Forecasting Electric Vehicle charging demand using Support Vector Machines

Road transport today is dominated by oil-delivered fuels and internal combustion engines and such a high level of dependence on one single source of primary energy carries strategic, climatic and economic risks [1]. Electric mobility offers an opportunity for diversification of the primary energy sources used in transport, but also brings new risks, technological challenges and commercial imperatives. Large penetration of Electric Vehicles (EV) will increase the electricity demand and load forecasting plays a central role in the operation and planning of electric power. This paper proposes a short-term load forecast model using Support Vector Machines, an artificial intelligence technique. A realistic scenario is studied to test the performance of the suggested model. The accuracy of the method is evaluated through a comparison with a Monte Carlo forecasting technique.