A BP neural network-Ant Lion Optimizer and UKF method for SOC estimation of lithium-ion batteries

Accurate state of charge (SOC) estimation is great significance to promote the development new energy vehicles. And a battery model’s accuracy for SOC accuracy. To this end, method based on back propagation neural network-Ant Lion Optimizer (BPNN-ALO) and unscented Kalman filter (UKF) proposed. First, 2-RC model established BPNN used fit OCV-SOC corresponding relationship. Second, ALO are combined complete parameter identification, UKF estimation. Finally, verify BPNN-ALO-UKF under two working conditions, compare it with other methods. The test results show that proposed has higher accuracy, minimum values Root Mean Square Error (RMSE) Absolute (MAE) only 0.51% 0.40%, respectively. different also better generalization robustness.