loading page

AI based Approach for State of Charge Estimation of Batteries for EV Applications
  • Deepak Kumar,
  • M Rizwan,
  • Amrish K. Panwar
Deepak Kumar
Delhi Technological University Department of Electrical Engineering

Corresponding Author:[email protected]

Author Profile
M Rizwan
Delhi Technological University Department of Electrical Engineering
Author Profile
Amrish K. Panwar
Delhi Technological University Department of Applied Physics
Author Profile


In recent years Electric Vehicles (EVs) have been recognized as the most excellent substitute for petroleum and diesel-powered vehicles in the automotive industry, practically contributing to almost zero carbon emissions. Li-ion batteries (LIBs) are mainly utilized to power EVs, but certain drawbacks include temperature dependency, sluggish charging, battery ageing, etc. Estimating the state of charge is important to improve the performance and robust utilization of LIBs. The state of charge of lithium-ion batteries is directly related to their safety and efficiency, while practical calculation remains challenging for real-time applications. Here, this paper uses artificial neural network-based machine learning and deep learning approaches to estimate the battery state of charge. The battery parameters have been accurately integrated as input for the models. The proposed model’s accuracy, reliability, and robustness are evaluated using available datasets. The mean absolute error was found in the range of 0.0030 to 0.0035, and root mean square errors 0.0043 to 0.0047 were obtained at 0 and 10 °C operating temperatures. The results demonstrate that the proposed model achieves more satisfactory accuracy and robustness, which is essential for the battery management system (BMS) to make accurate decisions.