A Comprehensive SOC-OCV Hybrid Model for Turnigy Graphene 5000 mAh Lithium-Ion Batteries Using Gradient Iterative Algorithm

In Lithium-Ion Batteries (LIBs), the relationship model between State of Charge (SOC) and Open Circuit Voltage (OCV) must be adaptive, simple, robust, and accurate especially in real-time SOC estimation. Besides, the model must be able to characterize LIB in large-scale temperatures to ensure its validity at any ambient conditions. Thus, this paper presents a hybrid model developed to express the relation between SOC and OCV. The improved model has proven its efficiency when applied to Turnigy Graphene 5000 mAh LIB. The model combines logarithmic, linear, and exponential functions with six parameters. The gradient-based iterative search method which is considered one of the best algorithms is used to optimize and adapt the model parameters. The RMSE of the developed hybrid model (13 mV) is always less than the RMSE of the old polynomial model except at -10 ºC. Finally, the developed hybrid model demonstrated great efficiency at most temperatures and was superior to the old polynomial model.