Design of an Effective State of Charge Estimation Method for a Lithium-Ion Battery Pack Using Extended Kalman Filter and Artificial Neural Network

Currently, Lithium-ion batteries (LiB) are widely applied in energy storage devices smart grids and electric vehicles. The state of charge (SOC) is an indication the available battery capacity, one most important factors that should be monitored to optimize LiB’s performance improve its lifetime. However, because SOC relies on many nonlinear factors, it difficult estimate accurately. This paper presented design effective estimation method for a LiB pack Battery Management System (BMS) based Kalman Filter (KF) Artificial Neural Network (ANN). First, considering configuration specifications BMS pack, ANN was constructed estimation, then trained tested using Google TensorFlow open-source library. An model extended KF (EKF) Thevenin developed. Then, we proposed combined mode EKF-ANN integrates EKF into ANN. Both methods were evaluated through experiments conducted real pack. As result, showed maximum errors 2.6% 2.8%, but better with less than 1% error.