Answers are grounded in this paper and cite their sources.
This study evaluates wind speed forecasting methods using artificial neural networks, highlighting performance metrics like error rates.
Wind energy is critical to meeting the power requirement of the global population. Accurate wind speed forecasting is vital in energy trading, power system operations, and enhanced market balance. This study examines three Artificial Neural Network (ANN) algorithms using meteorological parameters as inputs. The three methods of training that have been discussed in the current paper are the Scaled Conjugate Gradient Backpropagation technique, the Bayesian regularization algorithm, and the Levenberg-Marquardt (LM) training algorithm. In this work, 42 datasets of the total 60 datasets procured from the National Renewable Energy Laboratory (NREL) over five years were used for training, 9 for testing, and the remaining 9 for validation. Wind speed will be the study’s dependent variable while surrounding temperature, barometrical pressure, wind orientation, relative humidity, and rainfall amount are the independent variables. In the present study, the performance of the ANN algorithms is assessed using measures such as Mean Squared Error (MSE) and correlation coefficient (R). The results indicate that all three ANN algorithms exhibit excellent performance, enabling precise wind speed predictions. This reliability supports their potential application in optimizing wind energy operations. Among these, the LM algorithm provided the most precise predictions, exhibiting the lowest error rates. Therefore, it is concluded that the LM algorithm was the most accurate in predicting wind speed for the NWTC, providing valuable insights for renewable energy planning and management.
No takes yet. Share an insight, caveat, or question.
Altork et al. (2025) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: