Erroneous data detection in the harmonic state estimation of a radial distribution system with distributed generation

Abstract

The present article extends a previous work where a state estimation technique was applied to a power distribution system to determine the amplitudes and phase angles of the harmonic currents injected at its buses. The nodal harmonic voltage amplitudes and phase angles were presented as the state variables and the corresponding current parameters were then obtained indirectly through calculation from the resulting estimates. The model of a 50-Hz 13.2-kV 4-bus radial system was used, where the grid and a generator were connected at opposite ends of the system in a typical distributed generation configuration. In this article, said implementation is expanded with algorithms for the detection and identification of erroneous data (bad data) such as those which might be obtained from faulty or disconnected measuring instruments. Although traditional algorithms are implemented, this work is a contribution toward the validation of harmonic state estimation techniques applied to electric distribution systems.