Machine Learning Approaches for Predicting Electric Vehicle Charging Demand: A Comparative Review

The increasing demand of electric vehicles requires precise charging demand estimation for energy management and proper infrastructure development. Various machine learning techniques are examined in this paper's comparative review which can help forecast EV charging demand based on accuracy, scalability, and computational cost. Random Forest models offer a lower computational cost and resource demand than deep learning models, while Long Short-Term Memory models demonstrate accuracy in managing long-range dependencies. Hybrid approaches are proved to be the most suitable for spatialtemporal prediction of EV demand because they offer the best balance between accuracy and scalability. Reinforcement Learning models possess superior real-time responsiveness; however, traditional statistical models like ARIMA are hindered by the scalability problem despite being elementary. This work adds to the identification of the most suitable machine learning models for predicting electric vehicle charging demand, supported by a thorough review of literature.