Smart grid stability prediction using binary manta ray foraging-based machine learning

Modern information and communication technologies integration has had a tremendous impact on power grids. The creation of smart grids, which are notable for their higher efficiency and reduced running costs, is being facilitated by this advancement. Maintaining the integrity of these power networks is the top priority in this paradigm shift due to their critical role in meeting the expanding energy demands of smart cities, residences, industrial sites, and more. To get around this problem, several Machine Learning and Deep Learning models may be employed to predict stability in energy networks. The crucial role that IoT technology plays in providing electricity grid networks with intelligence is highlighted by this research. To do this, the study employs a Weighted Extreme Learning Machine (WELM) with the innovative Binary Manta Ray Foraging for weight selection. Notably, the research compares the efficacy of several prediction models using key metrics including accuracy, precision, recall, and the F1 score. Additionally, the dataset undergoes extensive preprocessing employing data augmentation and feature scaling techniques, yielding excellent results. Particularly, the extended dataset exhibits a tremendous boost in performance, with an astounding accuracy rate of 99%. This investigation therefore demonstrates unequivocally that the proposed WELM model beats rival predictive models in the domain of forecasting energy grid stability, holding the possibility of increased grid resilience and dependability.