摘要
Vondrak滤波是基于测量数据本身的一种数据处理方法,该方法能够在未知拟合函数的情况下,通过选择不同的平滑因子控制数据平滑的程度,对测量数据序列进行合理平滑。本文从Vondrak滤波的基本原理出发,将Vondark滤波与最小二乘估计统一起来,推导出基于Helmert方差分量估计的Vondark滤波方法(简称HVF法)。通过对模拟数据资料的分析,表明该方法能合理有效的选取平滑因子,最大限度的削弱数据序列中的随机误差,将信号与噪声合理分离。
Vondrak Filter is a method based on the observed data, which can smooth data through selecting the smoothing factor to control its degree even in the condition of unknown serials reasonably fit function. This paper, from the principle of Vondrak Filtering and together with the least square, estimates this way based on Helement type variance component estimation （HVF）. Numerical simulation experiments indicate that this method can select the smoothing factor reasonably and mitigate the random error of measurement effectively. Also, it can separate the signal from the noise successfully.
出处
《工程勘察》
CSCD
北大核心
2007年第6期42-45,48,共5页
Geotechnical Investigation & Surveying
基金
国家自然科学基金资助项目(40574003)
武汉大学地球空间环境与大地测量教育部重点实验室开放基金资助项目(905152533-05-01)
