State of the Art of Machine Learning in Energy Management System

The rapid growth in electric vehicle (EV) adoption has emphasized the need for intelligent energy management systems (EMS) that can enhance efficiency and driving performance. This study introduces a Machine Learning-Driven EMS that employs regression models, neural networks, and reinforcement learning to dynamically optimize power distribution in EVs. Implemented and evaluated through MATLAB and Simulink simulations, the proposed EMS outperformed conventional rule-based systems by reducing energy consumption by up to 12% and increasing the driving range by 8–10%. The system was particularly effective in low-load and steady-speed conditions, demonstrating adaptive real-time control and power optimization. Sensitivity analysis revealed that factors such as battery capacity, motor power, and driving cycle profiles significantly affect EMS efficiency. The findings underscore the potential of machine learning to revolutionize EV power management by enabling more sustainable, efficient, and intelligent mobility solutions. Future work will involve model refinement, real-world data integration, and compatibility with smart grid systems for broader application.