Parametrização e aplicação de métodos heurísticos no ajuste da curva de distribuição de Weibull em energia eólica

• Autor(a) principal: Santos, Lindemberg Ferreira dos
• Data de Publicação: 2018
• Tipo de documento: Dissertação
• Idioma: por
• Título da fonte: Repositório Institucional da Universidade Federal do Ceará (UFC)
• Texto Completo: http://www.repositorio.ufc.br/handle/riufc/36082
• topic: Engenharia mecânica; Programação heurística; Energia eólica; Heuristic methods; Distribution of Weibull; Wind energy

A few years ago, the wind energy stands out as one of the most widely used renewable energy sources in the world. The wind regime characterization according to a probability distribution is very important in the wind resource analysis. Commonly the two parameters’ Weibull distribution is used, however, a major challenge in the application of this distribution is to estimate the parameters k and c, so that a negligible error setting is obtained. In this study, It was tried to estimate these parameters for Petrolina-PE and São Martinho da Serra-RS regions by the heuristic optimization methods Harmonic Search, Cuckoo Search, Migratory Birds and Imperialist Competition, and by the deterministic methods that are usually used Least Squares Method, Moment Method, Empirical Method, Energy Pattern Factor Method, Modified Maximum Likelihood Method, Maximum Likelihood Method, Equivalent Energy Method, and Chi-Squared Method. The methods were implemented using R language. The objective function was determined and the parameterization process was performed for each heuristic method applied, which defined the ideal value of the main parameters. The fit’s quality between the generated curve and the data histogram was determined by the statistical tests RMSE (Root Mean Square Error), MAE (Mean Absolute Error), R2 (Determination Coefficient) and by the WPD (Wind Production Deviation) test . The statistical tests showed that the proposed four heuristic methods are effective and can be used to estimate the parameters k and c at the two analyzed regions. Among them, it was highlighted the Harmonic Search method, which required the smallest time to perform the estimation.