Research On SOC And SOP Estimation Method Of Lithium Battery Based On MSOA-Optimized EKF

Against the backdrop of energy conservation and emission reduction,Beijing has made it clear that it will gradually achieve carbon neutrality and carbon peak.In China,many domestic industries such as solar photovoltaic and new energy vehicles are developing rapidly.The onboard battery pack is the power for the stable operation of new energy vehicles,and it can be guaranteed to run in a reliable way supported by BMS.At the same time,the system can dynamically adjust the control strategy based on the estimation results of the battery-related state,so as to effectively use the performance of the power battery.To obtain more accurate related state parameters,the paper develops research regarding the method of obtaining model parameters and accurately estimating battery SOC and SOP based on building the battery research model.Firstly,the static capacity test of lithium-ion cell,open-circuit voltage test,HPPC test and dynamic working operation test including DST,BJDST,FUDS,US06 are completed.Secondly,equivalent circuit model with two RC loops is applied to estimate the model parameters and verify the precision of the results.Then the parameter results of offline parameter identification are verified based on MATLAB/Simulink simulation model.The variation trend of the obtained voltage simulation data is consistent with the measured data,and the maximum voltage error is less than 50 m V.Then the online parameter identification method of KF optimized by EHO is introduced to obtain unbiased estimation of the model parameters.The maximum voltage error when verifying the parameter results by simulation model is less than 32 m V.In contrast,the online parameter identification method has better working operation applicability and result accuracy than that of the offline method,laying a good foundation for the subsequent battery state estimation.Then,in the research of battery SOC estimation,the SOA is used to find the optimal initial value of the covariance of the noise contained in the EKF state equation to reduce the effect of the noise statistical characteristics on the precision of the estimation results of the filter algorithm.Meanwhile,the Logistic chaos initialization,nonlinear convergence factor and outof-bounds processing strategy are used to ensure the population species diversity of SOA and improve algorithm’s global search ability.The results under 5 standard test functions show that MSOA’s solving ability is obviously improved.Then,the performance of SOC estimation algorithm is verified based on the working operation data of DST,BJDST,FUDS and US06.The SOC estimation result obtained by MSOA optimized EKF algorithm is closer to the real SOC value obtained by ampere-hour integration method,and its MAE and RMSE of the estimation errors are smaller than those of EKF,showing good estimation accuracy and robust.Finally,a peak SOP estimation method under multi-constraints is proposed,through which the cyclic DST working operation data under 95%～0% SOC limit is obtained,and the continuous peak SOP estimation under 30 s,120s and 300 s is completed.Compared with a single constraint,the continuous peak SOP under multi-constraints is more practical and the estimation accuracy is obviously improved.