Deep Learning for Energy Demand Forecasting in Electric Vehicle Charging Stations

Electric vehicle (EV) growth in the Indian market has resulted in an enormous rise in energy demand. Because EVs are becoming more popular, effective charging infrastructure and energy management technologies are essential. Intelligent energy management systems are necessary to reduce the risks of power outages and disruptions resulting from the increasing number of EVs on the road. This paper proposes a novel Deep Learning (DL) model called Fractional Gradient-Based Recurrent Neural Network (FG-RNN) for predicting energy demand at electric vehicle charging stations (EVCS) to solve these challenges. This study examines and preprocesses data from an adaptable charging network (ACN). The proposed FG-RNN model was tested against two DL models: Deep Neural Network (DNN) and Long Short-Term Memory (LSTM). The FG-RNN model is one of the better solutions because of its high R-squared (R2) value of 97.682%. Furthermore, the model has a remarkably low Mean Absolute Percentage Error (MAPE) of 2.51% and a Mean Squared Error (MSE) of 0.072. Based on these results, the FG-RNN model appears to be a promising option for intelligent energy management systems for predicting energy consumption for EVCS.