基于多时间尺度Cholesky分解AEKF的锂电池SOC估计  

Multi-scale AEKF algorithm based on Cholesky decomposition for lithium-ion battery SOC estimation

作　　者：徐洁玉 王冬青 XU Jieyu;WANG Dongqing(College of Electrical Engineering,Qingdao University,Qingdao 266071,China)

机构地区：[1]青岛大学电气工程学院,山东青岛266071

出　　处：《电工电能新技术》2024年第3期49-55,共7页Advanced Technology of Electrical Engineering and Energy

基　　金：国家自然科学基金项目(61873183)。

摘　　要：建立可靠的锂电池荷电状态估算模型,获取精确估算值已成为锂离子电池组能源和安全管理的核心。选择锂离子电池的二阶等效电路模型为研究对象,提出了一种基于Cholesky分解优化多时间尺度自适应扩展卡尔曼滤波算法。状态方程中,对应不同状态变量子方程,选择不同采样周期,解决不同状态变量的不同时间尺度问题。考虑噪声变化,在扩展卡尔曼滤波的基础上,引入噪声的迭代估计,实现噪声的自适应矫正,结合Cholesky分解方法以克服计算的舍入误差问题。在不同工况下,选用不同型号的锂电池进行实验验证,验证该算法的普适性和有效性。The SOC estimation model construction and the acquisition of accurate estimation value of lithium-ion battery have become the core of its energy and safety management.Taking the second-order equivalent circuit model of lithium-ion battery as the research object,a multi-time scale adaptive extended Kalman filtering algorithm based on Cholesky decomposition is proposed for battery SOC estimation.In the equation of state,corresponding to different sub-equations of different state variables and choose different sampling periods to solve the problem of different time scales of different state variables.Considering the noise change,based on the extended Kalman filter,the iterative estimation of noise is introduced,the adaptive correction of noise is realized,and the Cholesky decomposition method is combined to overcome the calculated rounding error problem.Under different working conditions,different types of lithium batteries are selected for experimental verification to verify the universality and effectiveness of the algorithm.

关 键 词：锂离子电池 多时间尺度 荷电状态 CHOLESKY分解 扩展卡尔曼滤波 

分 类 号：TM912[电气工程—电力电子与电力传动]