Adaptive H-infinity Filter Design for Implementing on Battery State-of-Charge Estimation Based on Fuzzy Modeling

This work suggests a new method for fuzzy modeling of lithium-ion battery and State of Charge (SOC) estimation using Adaptive HR∞R Filter. At First, a common linear dynamical model is opted for the battery. Then, for any specified interval in the charge process the free parameters of the model are identified. So, the parameters can be considered as functions of the SOC in the entire process. The parameters are modeled and determined based on simple fuzzy systems and least squares method, respectively. The proposed model is incorporated in adaptive HR∞R filter, for SOC estimation. Moreover, a new method for adjusting the tuning parameters of the filter is suggested which shows good performance in practice. The experimental tests show some advantages of the proposed estimator which include: a) good accuracy and b) fast convergence and robustness to far initial conditions from their actual values.

Keywords:

HR∞ , filter , Fuzzy modeling , State of charge , Lithium-ion batteries