Estimação estatica de estado em sistemas de energia eletrica utilizando fatoração ortogonal QR

AUTOR(ES)
DATA DE PUBLICAÇÃO

1994

RESUMO

This work is concerned with computational aspects of the problem of static state estimation and bad data processing, in electrical power systems, using the nonlinear weighted least-squares technique. Modified Newton type methods, for solution of the nonlinear least-squares problem, are briefly presented and analysed. For solution of the sequence of linear least-squares problems, produced by modified Newton methods, the methods normal equations, hybrid normal equations and orthogonal QR factorization are presented and analysed. AIgorithms using Givens rotations, for computation of QR factors in the methods hybrid normal equations and orthogonal QR factorization are presented and discussed. Comparative results are presented concerning the expected precision of the analysed algorithms together with estimates of their required computational effort - FLOPS. For bad data processing, and analysed some algorithms for error detection using chi-square test, error identification using normalized residues and error elimination by updating the pertinent matrices of the normal equations and QR factorization are presented. AIgorithms, compatible with the methods hybrid normal equations and orthogonal QR factorization, for computation of normalized residues using the R matrix and for updating QR factors using conventional Givens rotations are proposed

ASSUNTO(S)

sistemas de energia eletrica

Documentos Relacionados