Battery SOC Estimation Based On Equivalent Circuit Model And Improved Cubature Kalman Filter

Lithium-ion batteries(LIBs)have been widely applied in electric vehicles due to their advantages of long cycle life and high energy density.The State of Charge(SOC)of LIBs is regarded as the fundamental parameter of the battery management system,which are obtained by using the popular approaches with data-drive based,model-based et al.And it cannot be measured by sensor.In this thesis,the improved Cubature Kalman filter algorithm is proposed by combing the second-order equivalent circuit model for the SOC estimation of LIBs.The following research work has been carried out:(1)To gain a more comprehensive understanding of battery performance and optimize the design of SOC estimation algorithms,the working characteristics of the LIBs are analyzed.The effects of battery capacity are explored based on the discharge rate,temperature,and cycle times.And the corresponding test results are analyzed.(2)To address the issues of noise interference and parameter identification accuracy in the operation,the principle of bias compensation is introduced for online identification of battery model parameters.Firstly,a second-order equivalent circuit model and the battery state equation are built.Subsequently,the AFFBCRLS is used to identify the model parameters.The Root Mean Square Error(RMSE)and Mean Absolute Error(MAE)of the simulated model terminal voltage are 0.0076 and 0.0049 by using the proposed method,respectively.Compared to the contractional method,the values are reduced by 25.49% and 2%,respectively.The results demonstrate that the established second-order equivalent circuit model exhibits higher accuracy.(3)To address the issue of SOC inaccurate estimation of LIBs caused by fixed measurement and process noise covariance and the positive-definite decomposition of the covariance matrix,the adaptive square root covariance Kalman filter(ASRCKF)algorithm is proposed.Compared to other method,the proposed ASRCKF shows an accuracy SOC estimation under the dynamic stress test(DST)condition at room temperature.The RMSE and MAE are 0.0023 and 0.0019,respectively.Additionally,it also shows good robustness performance when faces the different temperatures,initial SOC values,and voltage and current biases.The results show that the proposed ASRCKF algorithm achieves satisfactory accuracy,robustness,and low computational cost in SOC estimation.(4)To address the problem of inaccurate selection of the moving estimation window in ASRCKF,the particle swarm optimization(PSO)algorithm is introduced to construct an improved ASRCKF.The proposed method can intelligently select the optimal adaptive window value by using PSO algorithm,which effectively overcome the problem of inaccurate window selection.The DST,FUDS and US06 conditions with different temperatures and initial SOC values are utilized.The results show that the proposed method achieves accurate SOC estimation under different temperatures,operating conditions,and initial SOC values.