Predictive Analytics in Smart Grids: Leveraging Machine Learning for Renewable Energy Sources
Pages : 677-683, DOI: https://doi.org/10.14741/ijcet/v.11.6.12
Download PDF
Abstract
Stability testing and forecasting under various situations is very relevant since it is one of the most essential factors for evaluating the utility of smart grid architecture. A future smart grid design that can anticipate stability and avoid unwelcome instabilities is necessary due to the proliferation of both residential and commercial buildings as well as the incorporation of renewable energy sources into smart networks. To tackle the problems that come with integrating renewable energy sources, this research looks at how smart grid systems may be made more stable with an employ of predictive analytics and ML models. A simulated smart grid stability dataset containing 60,000 entries and 14 features from Kaggle. Three models were employed: ANN, CNN, and CART. The ANN model achieved superior results, with an accuracy of 98.7%, precision of 98.03%, recall of 98.02%, and F1-score of 98.02%. Comparison of ANN, CNN, and CART models demonstrated the ANN’s efficacy in accurately forecasting grid stability. The results highlight the promise of DL models and other forms of ML in predictive analytics for making renewable energy smart grids more reliable.
Keywords: Smart grid, renewable energy resource, Solar Energy, Predictive Analytics machine learning.