Abstract：The weighted total least squares problem not only takes into account the errors of coefficient matrix but also consideres that the heteroscedasticity between the observable vector and the coefficient matrix.Seeing that the traditional iteration algorithm for solving the problem is too complicate,this paper considered the errors of observable vector as a function of the errors of coefficient matrix and the parametes.Through the linearization,it became adjustment of observation equations problems, hence the Lagrange multiplier method was used to carry out the iteration functions for this algorithm.And then,two numerical examples were given, and it was shown that the adjustment results with our approach is identify with Schaffrin & Wieser method.
