Battery internal resistance and state of health estimation based on modified Van Genuchten-Gupta model compared to a feedforward neural network

Recently, estimating the state of health of batteries has been widely investigated in several research works. These researches are often based on data acquired at constant temperature, which does not reflect battery operating conditions. In this study, battery internal resistance estimation based on parametric equations and feedforward network considering the cumulative effect of temperature is proposed. Subsequently, state of health is estimated using a modified Van Genuchten–Gupta model. This model parameters are related to ambient temperature by a polynomial function of third order. Finally, the results are compared with those of a feedforward network considering the same inputs.