Structure in total least squares parameter estimation for electrical networks

Abstract

© 2018 A total least squares approach to estimating the line admittances of an electrical network for power distribution is explored. It is observed that the Newton iterates for solving the corresponding non-linear Karush-Kuhn-Tucker (KKT) optimality conditions are structured for path and tree networks. In particular, it is revealed that a condensed form of the linear system of equations to be solved at each Newton step has a sparse block arrow-head structure, which can be exploited to improve the scalability of the approach. A numerical example for a path network is presented.