LSTM Based Load Prediction for Distribution Power Grid with Home EV Charging

Electric Vehicles (EVs) are getting above expectation popularity among consumers, mainly due to their environmental benefits and technological advancement. So, the fleet of EVs is growing fast and urging the electrical power infrastructural support to cope up in time. Consequently, the electrical energy authorities have to arrange for the large-scale energy supply that the gas stations have provided till now. Importantly, EV charging at home will be more dominant than centralized charging stations usage with time. This work provides a novel application of LSTM based load prediction for electric vehicle charging in a distribution power system. The utility can evade costly maintenance based on the predicted load by replacing and upgrading distribution system components. Our simulation results demonstrate the transformer upgrade and replacement decisions and the maintenance cost minimization of our method compared to two baseline policies.