SOC Estimation Of Lithium Battery Based On Adaptive Fractional Cubature Kalman Filter

New energy electric vehicles have become one of the development trends in the automotive industry and are supported by consumers.The battery management system is the key to ensure the normal operation of the power supply system and provides the required energy for the vehicle.In the battery management system,the state of charge(SOC)is an important state quantity that needs to be continuously detected to quantitatively evaluate the remaining battery power.Reasonable monitoring of the battery’s SOC can prevent overcharge and over-discharge damage to the battery.In addition,accurate and reliable battery SOC estimation results are an important basis for drivers to estimate the remaining driving mileage.However,due to the complex electrochemical reaction and highly coupled nature inside the battery,coupled with the uncertainty of complex traffic conditions,the estimation of battery SOC becomes more difficult.The research content of this paper :In order to accurately describe the dynamic behavior of the battery,an equivalent circuit model containing three fractional-order components is established to replace the ideal capacitance in the traditional integer-order model.Based on the offline parameter identification strategy,this paper uses an improved hybrid particle swarm simplex search optimization algorithm to identify model parameters.This method can effectively avoid the standard particle swarm and simplex search method falling into local optimal solution.At different temperatures,the parameters of the two battery samples are identified and verified.The results show that the fractional order equivalent circuit model based on the model structure has high accuracy and applicability.In this paper,two SOC estimation methods based on fractional order equivalent circuit model of lithium ion battery are improved,which are fractional order cubature Kalman filter(FOCKF)and fractional order square root cubature Kalman filter(FSCKF).The FOCKF method proves the feasibility of applying the cubature Kalman filter(CKF)to the battery fractional-order system.Through simulation experiments under different dynamic conditions,the results show that it has high estimation accuracy.On this basis,the FSCKF method is used to ensure the positive definiteness of the error covariance to improve the stability of the numerical calculation and give full play to the advantages of the fractional order equivalent circuit model in accurately describing the dynamic behavior of the battery.Accurate process noise and measurement noise are important guarantees for the estimation accuracy of Kalman filter method.The noise estimator is added to the FSCKF method to estimate the SOC of lithium-ion battery,which effectively avoids the inaccurate filtering result caused by the constant noise covariance,and overcomes the divergence problem of the filtering algorithm caused by the improper setting of the super parameter in the Kalman filter algorithm.The simulation experiments of SOC estimation of two battery samples are carried out under the initial SOC values of two algorithms and three different dynamic conditions.The results show that the adaptive fractional square root cubature Kalman filter(AFSCKF)with noise estimator has better performance in robustness and accuracy.