Helmert方差分量估计的虚拟观测法  

Virtual observation method based on Helmert variance component estimation

作　　者：朱军桃[1] 孙昌瑜[1] 焦元元[1] 

机构地区：[1]桂林理工大学测绘地理信息学院/广西空间信息与测绘重点实验室,广西桂林541004

出　　处：《测绘科学》2013年第5期152-153,161,共3页Science of Surveying and Mapping

基　　金：国家自然科学基金(41071294)

摘　　要：本文挖掘先验信息作为虚拟观测并与实际观测建立联合误差模型,对此模型利用Helmert方差分量估计理论求解。通过模拟算例、工程实例的对比和分析,得出了该法对参数的拟合效果优于最小二乘法、补偿最小二乘法,论证了其在物理解释方面和拟合效果方面的可行性和有效性。The paper excavated a priori information as virtual observation and the actual observation establish the error model, used Helmert variance component estimation theory to solve the model. Practical examp]es showed that this method was better than least squares or compensative least squares method in fitting results with its feasibility and efficiency in physical interpretation and fitting results.

关 键 词：半参数回归模型 虚拟观测 HELMERT方差分量估计 

分 类 号：P258[天文地球—测绘科学与技术] P207