VMD在GNSS高程时间序列中的周期性分析  

Periodicity Analysis of VMD in GNSS Elevation Time Series

作　　者：谯江容 饶爽 关乃瑜 雷新强 穆永洪 王新鹏 QIAO Jiang-rong;RAO Shuang;GUAN Nai-yu;LEI Xin-qiang;MU Yong-hong;WANG Xin-peng(School of Mining,Guizhou University,Guiyang Guizhou 550025,China;GNSS Research Center,Wuhan University,Wuhan Hubei 430079,China)

机构地区：[1]贵州大学矿业学院,贵州贵阳550025 [2]武汉大学卫星导航定位技术研究中心,湖北武汉430079

出　　处：《地矿测绘》2024年第4期1-6,共6页Surveying and Mapping of Geology and Mineral Resources

基　　金：贵州大学大学生“SRT计划”项目(贵大SRT字(2022)322号);贵州大学实验室开放项目(SYSKF2024-056)。

摘　　要：由于各种地球物理效应的影响,大部分IGS站的时间序列存在明显的非线性变化,其中高程方向的坐标变化更为明显,周期性也更为复杂。由此,引入VMD方法,对4个IGS站的高程时间序列进行分解,得出不同尺度的IMF,然后利用傅里叶变换对各个分量进行处理得到各站的频谱图,探测出各站都存在的周期项为2、4、6天,部分站含有年周期、月周期。EMD所得的周期项中各站存在季周期,月周期、双月周期、年周期、两年周期、四年周期。实验结果表明,与EMD相比,VMD在设置合适的收敛条件和迭代寻优的情况下可以有效地避免模态混叠、避免噪声干扰和抑制端点效应,更能自适应、准确地提取GNSS高程时间序列中的周期项。Due to various geophysical effects,the time series of most IGS stations exhibit significant nonlinear changes,with more pronounced coordinate changes in the elevation direction and more complex periodicity.From this,the VMD method to decompose the elevation time series of four IGS stations was introduced in order to obtain IMF of different scales.Then,fourier transform was used to process each component to obtain the spectrum of each station,and it was found that the period terms of each station were 2,4,and 6 days,with some stations containing annual and monthly periods.The period term obtained by EMD included quarterly,monthly,bimonthly,annual,two-year,and four-year periods for each station.The experimental results have showed that VMD can effectively avoid modal aliasing,noise interference,and suppress endpoint effects by setting appropriate convergence conditions and iterative optimization,and more adaptively and accurately extract periodic terms from GNSS elevation time series,compared with EMD.

关 键 词：EMD VMD 周期项 高程时间序列 

分 类 号：P228[天文地球—大地测量学与测量工程]