基于经验模态分解削弱多波束残余误差的方法  

Weakening effect of residual error based on EMD

作　　者：赵祥鸿 暴景阳[1] 黄辰虎[2] 黄贤源[2] 陆秀平[2] 

机构地区：[1]海军大连舰艇学院海洋测绘系,辽宁大连116018 [2]海军海洋测绘研究所,天津300061

出　　处：《海洋通报》2017年第6期662-667,共6页Marine Science Bulletin

基　　金：国家自然科学基金(41474012);国家重大科学仪器设备开发专项(2011YQ12004503)

摘　　要：针对多波束水深数据中存在的系统性残余误差,提出了基于经验模态分解方法来削弱残余误差的方法:首先利用经验模态分解方法对多波束测深数据作一维分解,将非线性、非平稳的多波束测深数据分解成准线性子波,然后构建水深数据趋势项与残余项,利用中央波束趋势项建立整体数据趋势项,最后加以水深数据残余项还原海底地形,削弱残余误差影响。通过实测多波束测深数据验证方法的有效性。This paper puts forward a method for weakening the effect of residual errors of multibeam bathymetric data based on Empirical Mode Decomposition（EMD）.Multibeam bathymetric data is deposed into one-dimensional linear waves by EMD.In that way,the trend and residual of seabed topography is constructed.The trend of whole data is constructed using the center beam trend.Then submarine topography is restored by merging of the data trend and the residual one,which weakens the influence of residual errors.The paper also shows an experiment to prove effectiveness of the method.

关 键 词：经验模态分解(EMD) 多波束 残余误差 趋势项 残余项 

分 类 号：P733.2[天文地球—物理海洋学] P229[天文地球—海洋科学]