基于分数阶模型多新息无迹卡尔曼滤波算法的超级电容SOC估计
SOC Estimation of Supercapacitor Based on Fractional Order Model Multi Innovation Unscented Kalman Filtering Algorithm
郑轶 1许永红 2张红光 3童亮 2李力华 4张兆龙5
作者信息
- 1. 北京汽车研究总院有限公司,北京 101300;北京新能源汽车股份有限公司,北京 100175
- 2. 北京信息科技大学机电工程学院,北京 100192
- 3. 北京工业大学环境与生命学部,北京 100124
- 4. 北京新能源汽车股份有限公司,北京 100175
- 5. 北京汽车研究总院有限公司,北京 101300
- 折叠
摘要
对超级电容的SOC估计展开了研究.首先,搭建了超级电容测试平台,用于超级电容的参数辨识,并对超级电容进行了常规性能测试;其次,在不同的环境温度和动态工况下采用多种算法进行超级电容SOC估计.结果表明,采用分数阶模型多新息无迹卡尔曼滤波(FOMIUKF)算法对超级电容SOC的估计精度最高,对超级电容的路端电压跟随情况最好,估计结果的均方根误差和平均绝对误差的最大值分别约为1.8%和1.73%.
Abstract
This paper conducts research on the SOC estimat ion of supercapacitor.Firstly,a supercapacitor testing platform was built,and parameter identification of supercapacitor was carried out,and routine performance testing of supercapacitor was carried out.Secondly,multiple algorithms were used to estimate the SOC of supercapacitor under different environmental temperatures and dynamic operating conditions.The results showed that the fractional order model multiple innovation unscented Kalman filter(FOMIUKF)algorithm had the highest estimation accuracy for the SOC of supercapacitors and the best tracking performance for the terminal voltage of supercapacitor.The maximum root mean square error and mean absolute error of the estimation results are about 1.8%and 1.73%,respectively.
关键词
超级电容/分数阶模型/参数辨识/多新息无迹卡尔曼滤波算法/荷电状态估计Key words
supercapacitor/fractional order model/parameter identification/fractional order model multi innovation unscented Kalman filtering algorithm/estimation of the state of charge引用本文复制引用
基金项目
北京市自然科学基金面上项目(3222014)
北京市自然科学基金青年基金(3244039)
出版年
2024
NETL
NSTL
万方数据