基于改进Kalman滤波和安时积分的SOC复合估算

SOC estimating method based on improved Kalman Filter and AH integrator

针对当前SOC估算算法中,安时积分法初始值难以估计、库仑系数拟合复杂和卡尔曼滤波法电流变化较快时易出现估值发散的问题,提出了基于闭环反馈的改进卡尔曼滤波算法和基于库仑系数拟合的改进安时积分法相结合的复合估算方法;在闭环卡尔曼滤波稳定状态时,选取此时SOC作为改进安时积分的初始值,后进行改进的安时积分。该方法避免了电流变化时卡尔曼滤波所引起的SOC估值发散问题,通过对12 V阳光A412/50 A电池进行仿真测试,其估值精度达到1.5%。

In accordance with the initial state of charge （SOC） and Coulombic efficiency and variation of battery, a vailable capacity was difficult to measure using the AH counting method. The anti-interference of Kalman filiter was not strong. Improved Kalman filter with feedback loop and improved AH were combined to estimate the compound estimate method. When the improved Kalman filter was stable, the moment of SOC was used as the initial value of AH integration approach. The SOC distortion of Kalman filter was avoided. In the practical application test, the accuracy and good estimate were obtained, and its error was less than 1.5%.