Lithium battery SOH estimation through FFNN

Abstract

The battery health plays a key and decisive role in the use of lithium batteries. In this paper, the battery data detected offline is used to predict SOH through the big data platform algorithm model, in order to acknowledge the current health of the battery. The construction of the learning model is carried out through the historical data of 1000 batteries data set, so that the MAE(mean absolute error) is lower than 0.095%.