Linear Parameter-varying Model for Prediction of Charge / Discharge Behavior of Tri-Electrode Zinc-air Flow Battery

Recently, energy storage systems have been a hot topic of research due to their potential to enable the use and integration of renewable energy into a power grid. Zinc-air flow batteries (ZAFBs) are viewed as having significant promise for usage as renewable energy storage devices due to their high energy density and low cost. The battery management system (BMS) for ZAFBs, on the other hand, is still in its infancy, as a precise prediction of their nonlinear behavior is necessary. To address this shortcoming, a linear parameter varying (LPV) model is constructed via a multiple linear time-invariant (LTI) model using charge/discharge current and state of charge (SOC) as scheduling parameters. The model created is validated using a variety of battery data from several experimental batches. According to the discharge current and state of charge, the results reveal that the established LPV model can accurately predict the nonlinear behavior of the ZAFBs. Thus, it is determined that in terms of local accuracy, the LPV model is comparable to the linear model. With respect to the global accuracy, the analysis demonstrates that the LPV model outperforms the linear model. This result provide arguments for the LPV model's capacity to anticipate the dynamics of ZAFBs and their suitability for usage in the BMS.