Globally convergent state estimation based on givens rotations

This paper proposes the combination of trust region methods and sequential-orthogonal techniques in order to devise globally convergent state estimators. The general nonlinear least squares problem in the context of power system state estimation is formulated so as to include inequality constraints which model the trust region. It is shown that the required changes on the state estimation equations solved in each iteration are equivalent to considering the contribution of properly defined a priori state information to the estimation process. Since a priori information are easily taken into account by the three-multiplier version of Givens rotations, the latter are employed to solve the linearized problem. This imparts numerical robustness to the iterative process in addition to the algorithmic robustness of the trust region approach, thereby improving the state estimator capability to converge even in the presence of severe modeling errors.