Deferred Supplier Energy Amount Prediction Using Neural Network based on Switching Strategy for Resilient Smart Grid

Authors:

Abstract
With the rapid growth in population and economy, the global demand for electricity has increased significantly. To ensure that electricity is efficiently distributed to households and industries and minimize power losses, the adoption of smart grid technology has become essential. Smart grids offer the potential to reduce power losses during the distribution process and improve the overall efficiency of the electricity distribution system. Machine learning techniques have been successfully implemented to efficiently distribute power among various entities and satisfying customer demands. In the present work, several machine learning algorithms and MLP neural network have been deployed for predicting the amount of energy carried over to next period. The experimentation results highlighted the superiority of the MLP neural network algorithm, which outperformed the other state of the art machine learning algorithms yielding 0.998866 of R-squared and 44.228105 of RMSE. Moreover, we aim to introduce a distribution planning approach that fosters collaboration among electric distribution centres to effectively optimize the logistics of energy flows while minimizing energy deferment.