Otimização do cabeamento interconector de aerogeradores em parques eólicos Onshore

Abstract

Throughout the history of civilizations, the use of wind energy has always had its place. It is estimated that, around 2800 BC, the Egyptians used this resource and, currently, the wind force can be used to generate electrical energy at different levels of demand, helping small isolated communities to use the benefits of this energy. On the other hand, wind farms with a large number of wind turbines contribute significantly to the Brazilian energy matrix, since the winds are more intense in the months of June to December, a period in which the rainfall intensity is lower. In this way, wind energy becomes an important supplementary source of energy as it is considered clean energy, avoiding the use of plants that generate environmental impact. In this context, the present research aims to develop the optimization of the interconnector cabling of wind turbines in onshore wind farms, since a considerable share of the cost of these farms is due to cabling. Thus, it is proposed to optimize the interconnection network of the wind turbines with the aid of the Minimum Generation Tree Problem, solved through three metaheuristics, these being the Firefly Algorithm, the Particle Swarm Optimization Algorithm and the Imperialist Competitive Algorithm. Cerro Chato and Chuí onshore wind farms were chosen for the application of optimization methods, both in the state of Rio Grande do Sul, each of which has a different type of interconnection between wind turbines, where the first has its overhead collection network, while the second presents this in an underground form. For both cases, the positioning of the wind turbines, as well as the substations, was respected, with only the interconnection cabling path between the turbines being analyzed. In this study, possible obstacles between generators were not considered. Therefore, it was found that all methods were efficient in terms of optimization, when compared to the current layout and the cabling value. Finally, we highlight the performance of the Firefly Algorithm, which presented in its response the shortest cabling length in both cases, showing that metaheuristics are effective for solving this problem.