GPS相关观测量稳健估计探讨

用GPS采集的观测数据与常规方法有很多不同之处,主要是所建立的函数存在模型误差,其次,随机模型也存在误差,一般指的是模型中的系统误差和粗差。在测量数据服从正态分布情况下,最小二乘估计具有最优无偏的统计性。当测量数据受到粗差干扰而偏离正态分布时,由于最小二乘估计不具备抗粗差的能力,就要寻找具有抗粗差能力的估计理论和方法。通过实验数据分析和比较,认为等价方差-协方差稳健估计法对粗差具有较强的抵抗能力,从而可以提高大规模整体平差成果精度。

Probing into robust estimation of GPS correlated observation data

Collecting observation datas by GPS are very different from the conventional methods,mainly the function model errors exist,second,there are stochastic model errors that include blunder and systematic error.When surveying data obey normal circumstances,RLS theory is the best.If the measurements including the blunder that go far from the normal circumstances,RLS theory do not have the rough-poor,it is necessary to found anti-poor estimate method as a solution gross error problem.Through analysis of the exper...