有界不确定性平差模型的迭代算法

针对现有的有界不确定性平差模型算法较为复杂且没有顾及权重的问题,该文提出了一种无需奇异值分解的迭代算法及其一种加权方法。直接采用了迭代算法求解有界不确定性平差模型的min-max准则,推导出了未知参数估值,算法概念简单,易于实现,收敛速度更快。基于该文提出的迭代算法,当系数矩阵和观测向量各自均不等权时,采用了一种加权方法,并推导了其解算过程。算例结果表明:该文提出的迭代算法是可行的,并且解算效率更高;加权后的迭代算法是有效的。

Iterative algorithm for adjustment model with bounded data uncertainties

According to the fact that the existing algorithm for adjustment model with bounded data uncertainties is complex and doesn't take into account the weight,this paper presented an new iterative algorithm and a weighted method without applying singular value decomposition to calculate the min-max criterion of the adjustment model.The unknown parameter estimation was derived from the directly iterarive algorithm.And the algorithm is simple in the concept,easy to implement,and fast in the convergence.Moreover,when the weight was different for each element of coefficient matrix and observation vector,the contribution obtained the unknown parameter solution,which was based on the use of the proposed algorithm and weighted method.The results illustrated that the proposed algorithm could be practiced and had higher calculation efficiency;the efficiency of the weighted algorithm was proved either.