摘要
本文基于广义EIV(Generalized Error-In-Variables,GEIV)模型,提出一种矿区高程异常拟合的无缝推估算法。该方法可以顾及观测值之间的相关性,实现求解二次多项式系数的同时并推估出未知高程异常的一步法。通过实验表明,相比较于GEIV模型和Gauss-Markov(GM)的两步法,本文提出的方法求解多项式系数的均方根误差与广义EIV模型的精度一致,且优于GM模型,推估未知高程异常的精度分别提升50%和10%。
In this paper,a seamless prediction algorithm for fitting elevation anomalies in mining areas is proposed based on the Generalized Error-In-Variables(GEIV)model.The algorithm achieves a one-step method of solving quadratic polynomial coefficients while predicting unknown elevation anomalies,taking into account the correlation between observations.Experiments show that compared with the step-wise method of GEIV model and the Gauss-Markov(GM),the method proposed in this paper solves the root mean square error of the polynomial coefficients with the same accuracy as the generalized EIV model and better than the GM model,and the accuracy of predicting unknown elevation anomalies is improved by 50%and10%,respectively.
作者
杨根新
王友昆
谢正明
Yang Genxin;Wang Youkun;Xie Zhengming(Yunnan Land and Resources Vocational College,Kunming 652501,China;Kunming Surveying and Mapping Institute,Kunming 650051,China)
出处
《工程勘察》
2023年第8期46-51,共6页
Geotechnical Investigation & Surveying
基金
云南省教育厅科学研究基金项目(2021J0964)
自然资源时空大数据应用技术科技创新团队(2021KJTD03)。
