方差分量估计的抗差总体最小二乘算法

应用等价权函数建立总体最小二乘抗差估计算法,算法的有效性受到随机模型的影响。为减小随机模型误差对抗差算法的影响,提出方差分量估计确定总体最小二乘随机模型的思想;利用等价权函数对观测元素进行迭代定权,分类确定等价权函数的域值,提高权函数抗差算法的有效性。最后,通过算例验证算法可行性,结果表明本文算法是有效的。

Solution for robust total least squares based on variance component estimation

Traditional algorithms of robust total least squares are based on weight function,while the stochastic model has influence on their effective.To overcome the influence that random error of stochastic model has on robust algorithms,the solution to application variance component estimation for computation the stochastic model of total least squares is proposed in this contribution.Besides,to improve the effective of robust algorithm based on weight function,the method of computing the threshold value of weight function separately is discussed especially,while weight function is applied to determining the weight of observation elements iteratively.At last,the feasibility of the proposed algorithm is verified through an example,and its numerical results demonstrate that the algorithm is effective.