Study On State Of Charge Estimation Of Lithium-ion Battery Pack Used On Pure Electric Vehicle

Estimation of the SOC(State Of Charge)of pure electric vehicle is one of the most important contents of BMS(Battery Management System),accurate estimation of the SOC values of the single cell will be helpful for the single cell to give full play to its performance and protect the single cell from overcharge and over discharge.So far most car companies use open circuit voltage-ampere hour to estimate the soc of a battery pack,the measuring precision and sensitivity by this way is low.kalman filter combined with current time integral use in the estimation of battery SOC can help the SOC to correct its values in real time,its precision is higher in this way.Considering the instability of electric vehicle operating conditions,the author chose first.order thevenin model as single cell equivalent model.The nominal capacity of the sample single lithium iron phosphate cell is 5Ah,the nominal voltage is 3.2V.Take the temperature and currents into consideration,the author identifies the single cell model parameter in HPPC condition under different temperature.Then verify the model accuracy.When estimating the single cell SOC,adding current time integral formula to single cell state space equation,and disperse the obtained complex equation,then rewrite the complex equation into the form of kalman filtering algorithm.Write the m program refer to the obtained complex equation,then run the m program,the result is a curve reflect the relationship between times and SOC.When estimating the battery pack SOC,each point time we have 24 voltage values as reference,we use the mean value of the 24 voltage values as the reference voltage to estimate the battery pack SOC.Then we define the 0 SOC where there is a single cell with its voltage value less than or equal to the discharge cut.off voltage and the 100 SOC where there is a single cell with its voltage value greater than or equal to the charge cut.off voltage.Finally we test the algorithm,we connect two single cells in parallel and then connect 24 parallels in series as the test object.Charge and discharge the battery pack with constant current and variable current,analysis the model output SOC curve.The results show that kalman filter combined with current.time integral algorithm use in SOC estimation can make sure the SOC values to correct its values in real time and it has a higher correction precision.the battery pack SOC can well reflect the dump energy of the battery pack,and the single cell will never overcharge and over discharge,the SOC curve is smooth as well.