【摘要】 针对电动汽车锂离子电池状态估算问题,提出了一种复合的电池荷电状态(state of charge,SOC)估算算法。在固定参数卡尔曼滤波算法的基础上,引入基于遗忘因子的递推最小二乘法(forgotten factor recursive least square,FFRLS)进行电池模型参数在线辨识;基于在线模型参数,利用无迹卡尔曼滤波算法(unscented Kalman filter,UKF)估算电池SOC,实现电池模型参数和SOC的实时联合估算。采集电池充放电实验数据进行离线仿真,结果表明该算法能较好地跟踪电池工作状态,SOC估算误差基本稳定在3%以内。更多还原

【Abstract】 Aiming at solving the problem of battery state estimation in electric vehicles,a complex estimation algorithm of battery state of charge(SOC)was presented.Based on the Kalman filter algorithm of fixed parameters,forgotten factor recursive least square(FFRLS)was adopted to identify the battery model parameters online.Based on the online model parameters,untracked Kalman filter(UKF)was adopted to estimate the battery SOC.Real-time combined estimation of battery model parameters and SOC was realized.The off-line simulation was conducted after collecting the experimental data of battery charging and discharging.The results show that the algorithm can track the battery working state well,and the error of SOC estimation will be basically stable within 3%.更多还原