Research On Estimation Of Lithium Battery SOC Considering The Influence Of Temperature

In recent years,countries are vigorously developing new energy technologies and products that can replace fossil fuels.Lithium batteries have become a representative of new energy products due to their excellent performance,especially in the field of electric vehicles.State of charge(SOC)of power battery is one of the important parameters of electric vehicle,which reflects the amount of its remaining charge.Electric vehicles will be affected by environmental factors during driving,which will bring errors to the estimation of battery SOC.Since most of the current estimation of battery SOC is carried out at room temperature,the research on estimation strategies applicable to different temperatures of SOC has extremely important theoretical and practical significance.In this paper,ternary lithium ion battery is selected as the research object.Firstly,the structure and working principle of lithium ion are analyzed.Based on the self-built experimental platform,the battery measurement experiment was designed to study the working characteristics of the battery.The hybrid pulse power test(HPPC)experiments were carried out at six temperatures(0℃,10℃,20℃,25℃,30℃ and 40℃).Then,the model parameters of the second-order RC equivalent circuit of the battery were identified by the least square method,and the function relationship between the model parameters and the SOC was obtained at different temperatures.The simulation model of the battery equivalent circuit was built through Matlab/Simulink,and the accuracy of the battery equivalent model was verified.Based on the battery equivalent model,the Kalman filter algorithm is used to estimate the SOC.Considering that noise has a great influence on SOC estimation in traditional Kalman filter(KF)and extended Kalman filter(EKF)algorithms,this paper adopts adaptive Kalman filter(AEKF)algorithm.The AEKF algorithm is proved to be accurate and robust by estimating the SOC of lithium battery at room temperature under HPPC and UDDS conditions.Then,the results of estimation of lithium battery SOC by EKF and AEKF algorithms are compared under different temperature conditions,and it is found that the error of AEKF algorithm in estimation of SOC is smaller under high temperature conditions.The accuracy of SOC estimation of EKF algorithm and AEKF algorithm largely depends on the accuracy of battery equivalent model.In the case that the battery equivalent model is not needed,the BP neural network is used to estimate the SOC.Three variables,voltage,current and temperature,were selected as the input and SOC as the output of the neural network.The optimal number of hidden layer nodes of neural network is found by trial and error method,and BP neural network is trained and tested by sample data.The test results show that the SOC estimation based on BP neural network has a good accuracy.Figure[54]table[18]reference[66]...