SOC estimation of lithium-ion battery using adaptive extended Kalman filter based on maximum likelihood criterion

A method for estimating the State Of Charge (SOC) is presented, which combines the backward difference equation model of the power battery with the adaptive extended Kalman filter based on the maximum likelihood criterion algorithm (MLC-Based AEKF). The discrete model of the Thevenin equivalent circuit for the power battery is built by the backward difference method. Compared with the bilinear transform equation model, the proposed method has the advantages of a simple structure and low computational complexity. The forgetting factor recursive least squares algorithm (FFRLS) is used to identify the parameters of the model. The MLC-Based AEKF is applied to realize an online SOC estimation under the unknown interference noise environments. The simulation experiment results verified the effectiveness of the SOC estimation method; that is, the SOC estimation average error is less than 0.15% and its maximum error is less than 0.5% in the Dynamic Stress Test (DST) condition.