ABSTRACT
This chapter includes generalized models that forecast the energy consumption of electrical vehicles (EVs) using real-world datasets and machine learning (ML). The objective of the chapter is to predict the number of EV registrations and the amount of power consumption in the next three years. Combining real-world data and a variety of ML approaches to identify potential strategies of problem solving is the focus of this chapter. First, it forecasts the number of EV sales in the upcoming years using the ML model, which is developed considering the number of EV sales in past years. Second, forecasting of future energy consumption is done by taking historical energy consumption into account. Two approaches are blended with various ML techniques such as LSTM, SARIMAX, and NeuralProphet to make the model more flexible and fit real-world problems. Additionally, the chapter evaluates the performance, which considers the different assumptions about how much energy is demanded by EVs.
