基于多面核函数配置型模型的参数估计

Parameter Estimation Based on Multi-quadric Collocation Model

  • 摘要: 针对同时出现非随机参数、已测点和未测点信号的配置型模型精确建立协方差函数的困难,在研究和应用最小二乘配置和多面函数拟合法的基础上,将两法综合,提出了多面核函数配置法,导出了这种平差方法的解算公式,并与最小二乘配置、多面函数拟合作了比较分析。

     

    Abstract: On the basis of the research of least squares collocation and multi-quadric equations interpolation,a new estimation method is presented.Penalised least squares technique is proposed to estimate parameters and the multi-quadric equation is chosen as R matrix.At the same time,the relationship of the observed points and non-observed points is also expressed as multi-quadric equation.The formulas of this method are deduced.Without the covariance equation of sign,the result obtained by the new method corresponds with that by the collocation method.It is also the expand of multi-quadric equations interpolation.So the new method can enlarge the use of collocation model.

     

/

返回文章
返回