Evaluating impact of different factors on electric vehicle charging demand

Electric vehicles (EVs) are emerging as majorenergy consumers, offering numerous environmental andoperational advantages such as reduced greenhouse gasemissions and lower reliance on fossil fuels. As the adop-tion of EVs accelerates globally, accurate forecasting of EVcharging demand becomes increasingly critical for maxi-mizing the efficiency, reliability, and profitability of char-ging infrastructure. However, many existing forecastingmodels fall short by neglecting the complex and dynamicinfluence of external factors– particularly weather condi-tions and calendar variables– which can significantly affectusage patterns. This study presents a robust forecasting fra-mework that integrates historical charging data with bothtemporal and meteorological information to comprehen-sively evaluate their individual and combined impacts onEV charging behavior. Leveraging long short-term memorynetworks– effective in modeling time-series data– we eval-uate the impact of contextual features on forecasting per-formance. Results show that calendar information notablyimproves accuracy, surpassing the effect of weather data.These insights help EV station operators optimize sche-duling, reduce uncertainty in day-ahead energy planning,and support sustainability and grid stability.