Authors: Dr Yaganti Krishnapriya, Talari Manohar
Abstract: The increasing demand for energy worldwide and the introduction of renewable energy sources make it necessary to move from conventional electrical distribution systems to smart electrical distribution systems. This study proposes an Internet of Things (IoT)-based framework for continuous real-time monitoring of electrical distribution systems. It employs IoT sensors, edge computing nodes, and a cloud-based analytical system for the continuous monitoring of electrical distribution systems. Moreover, the system uses a novel adaptive load scheduling (ALS) algorithm that is based on a hybrid LSTM-XGBoost model to accurately forecast future consumer loads. With the ALS algorithm, the system can predict consumer loads with an accuracy of 96.2% (RMSE = 0.034). Finally, the Model Predictive Control (MPC)-based load control system lowers peak demand and energy expenses by 27.4% and 19.8%, respectively, in a testbed with 200 residential customers.
DOI: http://doi.org/10.5281/zenodo.20666042
