SOC estimation of lithium battery based on AUKF algorithm of third-order RC model

Due to the complexity of the internal electrochemical principle of lithium-ion batteries, the second-order RC equivalent circuit model cannot describe the battery characteristics well, which brings certain errors to the SOC estimation. In this paper, the third-order RC model is used to improve the accuracy of the battery model. Parameter identification was performed using the Hybrid Pulse Capability Test (HPPC) combined with the least squares method. Then, aiming at the problem that the Unscented Kalman Filter (UKF) algorithm ignores unknown noise in estimating the battery SOC, an Adaptive Unscented Kalman Filter (3RC-AUKF) algorithm is proposed combined with the third-order RC model. Finally, a third-order battery model is built and the algorithm is simulated. Through the comparison between 2RC-UKF and 3RC-AUKF, it is proved that the 3RCAUKF algorithm can converge quickly, with an average absolute error of 0.69% and a root mean square error of 0.29%, indicating that 3RC -AUKF algorithm has better estimation accuracy and robustness.