Study On Charging State Of Lithium Iron Phosphate Batteries

With the development of new energy vehicles,higher requirements have been placed on the battery management system,estimation of battery state of charge is one of the most important aspects of battery management systems.The prediction of the remaining capacity of the battery,calculation of the output power,remaining years and judgment of the battery health are of great significance.The traditional SOC estimation method is susceptible to noise and the error is large,but Kalman filter is a closed-loop process for estimating the SOC of the battery.It has strong anti-interference ability and the estimation result is reliable.Therefore,Kalman filter is selected for the SOC estimation.In this paper,the 18650 lithium iron phosphate battery produced by Lishen Company was selected as the research object.Firstly,its performance indexes were analyzed according to the battery selection.The SOC-OCV fitting curve and fitting function were mainly introduced.The relationship has a reference significance for the SOC initial value setting of the subsequent Kalman filter.Then the model is established for the characteristics of lithium iron phosphate battery.Since it is very difficult to establish the electrochemical model,the internal parameters need nearly twenty,and the parameter identification is very difficult.Therefore,based on the second-order RC equivalent electrical model,the HPPC test method is used.The battery is pulse tested to obtain the identification results of the parameters of the model.Then the Simulink is used to build the second-order RC model,and the corresponding identification parameters are input.Comparing the results with the actual measured values of the battery,the model has a good simulation of the actual battery operation.When using the Extended Kalman Filter to simulate the SOC,the EKF algorithm achieves a good follow-up effect under constant current discharge,but the EKF algorithm has a large following error to the SOC under the operating conditions.So according to the specific situation,the measurement noise covariance _kR is given corresponding coefficients,and the working condition noise is compensated to achieve a good following effect.However,the applicable range of the noise compensation EKF algorithm has certain limitations,and it is necessary to assume that the statistical characteristics of the noise are known,white noise with a mean of zero.However,in practical applications,the noise is usually unknown,so an adaptive algorithm is added based on Kalman filter.The adaptive Kalman filter can adjust the system process noise covarianceQ_k and measurement noise covarianceR_k online.The value of the covariance reduces the influence of unknown noise on the SOC estimation,so the anti-interference ability is strong.Compared with the noise compensation EKF algorithm,the adaptive Kalman filter algorithm has a good follow-up effect on SOC when the noise is abrupt.Figure[44]Reference[45]Chart[8]...