EV-Based Reconfigurable Smart Grid Management Using Support Vector Regression learning technique Machine Learning

Abstract Renewable microgrid (RMG) is a new solution to improve the safety, reliability, quality, and performance of energy in electrical systems. By implementing different renewable energy sources, like photovoltaic panels and wind turbines, RMGs can reduce greenhouse gas emissions and enhance proficiency. The paper suggests a RMG energy management method based on machine learning, which considers a reconfigurable frame according to the remote link switching and segmentation. The proposed scheme investigates advanced support vector machines to model and estimates the charging demand (ChD) of hybrid electric vehicles (HEVs). To reduce the load impact of HEV on the network, 2 various cases have been implemented; one is coordinated, and the other is smartly charged. Because of the complex construction of the problem statement, an improved optimization technique according to the gray wolf is proposed. The simulation outcomes of the IEEE microgrid (MG) trial system depict that it is of adequate and efficient quality in both cases. Based on the result of the forecast for the total charge demand of HEVs, the average absolute percentage error is 0.978 that is so small. Furthermore, the outcomes are shown that, compared to the coordination scheme, the total operating cost of the MG in smart load is reduced by 2.5%.