Renewable energy sources have gained great popularity due to the increasing importance given to a sustainable environment and economic development. Because of the environmental friendliness of renewable energy compared to fossil fuels, the tendency and investments in this field have increased. Wind energy comes into prominence among renewable energy sources because of its potential power that is used in various areas currently. Wind energy having stochastic nature is more sensitive to the extreme values of wind speed. Therefore, in order to create wind energy effectively, an accurate wind speed forecast is needed. In this study, nonlinear dynamical system approaches have been implemented by using reconstructing of phase space based on specifying minimum embedded dimension and delay time. In order to find out performance, different error metrics (MSE, RMSE, MAE, and MAPE) have been implemented. According to results, RMSE has been found 0.47 and 0.85 in hourly and daily dataset, respectively. Also, the correlation coefficient between the measurement and the obtained data set was as high as 0.92 in the hourly wind variable. In addition, a lesser correlation coefficient of 0.62 was found in the daily wind speed.