity and robustness of data processing,this paper proposes a denoising method that combines the IQR and wavelet decomposition.Firstly,IQR and wavelet analysis are used to denoise the original data.On this basis,a nonlinear programming combined denoising model is established for the purpose of minimizing the residual error.The model coefficients are solved by using sample data,and then the experimental data are finely denoised.The results show that this combined method could effectively eliminate the noise and improve the denoising of deformation data.