Application of the Cuckoo Search in the Adjustment of Weibull Curves for Wind Energy Using Wind Data of Petrolina City

In this work a Cuckoo Search (CS) applicability study was carried out using the methodology described by Yang and Deb (2009) in the optimization of Weibull curves, being implemented from an algorithm written in programming language R. The real velocity data were obtained from the National System for the Organization of Environmental Data (SONDA) platform for the city of Petrolina - PE, because it is located in a state with great potential for wind energy generation. The Cuckoo Search heuristic method was compared with six deterministic methods widely used in the literature from the results of the statistical tests Root Mean Square Error (RMSE), R2, Mean Absolute Error (MAE) and Power Density Error (EDP). With the objective function used, Cuckoo Search was satisfactory in the wind representation of the studied region, presenting RMSE of 0.006052, R2 of 0.99348 and MAE of 0.003979, more desirable in relation to the other methods. Also, an EDP of approximately 1.12% above the actual power density was obtained, which makes it difficult to use for problems in which energy oversizing is not allowed.