State estimation of a three-phase four-wire secondary distribution network

Abstract

A state estimation algorithm for assessing the state of a power delivery network according to the measurement information performed by smart meters is proposed and its efficient use in a real low-voltage three-phase four-wire secondary distribution network demonstrated. Nonlinear state estimation is performed by a method of simple iterations, at each step of which an overdetermined system of linear equations of measurements is solved by the method of weighted least squares. State estimation is performed independently for each phase and the neutral wire; phase voltage estimates relative to the neutral wire are determined according to estimates of the phase wire voltages and the neutral wire relative to the ground. Testing of the method is carried out on the example of a real 11-node main feeder, on the poles of which MIR S-04, MIR S-05 and MIR S-07 single-phase and three-phase meters of Russian production are installed. For the state estimation, information was used on the hourly average power of loads and voltage modules for 576 measurement sections taken from the protocols of an automated commercial electricity accounting system. The high accuracy of the obtained estimates is confirmed by no more than 1.2 V residues between the measured values of the variables and their estimates. Two methods for determining energy losses by estimating voltages - with respect to earth and a neutral wire, respectively - are presented. The possibility of balancing loads by moving single-phase loads of the most loaded phase to a less loaded phase is demonstrated. The calculations performed for a real network show the proximity of variables measured by smart meters to their estimates, confirming the effectiveness of the presented algorithm for assessing the state of the secondary distribution network with explicit consideration of the neutral wire.