A Parametric Study to Predict Wind Energy Potential from Neural Network

The assessment of wind energy potential is a topic that has not yet been thoroughly explored. The complex nature of wind potential makes it challenging to predict and evaluate. To address this issue, researchers have proposed various methods, including the use of artificial intelligence and algorithms. In this study, experimental data on wind energy conversion system performance was utilized to develop a new model based on neural networks. Specifically, a CNN-type neural network was employed due to its high capacity. Three different artificial neural network architectures were designed, trained, and evaluated for wind speed prediction. The proposed neural network model was validated by comparing it with experimental data, allowing for the prediction and analysis of wind energy conversion system performance and trends. The results indicate that the proposed model demonstrates a low error percentage (ranging from 1 to 0.97) in predicting the increasing efficiency of the wind energy conversion system, suggesting its strong alignment with the real model and network efficiency.