PROPOSED METHODOLOGY FOR ANALYSIS OF HARMONIC IMPACTS IN ELECTRICAL GRIDS USING DECISION TREE TECHNIQUE - PART 1

Item

Tipo do ITEM
Artigo Ciêntifico
Título do Artigo
PROPOSED METHODOLOGY FOR ANALYSIS OF HARMONIC IMPACTS IN ELECTRICAL GRIDS USING DECISION TREE TECHNIQUE - PART 1
Descrição
The progressive use of power electronic devices, connected to the electrical system, has
contributed to the increase of problems caused by harmonic disturbances in electrical networks.
Thus, this work aims to develop a methodology to analyze the harmonic impacts on the electrical
grids caused by non-linear loads with the use of the Decision Tree technique. The model used
was the regression tree through the use of the Classification and Regression Tree (CART)
algorithm, which relates the harmonic current of a non-linear load and the harmonic voltage at the
common connection point of the electrical grid. The technique used was applied in a case study in
order to show the validity of its application in solving problems that seek to investigate the
dependency relationship between the two variables. Among the feeders used in the research for
the 3rd harmonic analysis, feeder DIAL2-19 contributed the impact factor for phase A with
43.587%. The feeder DIAL2-16 contributed with impact factor 36.407% for phase B. For phase C
feeder DIAL2-17 contributed 26.434% impact factor
Abstract
The progressive use of power electronic devices, connected to the electrical system, has
contributed to the increase of problems caused by harmonic disturbances in electrical networks.
Thus, this work aims to develop a methodology to analyze the harmonic impacts on the electrical
grids caused by non-linear loads with the use of the Decision Tree technique. The model used
was the regression tree through the use of the Classification and Regression Tree (CART)
algorithm, which relates the harmonic current of a non-linear load and the harmonic voltage at the
common connection point of the electrical grid. The technique used was applied in a case study in
order to show the validity of its application in solving problems that seek to investigate the
dependency relationship between the two variables. Among the feeders used in the research for
the 3rd harmonic analysis, feeder DIAL2-19 contributed the impact factor for phase A with
43.587%. The feeder DIAL2-16 contributed with impact factor 36.407% for phase B. For phase C
feeder DIAL2-17 contributed 26.434% impact factor
Língua do arquivo
inglês
Data da Publicação
Ano 2022
Palavra-chave
Methodology
Harmonic Impacts
Electrical Networks
Computational Intelligence
Decision Tree
Autores
Pedro Ferreira da Silva Filho
Jandecy Cabral Leite
Rivanildo Duarte Almeida
Local
UFPA - Manaus, 2022
Coleções
ARTIGOS