磷酸铁锂电池SOC在线估算研究  

Research on On-line SOC Estimation of Lithium Iron Phosphate Battery

作　　者：王俊 尹志勇 王勇 郭鑫 任晓琨 WANG Jun;YIN Zhiyong;WANG Yong;GUO Xin;REN Xiaokun(Electrical and Power Engineering Teaching and Research Section,Shijiazhuang Campus of Army Engineering University,Shijiazhuang 050003,China)

机构地区：[1]陆军工程大学石家庄校区电气与电力工程教研室,河北石家庄050003

出　　处：《电工技术》2022年第17期45-49,52,共6页Electric Engineering

基　　金：河北省自然科学基金项目(编号E2019506033)。

摘　　要：针对扩展卡尔曼滤波算法中近似线性化处理及电池实际使用中出现的容量衰减引起的电池状态估算误差,提出基于容量修正的无迹卡尔曼滤波算法完成电池荷电状态估算,实现电池全寿命状态监测。通过分析锂电池工作特性和建立二阶RC等效电路模型,使用在线参数辨识法实现电池动态参数计算,为电池荷电状态估算奠定基础。结合电池实际工作中容量衰减特性,利用容量修正和无迹卡尔曼滤波算法完成电池荷电状态的在线实时估算,提高了估算精度。在恒流放电工况下,利用MATLAB仿真验证,表明无迹卡尔曼滤波算法的估算精度和鲁棒性优于扩展卡尔曼滤波算法,估算误差在3.5%以内。引入容量修正后,对老化情况下容量衰减的电池进行荷电状态估算,相比容量未修正时最大估算误差减小了3%,满足电池全寿命状态估算使用需求。Aiming at the battery state estimation error caused by approximate linearization in the extended Kalman filter algorithm and capacity attenuation in the actual use of the battery,an unscented Kalman filter algorithm based on capacity correction is proposed to estimate the battery state of charge and realize the whole life state monitoring of the battery.By analyzing the working characteristics of lithium battery and establishing the second-order RC equivalent circuit model,the on-line parameter identification method is used to calculate the dynamic parameters of battery,which lays the foundation for battery state of charge estimation.Combined with the capacity attenuation characteristics in the actual work of the battery,the on-line real-time estimation of the battery state of charge is completed by using the capacity correction and unscented Kalman filter algorithm,which improves the estimation accuracy.Under the condition of constant current discharge,MATLAB simulation shows that the estimation accuracy and robustness of Unscented Kalman filter algorithm are better than that of extended Kalman filter algorithm,and the estimation error is within 3.5%.After introducing the capacity correction,the state of charge estimation of the battery with capacity attenuation under aging is carried out,and the maximum estimation error is reduced by 3%compared with that without capacity correction,which meets the needs of battery life-cycle state estimation.

关 键 词：SOC估算 等效电路模型 扩展卡尔曼滤波算法 无迹卡尔曼滤波算法 容量修正 

分 类 号：TM912.9[电气工程—电力电子与电力传动]