Método para otimização múltipla e combinada de recursos em sistemas de distribuição

Abstract

This work presents an approach to the joint optimization of resources applied to three-phase and unbalanced electrical distribution systems, focusing on the reconfiguration of feeders; allocation and sizing of fixed capacitor banks; and allocation, sizing and adjustment of voltage regulators. The model uses the modeling and simulation of the systems through the use of OpenDSS and subsequent data processing with the use of MATLAB®. The work innovates by performing the optimization of the three problems in a single method, using for this two distinct optimization techniques, more appropriate to each of the two subproblems considered. In addition, it considers objectives more appropriate to the reality of energy distributors than is commonly seen in the literature, adopting for this a cost modeling compatible with the reality of the sector, allowing the obtainment of constructive, technical and financially viable solutions for the application by energy distributors in their systems. The first of the phases, based on a modified Branch and Bound algorithm, is responsible for solving the reconfiguration problem, obtaining multiple possible network configurations to be used in the second optimization phase, based on a modified Genetic Algorithm and responsible for the allocation of capacitors and voltage regulators. In this sense, different network alternatives provided by the first optimization process are analyzed by the second process without convergence for situations with a single or reduced possible network configurations. Both problems have slightly different objective functions, and in both situations it is sought to maximize the profitability of the feeder through the commercialization of electricity and with the realization of investments for the installation of equipment aimed at improving the operating characteristics of the network. In order to reduce subjectivity, common to conducting studies of this type, energy costs and standardized real equipment for use in the area of operation of utilities, which includes a detailed modeling of the composition of costs as a function of different powers and numbers of phases. Finally, techniques are also proposed by this work that aim to optimize the use of temporal and computational resources for the realization of studies, allowing the method to simulate different systems in order to extract results of interest even with reduced hardware or time.