All state of charge (SoC) estimation algorithms based on equivalent circuit models (ECMs) estimate the open circuit voltage (OCV) and convert it to the SoC using the SoC-OCV nonlinear relation. These algorithms require the identification of ECM parameters and the nonlinear SoC-OCV relation. In literature, various techniques are proposed to simultaneously identify the ECM parameters. However, the simultaneous identification of the SoC-OCV relation remains challenging. This paper presents a novel technique to construct the SoC-OCV relation, which is eventually converted to a single parameter estimation problem. The Kalman filter is implemented to estimate the SoC and the related states in batteries using the proposed parameter estimation and the SoC-OCV construction technique. In the numerical simulations, the algorithm demonstrates that it accurately estimates the battery model parameters, and the SoC estimation error remains below 2%. We also validate the proposed algorithm with a battery experiment. The experimental results show that the error in SoC estimation remains within 2.5%.

The key indicator to assess the performance of a battery management system is the state of charge (SoC). Although various SoC estimation algorithms have been developed to increase the estimation accuracy, the effect of the current input measurement error on the SoC estimation has not been adequately considered in these algorithms. The majority of SoC estimation algorithms are based on noiseless current measurement models in the literature. More realistic battery models must include the current measurement modelled with the bias noise and the white noise. We present a novel method for mitigating noise in current input measurements to reduce the SoC estimation error. The proposed algorithm is validated by computer simulations and battery experiments. The results show that the proposed method reduces the maximum SoC estimation error from around 11.3% to 0.56% in computer simulations and it is reduced from 1.74% to 1.17% in the battery experiment.