Design of A Smart Microgrid for Predictive Energy Generation

The exploration of artificial neural networks (ANNs) in the realm of renewable energy-based smart grids is the primary focus of this study. The difficulties associated with the integration of renewable energy sources into the electrical grid, primarily due to their inherent variability and intermittency, are identified as the main problem. The smart grid technologies development is recognized as a solution to enhance the grid’s efficiency and reliability. It is suggested that ANNs could significantly contribute to the smart grid system performance improvement. An examination of renewable energy-based smart grid key features, integration-related challenges, and potential benefits of incorporating ANNs are presented. Research on ANNs’ application in various smart grid systems components, such as energy demand forecasting, energy generation and storage optimization, and grid stability management, are reviewed. Data division for this study was accomplished through random subsampling, with 80% of the data forming the training set and the remaining 20% comprising the testing set. A comparison between these two groups was subsequently performed. Following the computations to determine the Mean Squared Errors (MSEs) for a percentage range between 60% and 90%, the study concludes by highlighting areas requiring additional research to fully exploit the potential of ANNs in renewable energy-based smart grids. In this study, the main objective is to understand the role of ANNs in optimizing the performance of renewable energy-based smart grids and identify areas for future research.