边坡GNSS变形监测数据去噪方法研究  

Denoising method for slope GNSS deformation monitoring data

作　　者：徐华明 孟飞飞 俞春燕 XU Huaming;MENG Feifei;YU Chunyan(Zhejiang Geotechnical Engineering Survey Institute Co.,Ltd.,Shaoxing,Zhejiang 312000,China;Hangzhou Xiaoshan Urban and Rural Surveying and Mapping Co.,Ltd.,Hangzhou,Zhejiang 311200,China;Shaoxing Keqiao District Surveying and Mapping Institute,Shaoxing,Zhejiang 312030,China)

机构地区：[1]浙江土力工程勘测院有限公司,浙江绍兴312000 [2]杭州萧山城乡测绘有限公司,浙江杭州311200 [3]绍兴市柯桥区测绘院,浙江绍兴312030

出　　处：《测绘技术装备》2023年第3期31-36,共6页Geomatics Technology and Equipment

摘　　要：为了能更加有效、精细地从全球导航卫星系统(GNSS)变形监测数据中提取真实信息,准确描述变形体的变形趋势,本文提出了一种自适应噪声完备集合经验模态分解(CEEMDAN)去噪方法。该方法去噪的流程为:首先,对目标信号进行CEEMDAN分解,得到若干本征模态函数(IMF);其次,将多尺度排列熵引入分解IMF分量的高频与低频区分中,保留低频IMF分量与残余项,使用小波分析去噪方法对高频分量做进一步处理;最后,将小波分析去噪后的高频分量、低频分量与残余项重构,得到最终的重构信号。本文还使用边坡GNSS监测数据对提出的方法进行验证。结果表明,相较于传统的经验模态分解(EMD)方法的去噪结果和小波分析去噪结果,本文方法的去噪结果的相关系数与信噪比更大,均方根误差更小,表现出更优的去噪效果,验证了本文方法的可靠性与优越性。In order to effectively and precisely extract real information from GNSS deformation monitoring data,and describe the deformation trend accurately,this paper proposes CEEMDAN denoising method,its denoising process is as follows:firstly,CEEMDAN decomposition is performed on the target signal to obtain several IMF;Secondly,multiscale permutation entropy is introduced into the high frequency and low frequency discrimination of IMF components,preserving the low frequency IMF components and residual terms,and using wavelet analysis denoising method to further process the high frequency components;Finally,reconstruct the high frequency component,low frequency component,and residual term after wavelet analysis denoising to obtain the final reconstructed signal.This paper also uses slope GNSS monitoring data to validate the proposed method.The results show that,compared with the denoising results from the traditional EMD method and wavelet analysis,the denoising results form this method have larger correlation coefficient and signal to noise ratio,smaller root mean square error and better denoising effect,which proves the reliability and superiority of this method.

关 键 词：变形监测 去噪 多尺度排列熵 自适应噪声完备集合经验模态分解 小波分析 

分 类 号：P258[天文地球—测绘科学与技术]