State-of-charge Estimation for Lithium Primary Batteries: Methods and Verification

Liqiang Zhang,Hezhen Liu,Xiangyu Wang,Ming Li

DOI: https://doi.org/10.1016/j.est.2024.111189

IF: 9.4

2024-01-01

Journal of Energy Storage

Abstract:An accurate estimation of the state of charge (SOC) for lithium primary batteries is rarely achieved, leading to resource waste and environmental pollution. In this study, we propose two methods, the stress accumulation method and the electrochemical impedance spectroscopy (EIS) graphical method, to address the shortcomings in SOC prediction for lithium primary batteries. The stress accumulation method utilizes discharge data in the time domain, and obtains the discharge current and capacity, which are subsequently used to determine the battery's actual discharge capacity under different current stresses. The EIS graphical method enables SOC estimation without historical discharge data based on EIS measurements at different SOC levels, and the resulting data are graphically transformed into seven parameters. A neural network model is trained to establish a relationship between the SOC and EIS. Verification experiments demonstrate that the stress accumulation method achieves an absolute SOC estimation error of +/- 0.05, while the EIS graphical method achieves an absolute SOC estimation error of +/- 0.12. Furthermore, applying both methods to the same battery results in an absolute SOC estimation error of approximately 0.036. Finally, we develop a " Lithium Primary Battery SOC Estimation " MATLAB App.