MODELO DE OTIMIZAÇÃO GA-FUZZY PARA MALHAS DE ATERRAMENTO DE SUBESTAÇÕES

Abstract

Abstract: This article used computational intelligence with the use of Matlab software to develop a ground loop optimization model. The model was developed using Fuzzy logic using parameters optimized by means of genetic algorithms. The optimized model seeks to find the smallest area and number of cables for the development of the mesh project. This model was applied to an elevating substation of a USF (Solar Photovoltaic Plant) and with the inclusion of four more values of soil resistivities. The optimization was obtained through the elaboration of an objective function and non-linear restrictions, with resolution through the application of genetic algorithms. For the USF substation grid design, with resistivity of 325 Ohms, an area of 1640 m² was found and the total cable length was 570 meters. After analyzing and certifying the results, the same procedure was adopted for the other resistivities. The Fuzzy model included five resistivity ranges where it was possible to correlate the input parameters with the results found by the AG with divergences below 3%.