Combined State and Parameter Estimation of Lithium-Ion Battery With Active Current Injection

Estimating the State-of-Charge (SoC) and State-of-Health (SoH), together with the parameters used in representing the dynamics of a lithium-ion battery, is essential to ensure optimal and reliable operation. However, this simultaneous estimation can take a significant amount of time to converge, and the estimation accuracy is limited by measurement noise and model inaccuracy. Note that for overactuated systems (e.g., hybrid energy storage systems and hybrid electric vehicles), the overactuation feature can be exploited to optimize the battery current profile for the estimation purpose. This article shows the potential to improve estimation accuracy when the desired current is actively injected and the estimation algorithm is properly structured. Specifically, by incorporating a high-pass filter, battery parameters can be independently characterized by injecting high-frequency and medium-frequency currents, and battery SoCSoH can then be estimated sequentially from the estimated parameters. A Cramer-Rao bound analysis shows that the accuracy of the proposed sequential estimation is much better than the case where all parameters and states are simultaneously estimated. The analysis is verified by simulation and experimental results. We point out that for battery-only applications (e.g., electric vehicles); the proposed method has limitations, as generally, the battery current profile cannot be changed.