测量参数估值和误差分析的最大最小准则法

Using Minimax Principle to Estimate Surveying Parameters and Perform Accuracy Analysis

本文建议一种平差方法.与最小二乘平差不同,它除了利用观测值以外,还使用我们对观测质量的评价即测量误差范围(可以根据所采用的观测方法的精度以及测量仪器的特性来确定).应用线性规划理论的单纯形法,该方法是以所估参数的最大可能误差最小为目标,最后可得未知参数的最佳估值及其最大可能的误差.

This paper proposes a new adjustment method.Different from the least- squares method,the new one computes a unique best estimate for a set of unknown parameters from both the observations and the known bounds on the observation errors,which can be determined in terms of the Precision of observation method and the properties of surveying instruments used. Using the simplex algorithm of the theory of linear programming,the method leads to an estimate minimizing the maximal possible errors of its components.