Lithium-Ion Battery Modeling And SOC Estimation Based On Fractional Theory

The increasingly serious environmental pollution problems and the shortage of fossil fuels such as oil have promoted the development of electric vehicles.As the auxiliary or main power source of new energy vehicles,the battery is a key component of the power train.So reliable and effective battery management system(BMS)can maximize the use of lithium-ion batteries under the same conditions.It is the key direction in the research field of electric vehicles at present.Modeling and state of charge(SOC)estimation are one of the main difficulties of BMS.The accurate SOC estimation can protect the battery,increase the endurance of the vehicle,improve the performance of the vehicle,reduce the demand for the power battery,and ensure the safe and stable operation of the vehicle.In this paper,the lithium ion battery is modeled by analyzing the impedance spectrum of lithium ion battery based on the fractional calculus theory,and the SOC estimation of the battery is studied based on the fractional order model.The main research work of this paper is as follows.(1)The structure principle and working principle of lithium-ion battery are introduced.The main characteristics of the lithium ion battery are analyzed.Then the battery test system and the lithium-ion battery parameters used in this paper are introduced.Finally,the the main uses of basic performance experiment,working condition experiment and the cycle life test are designed in this paper.And Their specific experimental steps are introduced.The initial capacity and the open circuit voltage(OCV)and other battery performance parameters are obtained.(2)First of all,this paper analyzes the electrochemical impedance spectrogram of lithium-ion battery,and adopts impedance elements with fractional order characteristics to improve the second-order RC integer equivalent circuit model based on the fractional calculus theory,Secondly,a fractional-order equivalent circuit model of lithium-ion battery is established,which can accurately describe the electrochemical processes such as charge transfer reaction,double-layer effect,mass transfer and diffusion of lithium-ion battery.Thirdly,based on the mixed-swarm-based cooperative particle swarm optimization(MCPSO),parameter identification of the fractional-order equivalent circuit model is conducted using the federal city driving schedule(FUDS)lithium-ion battery experimental data in the time domain.The simulation results show that the model has higher accuracy and better robust against different driving conditions and SOC than the 2-RC equivalent circuit model.(3)The accuracy of the SOC estimation based on the combined algorithm of the time integral method and the open circuit voltage is verified by the operating conditions.The results show that the cumulative SOC estimation error of the method increases with time so that the vehicle's static requirement for SOC estimation can not be met.Then two kinds SOC estimation methods based on fractional order model and Kalman filter(KF)algorithm are used to compare the model precision and SOC estimation precision.algorithm.The results show that the two algorithms have advantages and disadvantages.Combined with the advantages and disadvantages of the two algorithms,a SOC estimation method based on adaptive dual Kalman filtering algorithm(DKF)is proposed and verified by simulation.The simulation results show that the model precision is further improved.The algorithm has the performance of auto correction to true value in the case of system deviation,and has a high SOC estimation precision after stability.Based on the difference working conditions the algorithm can still accurately estimate the SOC value and has a strong applicability.(4)With the increase of charge and discharge times,the lithium-ion battery will be aging and the capacity will decay continuously.The error of SOC estimation will gradually increase without considering the aging of the battery,which seriously affects the precision of the SOC estimation.First,the physical failure model of exponential regression is established.The initial parameters of the model are obtained by MATLAB fitting,then the parameter value of the model is updated by particle filtering(PF)and the capacity of the battery is predicted.The SOC estimation method based on the battery capacity prediction is proposed.The simulation shows that the SOC estimation can still maintain a high precision in the case of battery aging.It further illustrates that the SOC estimation method based on capacity prediction proposed in this paper effectively inhibits the effect of battery aging on the accuracy of SOC estimation.