Fractional Order Model-based Estimation for State of Charge in Lithium-ion Battery

How to accurately estimate the state of charge is one of the key technologies for safe application in lithium-ion batteries. However, the current approaches are not fully fit for lithium-ion battery systems, and there needs improvement in accuracy, stability and practicality. In order to describe the dynamic characteristics of lithium-ion battery systems and improve the accuracy and stability, an adaptive extended Kalman particle filter is proposed based on fractional order battery model to estimate the state of charge of lithium-ion. In the process of parameter identification of fractional order battery model with immune genetic algorithm, the "memory vault" is used to reduce the calculation amount of the algorithm, and the "affinity degree" is introduced to solve the problem of local convergence of the algorithm. The proposed control algorithm was downloaded into the battery management system controller by means of model-based development and compared and verified by ECE and UDDS condition tests. The terminal voltage error of the second-order fractional battery model is not more than 13.96 mV, and the average error is 2.4-4.2 mV, which indicates that the fractional battery model is more sensitive to the change of current, and can better represent the performance of the battery voltage change, and can effectively ensure the calculation accuracy of the battery SOC. Compared with the EKF, the accuracy of SOC estimation is improved by more than 50%, and the convergence time is greatly reduced, which indicates that the introduction of adaptive Kalman filter in particle filter for correction can filter out noise and enhance the accuracy and robustness. © 2024 Chinese Mechanical Engineering Society. All rights reserved.