A multi-scale fractional-order dual unscented Kalman filter based parameter and state of charge joint estimation method of lithium-ion battery

Accurate estimation of lithium-ion batteries' state of charge (SOC) is the key to the battery management system (BMS). A multi-scale fractional-order dual unscented Kalman filter is proposed to promote the accuracy of the battery SOC estimation. First, a fractional-order model (FOM) based on the fractional calculus theory is proposed to represent the characteristics of lithium-ion batteries. Its parameters are identified by the adaptive genetic algorithm (AGA). The Root Mean Square Error (RMSE) of the model is less than 5 mV under test conditions. Then, a multi-scale fractional-order dual unscented Kalman filter (FODUKF) is developed and employed to achieve the parameter and SOC joint estimation regarding the slow variation of battery parameter and fast variation of battery SOC. Finally, the experimental data acquired from the BTS-2000 based battery test platform have verified the effectiveness of the method. The accuracy and robustness of the proposed methods are shown by comparing the results computed by different unscented Kalman filter (UKF) approaches. The RMSE and average estimation errors of battery SOC are controlled within the range of 1%.

Automated summary

This study proposes a multi-scale fractional-order dual unscented Kalman filter to improve the accuracy of SOC estimation for lithium-ion batteries. The characteristics of lithium-ion batteries are represented using a fractional-order model based on the fractional calculus theory, and the parameters are identified using an adaptive genetic algorithm. The fractional-order dual unscented Kalman filter is employed to jointly estimate the battery parameters and SOC. Experimental results verify the effectiveness and accuracy of the proposed method.