Research On SOC Estimation And Active Balancing Strategy For Lithium-ion Batteries

With the rapid rise of the new energy automobile industry and the increase of the production capacity of major battery manufacturers,automotive lithium-ion batteries will gradually retire after long-term use,and the retired lithium-ion power batteries still have utilization value.How to rationally utilize the retired batteries step by step will become a key research issue in the future.Aiming at the retired lithium-ion battery,this thesis studies the charge estimation method and active balancing technology of retired lithium-ion battery around its battery management system,and finally designs a retired lithium-ion battery management system for energy storage.Take ternary lithium battery as the research object,build a test platform,conduct charge-discharge experiments and pulse test experiments,etc.,and get the real capacity,parameters and other data of lithium-ion battery.You can also obtain parameters such as open-circuit voltage of lithium-ion battery,test the working condition of the battery,and obtain battery data,so as to provide data for electric estimation.The battery models of lithium batteries are explored,and the advantages,disadvantages and state space equations of various battery models are analyzed.In order to improve the accuracy of SOC estimation and active balancing strategy,the third-order RC model with higher accuracy is finally selected as the main research object,and the internal parameters of batteries are identified by experimental method and fading memory least square method,and then their accuracy is compared.The methods of SOC estimation of lithium-ion batteries and their advantages and disadvantages are introduced.Finally,the unscented Kalman filter(UKF)algorithm is selected as the research object,combined with the second-order RC and third-order RC equivalent circuit models.In order to solve the problem of ignoring unknown noise in SOC estimation,an adaptive unscented Kalman filter(3RC-UKF)algorithm based on the third-order RC model is proposed.In order to verify the accuracy and robustness of the algorithm,a third-order battery model is built and the algorithm is simulated according to the working conditions.By comparing four model algorithms(2RC-UKF,2RC-AUKF,3RC-UKF and 3RC-AUKF),the analysis results show that 3RC-AUKF algorithm has stronger estimation accuracy and robustness.In order to solve the obvious inconsistency problem of retired lithium batteries of electric vehicles,this thesis analyzes several active balancing strategies and their advantages and disadvantages,and finally selects the bidirectional flyback converter as the active balancing topology circuit.The equalization strategy adopts the active equalization strategy based on voltage,SOC and voltage and SOC.Using Simulink to build an active equalization system,the equalization effects of the three equalization strategies in this equalization topology are verified.According to the simulation results,the SOC-based equalization strategy has the fastest equalization speed.According to the proposed algorithm,the main control module based on MC9S12XEP100,the voltage and current acquisition module based on LTC6813 and the active equalization module based on LTC3300 are designed.Completed the design of hardware circuit and software algorithm.The proposed SOC estimation algorithm is transplanted into BMS,and the active balancing strategy is designed according to SOC estimation.It is verified under the condition of BBDST.The experimental results show that the SOC estimation error is less than 2%,which improves the active equalization efficiency and proves that the designed BMS have high accuracy.