Research On State-of-Charge Estimation And Active Equalization Technique Of Lithium-ion Battery

Battery system as the power source of electric vehicles,its performance affects the safety and power of car driving,lithium-ion batteries are widely used in electric vehicles because of their high specific energy,small self-discharge rate and light weight.However,the battery system is composed of a large number of battery cells in series and parallel,and the inconsistency between the battery cells often leads to problems such as performance degradation,capacity attenuation,and thermal runaway of the battery system.State of Charge(SOC)estimation and battery equalization control management of lithium-ion batteries are important to improve battery pack inconsistencies and enhance the performance of the vehicle.Therefore,relying on the key R&D plan of Shaanxi Province "Key Technologies and Safety Testing Equipment Development of High-performance Safe Charging of Electric Vehicles"(2021LLRH-04-03-02).This paper focuses on the study of SOC estimation and equalization techniques for lithium-ion battery.The main research contents are as follows:(1)Aiming at the characteristics and equilibrium control of lithium-ion batteries,the structural materials,charge-discharge reaction principle,characteristic indicators and influencing factors of battery inconsistency of lithium-ion batteries were analyzed,and the inconsistency characteristics of battery packs were analyzed by voltage,capacity and SOC.The equalization technology is used to solve the problem of battery pack inconsistency,and the battery pack equilibrium control technology was analyzed from three aspects: equilibrium topology circuit structure,equilibrium variable and equilibrium control algorithm,which lays a foundation for the research of active balancing technology.(2)Aiming at the lithium-ion battery modelling and SOC estimation of lithium-ion battery,the electrochemical model,neural network model and equivalent circuit model were compared and analyzed for their specific application scenarios.Thevenin’s equivalent circuit model was selected for the study,and capacity tests,OCV(Open Circuit Voltage)tests and DST tests were conducted in the laboratory.A Forgetting Factor Recursive Least Squares(FFRLS)method was proposed to identify the parameters in the model.The accuracy of the model and the effectiveness of the identification algorithm were verified under working conditions.The Singular Value Decomposition(SVD)was used to replace the square root decomposition in the traditional Unscented Kalman Filter(UKF)algorithm,thus solving the problem that the UKF algorithm is terminated due to the non-positive definite covariance matrix.The joint algorithm of FFRLS and SVD-UKF was also proposed to estimate the battery SOC value based on the SVD-UKF algorithm.(3)Aiming at the active equalization of lithium-ion battery packs,the Buck-Boost circuit topology was optimized in conjunction with the design requirements of the equalization system,and a bi-directional energy flow equalization circuit structure based on a power inductor was designed.In order to accelerate the consistency between battery packs and reduce the energy loss in the equalization process,the single-layer equalization topology was optimized and improved,and a hierarchical equalization topology was proposed.A fuzzy control algorithm was proposed to control the switch on/off frequency in the equalization circuit to dynamically adjust the equalization current level.Based on the accurate estimation of the SOC of lithiumion batteries,the SOC is used as the equalization index to design the equalization control strategy.Simulations were carried out in Matlab/Simulink,and the results show that the fuzzy control-based bi-directional hierarchical structure of inductive energy can effectively improve the inconsistency of the battery pack.