The Precision SOC Estimation Method of LiB for EV Applications Using ANN

Abstract

Lithium-ion battery(LiB) is being used in various fields due to their advantages such as high energy density, high power density, and longer cycle life. In order to optimize the performance of the LiB and improve the lifetime of the Electric Vehicle (EV), monitoring the State of Charge (SOC) is very important. Therefore, it is essential to estimate of the SOC of the battery. This paper proposed the SOC estimation model for lithium-ion batteries based on the Artificial Neural Network (ANN) model. In the proposed model, SOC estimation of lithium-ion batteries was performed through five steps include variable selection, data collection, data preprocessing, neural network paradigm, and neural network learning. The actual SOC and the predicted SOC of the EV battery model were compared to prove the validity of the ANN model. As a result, the ANN showed the maximum and average errors of 18% and 2.65%, respectively, and the accuracy was 97.35%.