观测数据中的粗差定位与定值算法

Algorithm for Location and Evaluation of Gross Errors in Surveying Data

根据粗差对验后方差的影响,将误差理论与可靠性理论相结合,采用逐步搜索的方法探测粗差的位置;提出方差比极大、偶然误差平方和极小法进行多维粗差定位与定值。实验表明,用两次相邻搜索的方差的比出现峰值作为搜索结束的条件,在偶然误差平方和极小的条件下进行估值是可行的。

According to the effects of gross error on posterior variance, on the basis of the er- ror theory and reliability theory, the multidimensional gross errors are located and evaluated with the method of maximum variance ratio and the minimum sum of the squares of random error. Through a large number of random simulation calculations, the threshold, which jud- ges the end conditions of the search, can be obtained. After gross errors are searched and lo- cated, an estimating model is given under the conditions of the minimum sum of the squares of random error. The estimating model is tight and not approximate. After many experi- ments, the result showes that using the method not only can detect the big gross error can also detect small gross error.