Comparative Short-Term Wind Energy Forecasting Using Hybrid JAYA-ANN, GRU-ANN, and VMD-ANN Models
International Journal of Smart Grid, vol.10, no.1, pp.14-29, 2026 (Scopus)
- Publication Type: Article / Article
- Volume: 10 Issue: 1
- Publication Date: 2026
- Doi Number: 10.20508/ijsmartgrid.v10i1.696.g418
- Journal Name: International Journal of Smart Grid
- Journal Indexes: Scopus
- Page Numbers: pp.14-29
- Keywords: Artificial neural network, forecasting, hybrid learning, wind energy
- Yozgat Bozok University Affiliated: Yes
Abstract
Wind power generation is directly dependent on weather conditions, so it is very difficult to predict how much energy will be generated in a given time period. The main objective of this study is to predict wind power generation more accurately by comparing an Artificial Neural Network (ANN)-based forecasting model with JAYA-ANN, Gated Recurrent Unit (GRU)-ANN, and Variational Mode Decomposition (VMD)-ANN hybrid models. Wind direction, particulate matter (PM10), temperature, and historical power data for summer and spring seasons were collected to estimate the daily wind power generation capacity. Of the collected data, 80% was divided into a training set, 10% into a validation set, and 10% into a test set, and appropriate modelling methods were applied. The performance of the models was evaluated with Mean Absolute Percentage Error (MAPE), Root Mean Squared Error (RMSE), and Mean Absolute Error (MAE) error measures. According to the prediction results of MAPE, RMSE, and MAE values of ANN, VMD-ANN, GRU-ANN, and JAYA-ANN hybrid models in summer and spring seasons, the JAYA-ANN hybrid model is better than other prediction models and provides higher forecasting accuracy. Such a forecasting model can be an important guide for energy planners and local electricity providers for generation planning and management of alternative energy sources.