IAGA辨识分数阶模型与FOAEKF算法的锂电池SOC估计  

IAGAfor identification of fractional order model and FOAEKF algorithm for lithium battery SOC estimation

作　　者：张梦龙 凌六一[1] 宫兵 邢丽坤[1] ZHANG Menglong;LING Liuyi;GONG Bing;XING Likun(School of Electrical and Information Engineering,Anhui University of Science and Technolog,Huainan Anhui 232001,China)

机构地区：[1]安徽理工大学电气与信息工程学院,安徽淮南232001

出　　处：《电源技术》2022年第6期638-642,共5页Chinese Journal of Power Sources

基　　金：安徽省高校自然科学基金项目重点项目(KJ2019A0106)。

摘　　要：锂电池荷电状态(SOC)的精确估计可以有效提高锂电池的使用寿命和电动汽车工作时的安全系数。为了降低模型误差与工况噪声对SOC估计精度的影响,依据分数阶理论建立锂电池分数阶等效电路模型,采用改进自适应遗传算法(IAGA)对分数阶模型进行参数辨识,结合基于噪声残差新息的分数阶自适应扩展卡尔曼滤波算法(FOAEKF)实现对锂电池SOC的精确估计。在城市道路循环工况(UDDS)下,分别采用FOAEKF、分数阶扩展卡尔曼滤波算法(FOEKF)和整数阶扩展卡尔曼滤波算法(IOEKF)对锂电池SOC进行估计,三者估计的平均绝对误差分别为0.00749、0.01117、0.01574,均方根误差分别为0.00916、0.0138、0.0186。结果表明IAGA-FOAEKF算法提高了复杂工况下电池模型和SOC估计的准确性。The accurate estimation of lithium battery state of charge(SOC)can effectively improve the service life of lithium battery and the operation safety factor of electric vehicle.In order to reduce the influence of model error and operating noise on the SOC estimation accuracy,a fractional order equivalent circuit model of Li-ion battery was established based on the fractional order theory,the improved adaptive genetic algorithm(IAGA)was used for the parameter identification of the fractional order model,and the fractional order adaptive extended Kalman filter algorithm(FOAEKF)based on the new interest of noise residuals was combined to achieve the accurate estimation of Li-ion battery SOC.Under the urban dynamometer driving schedule(UDDS),three algorithms,FOAEKF,fractional order extended Kalman filter algorithm(FOEKF)and integer order extended Kalman filter algorithm(IOEKF),were used to estimate the Li-ion battery SOC.The mean absolute errors of three algorithms are 0.00749,0.01117,0.01574,and the root mean square errors are 0.00916,0.0138,0.0186.The results show that the IAGA-FOAEKF algorithm improves the accuracy of the battery model and SOC estimation under complex operating conditions.

关 键 词：分数阶模型 自适应 遗传算法 扩展卡尔曼滤波 荷电状态 

分 类 号：TM912.9[电气工程—电力电子与电力传动]