卫星重力梯度数据粗差探测的组合法研究  

Research on Combination Method for Outlier Detection of GOCE Satellite Gravity Gradient Data

作　　者：李晓翔 罗晓丹 LI Xiaoxiang;LUO Xiaodan(Institute of Natural Resources and Remote Sensing of Guangxi Zhuang Autonomous Region,Nanning 530023,China)

机构地区：[1]广西壮族自治区自然资源遥感院,广西南宁530023

出　　处：《测绘与空间地理信息》2022年第S01期252-255,共4页Geomatics & Spatial Information Technology

摘　　要：对GOCE卫星重力梯度实测数据粗差探测做了相关研究,探讨了进行粗差探测的相关理论与方法,包括重力梯度数据粗差探测的迹准则、连续性准则,以及Dixon检验法、小波探测法、组合法等几种粗差探测方法,重点探讨粗差探测组合方法。在连续性准则下运用上述探测方法对实测数据分别进行了粗差探测计算,通过调整置信度、阈值等参数,分析这些参数对各种探测方法结果的影响。研究结果表明:1) Dixon检验方法算法简单、快速,可用于快速解的粗差探测;2)小波收缩阈值方法探测质量较高,可用于精密解的粗差探测;3)小波和Dixon检验的组合探测法探测质量优于2种单一探测方法。As an important part of satellite gravity gradient data pre-processing, the disciplines and the methods of the outlier detection are discussed in this thesis, including trace criterion, continuity criterion, Dixon test and wavelet outlier detection algorithm, while the combination methods are focused. Gravity gradient data was obtained by applying from the ESA, the GOCE(Gravity Field and Steady-State Ocean Circulation Explorer) satellite. By modifying the parameters such as the window width and threshold to analyze these parameters’ influence on the results, the following conclusions can be obtained:(1) Dixon test is sample and fast that can be the Quick-look solution.(2) Precise solution requires the outlier detection to have a better result. Wavelet outlier detection algorithm and hybrid method of wavelet and Dixon test are adapted to this solution;(3) Combination method of wavelet and Dixon test has a better result as compared with the wavelet independent and Dixon independent, which can be used in the GOCE pre-processing.

关 键 词：卫星重力梯度 预处理 粗差探测组合法 小波收缩阈值 

分 类 号：P223.6[天文地球—大地测量学与测量工程]