Study On The Equalization Method And SOC Estimation Algorithm For Lithium-ion Battery With LiNi_0.5Co_0.2Mn_0.3O₂ Cathode Materials

State of charge?SOC?estimation and equalization are the core of power battery pack control and automatic management.Based on accurate SOC estimation,problems could be identified in packs,and more energy of battery packs could be utilized with a better equalization method.LiNi_0.5Co_0.2Mn_0.3O₂?523?lithium-ion batteries are used as experimental samples,which are first connected in parallel for cycle discharge test,and then are analyzed for developing SOC estimation and balancing method in this paper.The details are as follows:?1?A constant current electronic load and a data acquisition device are built to collect the data of current and voltage in every brunch of parallel connected battery packs.For this experimental platform,the distribution of current in each branch is greatly influenced by the contact resistance between the battery and the battery seat.More specifically,the terminal voltage of the cell with the maximum contact resistance will be greater than others as its corresponding branch current is smaller than others.In the late cycle,the cell whose capacity is reduced heavily would suffer from low branch current because of reduced capacity and not be charged fully.The discharge interval of a more attenuated battery is extended and that of a less attenuated battery is reduced by the parallel connection mode,thus,this mode must increase the inconsistency between batteries.?2?The influences of nominal capacity and the inner OCV?open circuit voltage?-SOC curve on the algorithm estimation are analyzed.The results show that,the estimated error of H-infinity can exceed15%when the cell fades in the cycle.Considering that the capacity of the cell changes more quickly than the curve of OCV-SOC,the parameter of nominal capacity has a greater effect on the estimated results.An improved algorithm based on reverse recursion of historical data which could update the inner parameters of algorithm online is proposed,the results prove the method could ensure the estimated error of H-infinity within 3%,while the error estimated by the method of SMO?sliding mode observer?fluctuates around 4%,and the convergence rate of the initial SOC deviation is lower than the improved method in this paper.?3?24 batteries which are modeled based on the experiment data of 523 lithium-ion battery are connected in string in MATLAB/SIMULINK for studying on balancing methods.The equilibrium performance index,that is,the effective energy ratio u_e,has been redesigned.Compared to the equilibrium efficiency?,u_e can reflect the performance difference between various non energy transfer equilibrium methods.The method 2 based on redundant cells and the method 3 which is based on reconfiguration are proposed.The simulation results show that,the improved method 2provides an equilibrium efficiency which is beyond 99%.Although the consistency between batteries at the end of discharge is worse than method 1,due to the higher value of u_e,the discharge time of method 2 using 6Ah redundant battery is bettery than method 1;the reconfiguration-based equalization method 3 has greatly improved the energy efficiency of the battery pack,as is 35s more than the battery life time of method 1,and the energy loss on the armature resistor is less than that without equalization.However,such an equalization method requires batteries to be connected in series and then in parallel.The reliability of the connection method is not high and the service life of the battery pack cannot be improved.