Research On The Method Of Estimating The State Of Charge Online For Lithium-ion Battery Based On Multi-model

The state of charge(SOC)is an important parameter to describe the state of lithium-ion battery,and it is the core function of battery management system(BMS).The accurate SOC estimation can effectively prevent the overcharge or overdischarge and prolong the service life of lithium-ion battery.However,accurate battery modeling is the basis of estimating SOC.In this paper,we study the SOC online estimation method based on the multi model.The main contents are following:(1)The basic structure,working principle and basic characteristics of lithium ion battery are introduced.The performance test and SOC online estimation of battery system based on LabVIEW are verified.Battery empirical model and equivalent circuit model and off-line identification of their parameters were accomplished respectively.(2)Aiming at the problem of filter divergence,the square root Unscented Kalman Filter(SRUKF)method is studied based on UKF.The improved SRUKF algorithm using the Variable Forgetting Factor Recursive Least Squares(VFF-RLS)and the expanding KF was proposed to solve the problem that the algorithm needs an accurate model and a priori noise statistics.Firstly,the state space equation is obtained through the empirical model and the Thevenin model.Secondly,parameter adaptive adjustment was realized by VFF-RLS.Then,the noise was adjusted by the expanding KF.Finally,physical experiments which contain simple and complex conditions were designed.The experimental results reveal the improved approach's accuracy is excellent with acceptable robustness.(3)The SOC online estimation algorithm based on multi model was studied based on the improved SRUKF algorithm lithium battery.Firstly,according to the precision analysis of a single model,the multi model method is used to improve the model precision.Secondly,according to the research of multi model structure,we analyzed the corresponding implementation process of adaptive algorithm based on multiple models and adaptive algorithm.Then,physical experiments which contain simple condition and complex condition were designed.The experimental results reveal that this approach's estimate accuracy is excellent with acceptable robustness.Finally,in order to reduce the impact of aging on SOC estimation,we analyze the relationship between SOC and the SOH based on collected data and propose a BP method of battery capacity and compensate the aging of the SOC,improving the accuracy of the SOC estimation.