Redefining ground truth for estimating the charge level of lithium-ion batteries using relaxation-based voltage measurements

Accurate state-of-charge (SOC) estimation is essential for the safe and efficient operation of lithium-ion batteries in electric vehicles. However, the evaluation of SOC estimation methods remains inconsistent due to the absence of a standardized ground truth. Despite known limitations such as integration drift and current sensor bias, Coulomb Counting (CC) remains the most widely used reference in the literature. This study introduces a unified experimental framework for defining a physically consistent ground truth and validates the use of relaxation-based open-circuit voltage (OCV) curves as a superior alternative. Stepwise C/20 discharge tests were conducted on NMC prismatic cells across four temperatures (–10 °C to 40 °C), incorporating 7-hour relaxation intervals terminated when |dV/dt| < 0.1 mV/min. The stabilized voltages at each SOC level were used to construct high-fidelity open circuit voltage to state of charge)OCV–SOC( maps, which were then applied to a mixed drive cycle to compare SOC estimates derived from CC against the relaxation-based reference. Results show a cumulative CC error exceeding 3.5%, underscoring its bias under dynamic conditions. The framework is chemistry-agnostic and can be extended to other chemistries (e.g., LFP) and aging states. These findings support the adoption of relaxation-based SOC as a standard reference for model validation, contributing to more reliable battery management and improved energy efficiency in electric vehicle systems. • Introduces a unified framework for defining SOC ground truth in Li-ion batteries. • Uses long-duration voltage relaxation (7 h) to construct high-fidelity OCV–SOC curves. • Demonstrates that Coulomb Counting accumulates integration drift and sensor bias. • Establishes a physically consistent reference applicable across temperatures (–10 °C – 40 °C). • Enables more reliable validation and benchmarking of SOC estimation algorithms.