Coulomb counting state-of-charge algorithm for electric vehicles with a physics-based temperature dependent battery model

This paper proposes a new Coulomb counting state-of-charge (SOC) algorithm for a lithium iron phosphate battery that achieves accurate results (3.8% standard deviation) for several drive cycles and temperatures between -10°C and 20°C. The basis for this method is a physics-based battery model that compensates for the influences of drive dynamics and temperature on the SOC estimates, instead of the data-driven, empirical adjustments that have been reported in the literature to date. The method also provides compensation for battery aging effects by recursively updating the battery parameters using a Kalman filter algorithm. This method has been evaluated using both lab equipment and a converted Ford F150 electric truck. The results confirm the robustness of the proposed algorithm for suppressing the influences of drive dynamics and temperature on SOC estimates.