Solution Methods for Sparse Weighted Least Squares Problems in Power System Static State Estimation

摘要

The work described is part of a project to develop a state estimation for the main Norwegian electric power grid. The approach followed is to formulate the estimation problem as a nonlinear weighted least squares problem. This is then solved via a succession of linearized sub-problems. Methods are discussed for solving these linear least squares problems. The method of normal equations and a method called Hachtel's method have been applied to a 99-bus test system from the main grid. Hachtel's method is found to require 60-70% less computing time than the normal equations. It also greatly improves numerical stability, but it requires somewhat more storage. (ERA citation 11:013207)