Comparative Analysis of the Accuracy of Lithium-Ion Battery State of Charge Estimation Using Open Circuit Voltage-State of Charge and Coulomb Counting Methods with Simulink MATLAB

This study investigates the State of Charge (SOC) estimation of a battery using secondary data from the Samsung INR 18650-20R (2000mAh). The methods employed include the OCV-SOC, Coulomb Counting, and the 1RC equivalent battery model at temperatures of 0°C, 25°C, and 45°C. This research evaluates the accuracy of these methods while assessing the influence of temperature on SOC estimation performance, which is critical for battery management systems in various applications. The equivalent battery model was tested using a 1A current with 10% SOC intervals, while the SOC estimation was performed under a 0.1A current during discharge conditions. The results indicate that the 1RC model demonstrates the smallest error at 25°C and 45°C, establishing itself as the most consistent method for SOC estimation across these temperatures. The Coulomb Counting method exhibits superior performance, with an R² value nearing 1 across all tested temperatures, showcasing its reliability in accurately reflecting SOC. Conversely, the OCV-SOC method delivers an R² range of 0.9757–0.9864, with its best accuracy observed at 45°C but significantly lower accuracy at 25°C, especially at low SOC levels (0–10%). The Coulomb Counting method’s high accuracy is influenced by its reliance on ideal simulation data, which excludes real-world challenges such as current leakage and sensor fluctuations. Nonetheless, the combination of the 1RC model and the Coulomb Counting method proves more reliable for SOC estimation under diverse temperature conditions compared to the OCV-SOC method.