PREDICTION OF EV CHARGING PATTERNS USING HYBIRD MACHINE LEARNING ALGORITHMS

Abstract

Abstract—Electric cars are becoming increasingly popular in the transportation sector. Despite their growing acceptance, several factors still limit their widespread use. These factors include the relatively limited driving range of electric vehicles and the high costs related to battery production and maintenance. Understanding the energy usage of EVs becomes more important nowadays due to the rapid usage and emergence of EVs into the market. To address these challenges, researchers have turned to machine learning models to predict electric vehicle charging patterns accurately. Among these models, an ensemble learning approach has shown to outperform previous models significantly. This is evidenced by superior Symmetric Mean Absolute Percentage Error (SMAPE) scores, indicating more accurate and reliable forecasting of charging behaviours.