Hybrid Kalman Filter–Sliding Mode Observer for State-of-Charge Estimation in Lithium-Ion Batteries under Dynamic Load Conditions

The estimation of State-of-Charge (SoC) is crucial for the safe and efficient operation of battery management systems (BMS), but no single method is able to reliably estimate the SoC under all operating conditions. In this paper, a hybrid estimator based on the Extended Kalman Filter (EKF) and Sliding Mode Observer (SMO) is presented by employing a gated bumpless-transfer switching technique for the SoC estimation of a Lithium-ion battery. The EKF generates accurate estimates when the operating condition is normal, and the SMO is triggered when the measurement innovation is larger than an adaptive threshold, enabling the EKF to provide accurate estimates when large initial SoC error, high measurement noise, and transient sensor faults exist. In this paper, a second-order RC equivalent circuit model (ECM) of a 2.6 Ah lithium-ion cell was used to implement the proposed hybrid estimator as well as the EKF-only, SMO-only, and Coulomb counting methods. The results indicate that the hybrid EKF–SMO estimator does not significantly outperform a well-tuned EKF in an ideal Gaussian noise environment; however, it is significantly more robust against faults in the sensors and large uncertainty, which is suitable for practical BMS applications.