State of charge estimation for lithium-ion batteries: An adaptive approach

• We develop a reduced-complexity model from single particle model of Li+ batteries.
• We analyze the local observability/identifiability of the model.
• We propose an adaptive SoC estimation algorithm using IEKF.
• The algorithm can estimate SoC with easy implementation in the presence of unknown parameters.
• The analysis and results presented can be readily extended to other models.

State of charge (SoC) estimation is of key importance in the design of battery management systems. An adaptive SoC estimator, which is named AdaptSoC, is developed in this paper. It is able to estimate the SoC in real time when the model parameters are unknown, via joint state (SoC) and parameter estimation. The AdaptSoC algorithm is designed on the basis of three procedures. First, a reduced-complexity battery model in state-space form is developed from the well-known single particle model (SPM). Then a joint local observability/identifiability analysis of the SoC and the unknown model parameters is performed. Finally, the SoC is estimated simultaneously with the parameters using the iterated extended Kalman filter (IEKF). Simulation and experimental results exhibit the effectiveness of the AdaptSoC.