基于分数阶模型多新息UKF动力电池SOC估算研究  

Research on SOC estimation of power battery based on fractional order model multi innovation UKF

作　　者：郑轶[1,2] 许永红 张红光 童亮[3] ZHENG Yi;XU Yonghong;ZHANG Hongguang;TONG Liang(Beijing Automotive Research Institute Co.,Ltd.,Beijing 101300,China;Beijing New Energy Vehicle Co.,Ltd.,Beijing 100175,China;Mechanical Electrical Engineering School,Beijing Information Science and Technology University,Beijing 100192,China;Faculty of Environment and Life,Beijing University of Technology,Beijing 100124,China)

机构地区：[1]北京汽车研究总院有限公司,北京101300 [2]北京新能源汽车股份有限公司,北京100175 [3]北京信息科技大学机电工程学院,北京100192 [4]北京工业大学环境与生命学部,北京100124

出　　处：《电源技术》2024年第9期1777-1788,共12页Chinese Journal of Power Sources

基　　金：北京市自然科学基金面上项目(3244039,3222024);北京市教委一般项目(KM202411232021)。

摘　　要：动力电池管理技术是保障新能源汽车高效、安全和可靠运行的核心和关键。动力电池的荷电状态(SOC)是动力电池管理技术的基础,然而动力电池SOC的不确定影响因素太多,如何精确估算动力电池的SOC成为关键问题。针对SOC难以精确获得的问题,搭建了动力电池测试平台,开展了动力电池的常规性能测试、寿命测试,建立了基于分数阶理论的动力电池分数阶模型,将多新息理论与分数阶模型无迹卡尔曼滤波算法结合,提出了分数阶模型多新息无迹卡尔曼滤波(FOMIUKF)算法,并采用该算法对动力电池进行SOC估算。在不同的环境温度、动态工况、SOC初始值条件下对基于不同算法的动力电池SOC估算精度进行了对比分析。结果表明:基于FOMIUKF算法对动力电池SOC估算结果的平均绝对误差和均方根误差的值最小。在不同的动态工况下,采用FOMIUKF算法对动力电池SOC估算结果的平均绝对误差的最大值约为1.04%,对SOC估算结果的均方根误差最大值约为0.8586%,这表明采用FOMIUKF算法对动力电池SOC估算结果的精度高于EKF、UKF、FOUKF算法。Power battery management technology is the core and key to ensuring the efficient,safe,and reliable operation of new energy vehicles.The State of Charge(SOC)of power batteries is the foundation of power battery management technology.However,there are too many uncertain factors affecting the SOC of power batteries,and how to accurately estimate the SOC of power batteries has become a key issue.Regarding the difficulty in accurately obtaining SOC for power batteries,this paper establishes a power battery testing platform,conducts routine performance testing and power battery life testing,establishes a fractional order model of power batteries based on fractional order theory,and then combines multiple innovation theory with fractional order model unscented Kalman filtering algorithm to propose fractional order model multiple innovation unscented Kalman filtering(FOMIUKF)algorithm,and uses this algorithm to estimate the SOC of power batteries.Comparative analysis was conducted on the estimation accuracy of power battery SOC based on different algo‐rithms under different environmental temperatures,dynamic operating conditions,and initial SOC values.The results show that the average absolute error(MAE)and root mean square error(RMSE)of the power battery SOC estimation results are the smallest based on the FOMIUKF algorithm.Un‐der different dynamic operating conditions,the maximum MAE of the power battery SOC estimation results using the FOMIUKF algorithm is about 1.04%,and the maximum RMSE of the power bat‐tery SOC estimation results is about 0.8586%.This indicates that the accuracy of the FOMIUKF al‐gorithm for power battery SOC estimation results is higher than that of the EKF,UKF,and FOUKF algorithms.

关 键 词：动力电池 分数阶模型 多新息无迹卡尔曼滤波算法 荷电状态 

分 类 号：TM912[电气工程—电力电子与电力传动]