Research About Battery’s SOC Estimation Algorithm For Battery Management System

Battery management system(BMS) is an important part of electric vehicle(EV), for it ensures running safety of EV. Battery’s state of charge(SOC) is one of the most important parameters in BMS. Obtaining accurate SOC can prevent battery from overcharging and over-discharging, prolong battery’s service life and reveal EV’s driving range.In this article commonly used SOC estimation methods are introduced, then a new method called adaptive parameter Kalman filter(APKF) was put forward. This new method was composed by two kinds of mathematic algorithms named Kalman filter and least square method respectively. Kalman filter was used to predict battery’s SOC, and least square method was used to update parameters in Kalman filter. And mathematic deduction was used to prove that this new method was reasonable in mathematical theory. In order to test APKF method, data acquisition system was designed and built to get battery’s discharging data. The theory of this experimental system and each part of system was also introduced in article. According to given battery circuit model, battery’s impulse discharging data were processed to acquire parameters of circuit model.Program of APKF was developed by labview, main parts of program and their functions were presented in detail. Each part of program was developed respectively, and this article mainly introduced the compilation and debug of matrix operation program. Different SOC estimation results were got through computer program, and these results were made into curve graph. By comparing with SOC reference curve and other SOC estimation curves which were got by other methods, it finds out that the SOC estimation curve got by APKF method was closest to the SOC reference curve. And this proved that this new method has higher precision and smaller deviation, So APKF method is much better. By comparing SOC estimation curve got by APKF method with typical SOC estimation curves got by Kalman filter method, the existing imperfections of APKF method were also analyzed.