An optimal hierarchical algorithm for factored nonlinear weighted least squares state estimation

This paper proposes a novel hierarchical algorithm for nonlinear weighted least squares state estimation. The proposed method is based on the factored approach recently considered for hierarchical state estimation, but overcomes an approximation that leads to significant estimation biases in some situations. The newly proposed method is built on a new decomposition that exactly replicates the optimality conditions of the underlying weighted least squares optimisation problem. In addition to the improved accuracy, the new algorithm is also computationally faster, as it removes the need to solve additional nonlinear weighted least squares sub-problems within each iteration of the algorithm. Results of computational experiments justify these claims for two multi-area power systems.