Research On Modeling And SOC Estimation Of LifeO₄Battery For Electric Vehicle

With the global energy shortage and environmental pollution problems increasinglyserious, developing electric vehicles has become an important measure of the automotiveindustry to achieve sustainable development. As an important part of the electric vehicles,Battery Management System(BMS) through the effective control and management of thebattery to protect the safe and reliable operation of the electric vehicles. But the batterymanagement technology is far from mature and it is still an important and difficult problemto enhance the estimation accuracy of the battery State of Charge (SOC) in the BMSresearch. This paper takes the LiFeO4battery as the research object, separately in-depthstudy of the battery model and SOC estimation method which affect the SOC estimationaccuracy.Firstly, this paper introduces the internal structure and working principle of LiFePO4battery, analyzes the main factors which affect the performance of the battery, then throughexperiments studies the battery’s basic characteristics, such as voltage, internal resistanceand capacity. On this basis, combined with the analysis of the commonly used equivalentcircuit model, this paper proposes a kind of second-order RC equivalent circuit modelwhich considers the voltage hysteresis characteristics, and adopts index fitting method toestimate the initial value of model parameters offline. In order to adapt the batterytime-varying characteristics, this paper also studies parameters online estimation algorithmand uses improved recursive least squares algorithm online estimate the model parameters.Secondly, this paper analyzes the classical SOC estimation methods, and establishesthe LiFePO4battery state-space equations based on second-order RC equivalentcircuit model, then estimates SOC on the basis of extended Kalman filter algorithm andbattery model. In order to enhance the accuracy of SOC estimation, this paper also studiesthe battery SOC and model parameters joint estimation method and the implementationprocess.Finally, this paper validates the proposed theory by the designed BMS experimentalplatform and MATLAB simulation. Impulse discharge experiment shows that second-orderRC equivalent circuit model has the high accuracy and can track the voltage change，by theuse of improved recursive least squares algorithm to online estimate the model parametersfurther enhance the accuracy of the battery model. The experiment of simulated urban roadschedule indicates that the SOC estimation method, based on extended Kalman filter algorithm and battery model, is correct and effective. Through the three kinds of SOCestimation method contrast, verifies the battery SOC and model parameters joint estimationmethod can actually enhance the accuracy of SOC estimation.