AC network state estimation using linear measurement functions

Abstract

The real/reactive power and current magnitude measurements can be accounted for in an AC network state estimator using linear measurement functions. The nonlinearity in the conventional AC state estimator equations is transferred from the measurement functions into a system of nonlinear equality constraints which is independent of the measurement set. The new format of equations entails two advantages. First, it can be easily integrated in optimisation routines which employ first- and second-order derivative functions. Second, the linear measurement functions can benefit from scaling techniques which are well documented in the linear programming literature. This research details the implementation of a least absolute value state estimator employing the new format of equations. The optimisation method is based on a primal-dual interior-point method that can accurately account for zero injection measurements and power directions. Numerical testing is used to validate the approach for networks with measurement sets that are (i) conventional and (ii) have a high proportion of current magnitude measurements and power signs.