State-augmented adaptive sliding-mode observer for estimation of state of charge and measurement fault in lithium-ion batteries

Abstract

Estimating the state of charge (SoC) in lithium-ion batteries (LiB) encounters challenges due to model uncertainties and sensor measurement errors. To solve this issue, this study introduces an estimator based on an innovative adaptive augmented sliding mode approach. This approach incorporates measurement faults as additional state variables to minimize the impacts of uncertainties effectively. Furthermore, based on the sliding mode framework, the design of this estimator addresses resistance to model uncertainties. However, sliding estimators commonly face the chattering issue. To counteract this, the paper suggests employing adaptive dynamics to determine the estimator's gain. This adaptive approach allows the gain calculation to minimize estimation errors across all time steps, effectively reducing chattering and enhancing estimation accuracy. The performance of the proposed method is validated through simulations using two practical data sets. Results demonstrate superior accuracy compared to conventional sliding methods, with improvements in SoC and terminal voltage estimation.