A Novel Approach for Enhancing <scp>OCV</scp> Estimation Accuracy in <scp>EV</scp> Lithium‐Ion Battery Cell Employing Adaptive Segmentation With Polynomial Fitting

ABSTRACT In the domain of electric vehicles (EVs) and portable electronics, lithium‐ion batteries constitute a vital power source, necessitating precise state estimation within Battery Management Systems (BMS) to determine the State of Charge (SOC), State of Health (SOH), and terminal voltage. Electrical equivalent circuit models (ECMs) calibrated using temperature‐dependent Open Circuit Voltage (OCV) profiles and dynamic experimental data are extensively utilized to characterize the nonlinear electrochemical dynamics of the cell. Consequently, a simple and accurate method of OCV estimation is required for real‐time implementation in the BMS. Hence, this study introduces an adaptive segmentation approach employing a piecewise linearization method with first‐order polynomial fitting for static OCV estimation, thereby facilitating improved segmentation points of the OCV–SOC curve and enhancing estimation accuracy. The OCV estimation accuracy across various operating temperatures is evaluated by comparing it with existing methodologies and examining its impact on dynamic terminal voltage. The Root Mean Square Error (RMSE) value of the proposed method was compared with other methods, including data‐driven methods, model‐based, high‐order polynomials, and Gaussian regression, irrespective of operating temperature. Experimental validation of the proposed methodology was performed using MATLAB Simulation. The performance of the approach was assessed by RMSE and R 2 estimation metrics, demonstrating an improvement in the OCV‐SOC error across all temperature ranges compared with other well‐established methods. Furthermore, the RMSE value of the terminal voltage and behaviors of model parameters using this method are comparable to those of the existing model‐based approach.