摘要
将近年来统计界提出的并有较深入理论研究的t型估计引入测量数据处理中,提出了Gauss-Markov模型的t型抗差估计,给出了相应的EM算法,并进行了模拟计算。计算结果显示,t型抗差估计计算快捷稳定,收敛性好,并且t型抗差估计受粗差影响不大,具有较好的抗差能力。
t-type estimation was introduced to geodetic adjustments as a robust estimation, t-type robust estimation in Gauss-Markov Model was established and its corresponding EM algorithm was derived. A simulation example was given, which demonstrated that t-type robust estimation converges rapidly and stably, and suffered little from gross errors, showing good robustness.
出处
《测绘科学技术学报》
北大核心
2008年第1期57-60,共4页
Journal of Geomatics Science and Technology
基金
国家自然科学基金项目(40474007)
国家杰出青年科学基金项目(40125013)
“基础地理信息与数字化技术”山东省重点开放实验室课题(SD040202)
河南省自然科学基金项目(0511010100)
关键词
抗差估计
t型估计
EM算法
robust estimation
t-type estimation
EM algorithm
