Alocação otimizada de sensores indicadores de falta em redes de distribuição de energia elétrica

Carregando...
Imagem de Miniatura

Data

2016-12-16

Autores

Cruz, Héctor Manuel Orellana [UNESP]

Título da Revista

ISSN da Revista

Título de Volume

Editor

Universidade Estadual Paulista (Unesp)

Resumo

Este trabalho propõe abordar o problema de alocação de dispositivos indicadores de faltas (IFs) em redes de distribuição de energia elétrica por meio da técnica do Algoritmo Genético Adaptativo (AGA). O algoritmo busca obter uma configuração eficiente de instalação de IFs no sistema e reduzir o custo anual da energia não suprida (CENS) e por outro lado, o custo anual de investimento dos sensores (CINV). Estes custos conflitantes devem ser minimizados como um problema combinatório utilizando o AGA que possui taxas de recombinação e mutação dinamicamente calibradas baseadas na diversidade de cada população no processo. Os resultados mostram que o AGA é um método efetivo para encontrar soluções para o problema de alocação de IFs. Quando comparado com o AG clássico em vários testes, é observado que o AGA tem uma convergência mais rápida mostrando ser mais eficiente. Além disso, o AGA é utilizado para obter uma variedade de soluções numa Fronteira de Pareto aproximada variando os pesos do CENS e o CINV na função objetivo. Portanto, esta metodologia permite obter um conjunto de soluções em lugar de uma única solução e a partir da Fronteira de Pareto é possível escolher a solução que melhor satisfaz os interesses técnicos e econômicos da concessionária de distribuição de energia elétrica.
This work proposes the Adaptive Genetic Algorithm (AGA) to solve the problem of Fault Indicator (FI) placement in electric distribution systems. The algorithm attempts to obtain an efficient configuration for the installation of FIs in the system and reduce the annual costumer interruption cost (CIC) and on the other hand, the annual investment cost (CINV) of the fault indicators placement. These two conflicting costs must be minimized as a problem of combinational nature with the AGA which uses dynamically calibrated crossover and mutation rates based on the diversity of each population of the process. Results show that the AGA is an effective method to find solutions to the problem of fault indicator placement. When compared with the classic GA in various tests, it is observed that the AGA converges faster showing to be more efficient. Besides, in the last test the AGA is used to obtain a variety of solutions in an approximate Pareto Front by varying the weighting factors of the CIC and CINV in the objective function. Therefore, this methodology permits to obtain a set of solutions instead of one only solution and from the Pareto Front it is possible to choose the solution that best satisfies the technical and economic interests of the concessionary.

Descrição

Palavras-chave

Adaptive genetic algorithm, Fault indicators, Service quality, Electric energy distribution, Algoritmo genético adaptativo, Distribuição de energia elétrica, Indicadores de faltas, Qualidade do serviço

Como citar