Study On Modeling Of Lithium-Ion Battery Based On Temperature Effect And SOC Estimation Based On Ekf Algorithm

In recent years,under the background of energy shortage and environmental pollution caused by traditional fuel vehicles,electric vehicles develope rapidly.Because power batteries play an important role in electric vehicles,in order to ensure the power,safety and operating reliability of power batteries,the existence of a battery management system is particularly necessary.The accurate estimation of the state of charge(SOC)in the battery management system is the bottleneck restricting its development,in order to improve the estimation accuracy of SOC,extend the life of the battery and improve the efficiency of batteries,this article takes ternary lithium-ion battery as the research object,according to the influence of temperature on battery performance,and establishes an improved equivalent circuit battery model,estimate the SOC based on the extended Kalman filter algorithm.The main work and conclusions are as follows:(1)Take the ternary lithium-ion battery produced by LG as the research object,the batteries were tested at different ambient temperatures(-10?,0?,10?,25?,40?).The relationship between the Coulomb efficiency,open circuit voltage,battery capacity and battery surface temperature rise and ambient temperature are obtained.The results show: the Coulomb efficiency of the battery reaches a maximum of 100% when discharged at 0.25 C,25?,and a minimum of 93.1% when discharged at 0.25 C,-10?.The difference between the battery charging and discharging open circuit voltage becomes smaller as the ambient temperature and the SOC increase.When the average open circuit voltage of charging and discharging is within the range of 0.2-1,the corresponding average open circuit voltage at the same SOC value is very small,and when the SOC is within the range of 0-0.2,the difference is large.Under the same discharge rate,when the ambient temperature is low,the temperature rise of the center of the battery surface is greater,battery capacity is the largest when discharged at 0.25 C,25?,and the smallest when discharged at 0.5C,-10?.(2)Based on the traditional second-order Thevenin equivalent circuit model,an improved battery model with respect to ambient temperature and battery surface temperature is established.By using the HPPC test to identify the parameters of the battery model at different ambient temperatures,the variation rules of the battery parameters are obtained.Finally,the Matlab/Simulink is used to verify the improved battery model under open circuit voltage test conditions and dynamic stress test conditions.The results show: Under the condition of open circuit voltage test,when the ambient temperature is 40?,the mean absolute error(MAE)and root mean square error(RMSE)of the improved battery model error are compared to the battery model unimproved is slightly larger.Under dynamic stress test conditions,at an ambient temperature of 25°C,the MAE and RMSE of the improved battery model is the same as the original.At other ambient temperatures,the MAE and RMSE of the error of the improved battery model under two discharge conditions compared with the battery model unimproved has been significantly reduced in most cases.(3)The battery SOC is estimated by using the extended Kalman filter algorithm under different discharge conditions,and the estimation error before and after the battery model improvement is compared with the ampere-hour integration method as the standard.The results show: Under constant current discharge test conditions,only the MAE and RMSE of the SOC estimation error of the improved battery model are larger than the battery model unimproved when discharging at 1.5C,25°C,the MAE in other cases is mostly controlled within 1%,RMSE is mostly not more than 2%,which is better than the battery model unimproved.Under open circuit voltage test conditions,the MAE range of SOC estimation error of the improved battery model is reduced from 1.13%-3.36% to 0.54%-1.06%,and the RMSE range is reduced from 1.90%-4.92% to 0.84%-1.73%.Under dynamic stress test condition,the MAE of SOC estimation error of the improved battery model is reduced from 0.46%-4.17% to 0.37%-0.90%,and the RMSE range is reduced from 0.58%-4.98% to 0.53%-1.33%.Finally,the convergence of the extended Kalman algorithm is discussed,and it is proved that when the initial value and the actual value of the system have errors,the algorithm can make the errors disappear by iteration,and the iteration time is related to the size of the error.This thesis has 34 figures,10 tables,102 references.