Despacho econômico de energia na geração utilizando algoritmos evolutivos

Abstract

The present work aims to implement the evolutionary algorithms Particle Swarm Optimization (PSO), Grey Wolf Optimizer (GWO), and Genetic Algorithm (GA) in a specific problem of economic dispatch of electrical energy for thermal power plants. Initially, it presents an overview of the development of the national electricity sector, highlighting its limitations and the influence of thermal power plants on the security of the Brazilian electrical system, as well as the prospects for new contracts for this energy source. The work establishes the concepts of optimization and economic dispatch of electrical energy, detailing the studied algorithms and their implementation parameters. Using the Wilcoxon hypothesis test, comparisons were made between the medians of the costs obtained with the implementation of the algorithms. The results showed that the PSO algorithm provided the arrangement with the lowest total cost for the problem of thermal generation sources, while the GA stood out for requiring less time for its implementation. The application of the Wilcoxon test made it possible to identify that all three algorithms present statistically different cost values. It is concluded that all the evolutionary algorithms were effective in providing viable solutions to the problem, offering suitable arrangements to meet the demand for electrical energy.