Estimation Of SOC For Lithium-ion Battery In Electric Vehicle Based On MSTUKF Algorithm

With the increasing seriousness of energy crisis and environmental problems,electric vehicles have gradually became the direction of automotive development due to their advantages of saving energy,zero emission and environmental friendliness.Battery management technology,one of the three key technologies of electric vehicle,plays an important role for the normal operation of electric vehicle.The estimation of state of charge(SOC)is a core task in battery management.The accurate estimation of SOC can prevent over-charge or over-discharge,reduce the abuse and extend the service life of the battery.Lithium-ion battery is taken as the research object to carry out the SOC estimation.The basic contents are as follows:Firstly,in order to better understand the performance of Lithium-ion battery,the definition of SOC is described,the relevant characteristics is introduced and the influence of temperature,aging degree,charge and discharge rate on SOC are studied.Secondly,the advantages and disadvantages of the equivalent circuit models and the electrochemical models are analyzed,and the combined electrochemical model varying with temperature is selected as the battery model.According to the experimental data,the least square method is used to identify the model parameters,and the accuracy of the model is verified at different temperature.Then,the kalman filter algorithm and its several change forms are introduced.Aiming at the problem that the tracking ability of the untracked kalman filter algorithm is poor in model mismatch and state mutation,multiple fading factor matrices in strong tracking kalman filtering algorithm is introduced to force the innovation sequence to be orthogonal.Thus,the tracking ability of the algorithm is improved.On this basis,a strong tracking unscented kalman filter algorithm with multiple fading factors is used to estimate SOC.Finally,the estimation model is built in Simulink software,and the algorithm is applied into calculate SOC.The simulation results are compared with the experimental results to prove that this method can effectively realize the SOC estimation.