Determining the architecture of a neural network in the problem of estimating the state of the battery charge

The problem of estimating the state of charge of the battery, based on neural networks, is considered. Two types of recurrent nonlinear autoregressive neural networks were investigated in the problem of estimating the state of charge of a battery during its use. The main criterion for the quality of forecasting was the mean square error. According to the results of the study, the optimal structure of the neural network was chosen.