Combined Estimation Of SOC And SOP For Lithium-ion Battery Based On AUKF Algorithm

As the power source of pure electric vehicles,the most important functions of the battery management system are to optimize the charging and discharging control strategy,prevent the overuse of batteries,and improve the safety of the use of batteries.The realization of these functions all depend on the accurate estimation of battery SOC and SOP.Most traditional SOP estimation only considers the influence of a single factor such as battery terminal voltage or battery SOC.However,in engineering applications,battery voltage,current and SOC all affect the battery’s continuous output power.The coupling relationship between battery SOC and SOP makes SOC and SOP estimation algorithms interdependent.Therefore,based on the AUKF algorithm,this paper conducted joint online estimation of SOC and SOP for lithium ion batteries.The main work is as follows:(1)Taking a type of 2170 cylindrical lithium-ion battery as the research object,the basic structure and working principle of the lithium-ion battery were analyzed,the experimental platform was built,and the constant current discharge experiment,DST and FUDS dynamic condition experiment and HPPC discharge experiment were carried out under different ratios and different temperatures.The voltage,current and SOC experimental data of the battery under the above conditions were collected,which provided sufficient experimental data basis for subsequent battery model parameter identification and accuracy verification of SOC and SOP estimation results.(2)A second-order RC equivalent circuit model was selected,which took into account the complexity and accuracy of the model.The SOC-OCV relationship curve was obtained by polynomial fitting method based on the voltage data and SOC test data obtained by HPPC experiment.The resistance and capacitance parameters of the second order RC model are obtained by the off-line least squares identification algorithm.Since the offline least squares identification algorithm can only be used for specific working conditions and experimental samples,in order to solve the problem that the accuracy of the battery model based on the offline least squares identification is easily affected by the working conditions of the battery in practical application,a recursive least squares algorithm with forgetting factor is proposed to identify the resistance and tolerance parameters of the second-order RC model online.This method can weaken the influence of old data accumulation on parameters and enhance the feedback of new data to the system.After updating the resistance-capacitance parameters of the second-order RC model with the online identification algorithm,the theoretical and measured terminal voltage of the cell were compared and analyzed.It is proved that the established second-order RC equivalent circuit model and the online parameter identification algorithm have high accuracy.(3)Considering that KF and UKF algorithms could not deal with the divergence of filtering system,the UKF algorithm and adaptive algorithm were combined to introduce the adaptive adjustment function of noise variance to approximate the real noise distribution relationship,and the AUKF algorithm was written.The simulation results show that the average absolute error of SOC estimated by AUKF is less than 1.86% under different initial SOC values and different discharge rates.Under DST and FUDS conditions,the average absolute error of SOC estimated by AUKF is less than 0.52%,which verifies the good applicability and robustness of AUKF algorithm.(4)Based on the second-order RC equivalent circuit model,combined with the real-time SOC estimation results of the AUKF algorithm,and considering the terminal voltage constraint and the maximum charge and discharge current limit of the battery design,a combined SOC and SOP algorithm based on the AUKF algorithm was written.Through AUKF co-simulation under dynamic conditions of DST and FUDS at short,medium and long periods(30s,2min,5min)discharge duration,the effectiveness of SOC and SOP combined estimation of the battery was demonstrated,and the maximum safe charging and discharging current limit MAP of the power battery was obtained,providing a reference for the development of the battery management system.