利用半参数模型精化二次曲面GPS高程异常拟合模型  

Refining the GPS Height Anomaly Fitting Model of Quadric Surface with Semi-parametric Regression Model

作　　者：袁浩涛 张欢欢 张俊[1] YUAN Haotao;ZHANG Huanhuan;ZHANG Jun(School of Mining,Guizhou University,Guiyang 550025,China;School of Foreign Languages,Guizhou University,Guiyang 550025,China)

机构地区：[1]贵州大学矿业学院,贵州贵阳550025 [2]贵州大学外国语学院,贵州贵阳550025

出　　处：《测绘与空间地理信息》2019年第1期37-40,共4页Geomatics & Spatial Information Technology

基　　金：贵州大学大学生创新创业训练计划项目(贵大校创字2017(031));贵州大学教学改革项目:(JG201630);贵州大学测绘科学与技术研究生创新基地建设项目(贵大研CXJD[2014]002)资助;贵州省科学技术基础研究项目(黔科合基础[2017]1054)

摘　　要：针对利用多项式建立的GPS高程拟合模型不能很好地拟合高程异常变化趋势面的问题,提出了在常规最小二乘多项式模型的基础上,引入一个非参数模型补偿项,并参考重力测量中最小二乘配置模型方法,建立高程异常趋势面的半参数拟合模型。以小区域GPS采集数据为例,并分别运用两种模型进行拟合与推估,结果表明,基于半参数模型的高程异常拟合与推估效果更好。The GPS elevation fitting model,established based on polynomial,cannot effective fit height anomaly trend surface changing.So,consulting the model of least squares in gravity survey,introducing a nonparametric model compensation term to the conventional least square polynomial model to establish a semi-parametric Fitting Model that fitting height anomaly trend surface.Taking the data measured by GPS in small area as example and using two models mentioned above for fitting and pushing,the results illustrate the semi-parametric model is superior to the least square polynomial model.

关 键 词：半参数模型 二次曲面 补偿最小二乘 高程异常 

分 类 号：P228.4[天文地球—大地测量学与测量工程]