Year: 2021
Type: dissertation
Abstract: State estimation in power systems refers to the problem of evaluating voltage magnitudes and angles at all nodes given a set of measurements.Measured values can be active and reactive power at nodes (generation or load); active power, reactive power or current flowing through branches; and voltage magnitudes.Once all state variables are known, all other parameters that describe the power system can be determined immediately.The various types of error that affect measured values can be kept under control through measurement redundancy.The application of state estimation in transmission systems, where the assumption of balanced network and loads is customary, has already reached a mature stage of development.However, the same is not true for distribution systems, owing to a number of reasons: (i) the need for a three-phase representation, since neither loads nor network are balanced, (ii) the usually large number of nodes, much larger than the number of nodes in transmission systems, which have a considerable impact on the stability of the numerical solution, and (iii) the fact that redundancy is not normally guaranteed.This dissertation then represents an attempt to deal with the state estimation problem in large, unbalanced distribution systems.To validate the proposed methodology, it also presents applications on real distribution networks, including an underground network and comparisons to the actual methods used by utilities to include measurements on the network modelling.The results indicated that state estimation was able to properly include measurements on the network modelling.This way it could be an important tool for the evolution of power distribution sector.
Author: Gustavo de Gois Himeno
Language: Portuguese
Subfield: Electrical and Electronic Engineering
Field: Engineering
Domain: Physical Sciences
Open Access status: gold
OpenAlex ID: W4221138060