Forecasting the stability of A 4-node architecture smart grid using machine learning

Smart grid stability is one of the most important factors that can be used as a criterion for assessing the usability of smart grid architecture, so testing and predicting stability under various circumstances hold great importance. As a result of the increase in residential and industrial structures, and the integration of renewable energy into the smart grids, some intelligent solutions to predict stability to prevent unwanted instabilities in a future smart grid architecture is needed. In this study, we used various machine learning methods to predict smart grid stability. We approached the problem as a classification problem, we used a 4-node architecture smart grid dataset, and applied some well-known classification methods to classify the dataset into two classes which are stable and unstable. For the classification part, we used k-Nearest Neighbour (kNN), neural networks (NN), a support vector machine (SVM), and a decision tree. All four methods were tested under different hyper parameters. Finally, the ones with the best results were reported.