Modeling And State Of Charge Estimation Of Retired Lithium-ion Battery

The market of new energy vehicles is expanding year by year,which means that the ‘scrap tide’ of lithium batteries for vehicles is coming.Therefore,it is very important to retire lithium batteries.This thesis relies on the major science and technology innovation project of Shandong Province “Key Technology Research on Echelon Utilization of Retired Power Lithium Battery”(2019JZZY020810),this thesis took the retired lithium battery as the experimental object,tested its attribute parameters,explored the algorithm of SOC,and finally designed a battery management system for the base station standby power supply based on the theoretical test results.The specific work is as follows:(1)Build ternary lithium battery test platform,including Arbin EVTS,Xingyun battery test system,constant temperature and humidity box.The structure,characteristics and working principle of lithium battery are introduced.In this thesis,the separation of retired lithium battery is studied according to UL1974 document,and the property parameters of ternary lithium battery are tested,including actual capacity,open circuit voltage and so on.(2)The integer order model and fractional order model of ternary lithium battery are explored,and the advantages and state space equations of fractional order model are introduced in detail.Using experimental method and fading memory recursive least squares method to identify the second-order RC internal lithium battery model parameters,internal parameters using genetic algorithm to identify the fractional order model,and under the working condition of HPPC,the parameters identified by the three methods are combined with the corresponding model to estimate the terminal voltage,and the experimental results show that the error of the terminal voltage predicted by the fractional model is the smallest.(3)The concept and significance of lithium battery SOC was analyzed,and the standard Kalman filter,extended Kalman filter,unscented Kalman filter and H-infinite filter were analyzed and compared.Aiming at the problems of colored noise and slow operation speed in the estimation process,an unscented H-infinite filter algorithm based on fractional order model was proposed.The FOUHIF,FOUKF and UKF algorithms were compared and tested using DST,UDDS,NEDC and BBDST,and the errors and uncertainties of the test results are analyzed.Simulation results show that the FOUHIF algorithm has better accuracy and robustness compared with other traditional algorithms.(4)This thesis designs a high-precision BMS according to the parameters obtained in the first four chapters of the thesis and the design algorithm.The BMS can measure voltage,current,temperature and SOC of single battery.MC9S12XEP100 chip of Freescale is used in MCU.LTC6813 is used in voltage acquisition,Hall sensor scheme is used in current acquisition,and thermistor scheme is used in temperature acquisition.At the same time,the FOUHIF algorithm proposed in this thesis is transplanted to the designed BMS platform,and the algorithm is verified experimentally under DST conditions.The experimental results show that the average error of the algorithm is less than 2%,which has a good accuracy.