摘要
利用罚最小二乘原理构造加权惩罚平方和,将半参数回归模型与最小二乘配置结合起来,来处理GPS高程问题,导出了配置型模型中正规化矩阵正定时参数平差的计算方法,用直接法得到了参数和信号的估计量,给出了相应的公式。其次,通过选取合适的平滑因子,能使残差的分布更接近其真实分布,并用一个实例说明了此方法的有效性。
In this paper, based on penalized least squares, the penalized weighted sum of squares is set up. By combining semiparametric survey adjustment Models with the least squares collocation to deal with GPS elevation, the calculation method of a positively definite regularity matrix is deduced and the corresponding results are obtained. According to the foundational properties of the random errors, how to choice a reasonable smoothing parameter and regularizer matrix is studied. The adaptation of this technique is more flexible than the traditional model.
出处
《武汉工业学院学报》
CAS
2004年第2期99-101,108,共4页
Journal of Wuhan Polytechnic University
基金
国家自然科学基金资助项目(40274005)
湖北省教育厅重点科研计划项目(203X129)
