车载锂离子动力电池荷电状态与健康状态估计研究

发布时间：2018-05-25 07:35

本文选题：锂离子动力电池 + 荷电状态　；参考：《山东大学》2017年硕士论文

【摘要】：能源衰竭与环境污染问题日益严重,电动汽车是解决该问题的关键途径。动力电池作为电动汽车的储能部件,占电动汽车成本的1/3～1/2,其技术的发展仍无法完全满足电动汽车的需求,使其成为了制约电动汽车规模发展的关键瓶颈。因此,我国将动力电池专项技术攻关列入了国家"十三五"规划中。动力电池作为电动汽车的能量源泉,其荷电状态(State of Charge,SOC)与健康状态(State of Health,SOH)的准确估计是保障动力电池安全高效运行的前提,而温度、老化、工况等因素造成状态的精确估计难。本文以锂离子动力电池为研究对象,对SOC与SOH的精确估计进行了研究。首先概述了动力电池的性能及优缺点,以实验室电池测试平台为依托,设计了动力电池的测试方案,并根据实验数据对动力电池的温度特性、不同放电倍率特性、迟滞特性以及老化特性进行了分析。对现有的电池模型对比分析后,建立了二阶RC等效电路模型。针对动力电池固有的迟滞特性,对充电和放电方向的模型参数分别辨识。根据磷酸铁锂电池和三元电池的不同特性,选择了不同的开路电压函数。通过实验验证模型的精度,并分析了误差的成因以及老化对电池模型的影响。针对现有SOC估计方法的不足,本文采用自适应强跟踪无迹卡尔曼滤波算法估计SOC。该方法能有效的克服扩展卡尔曼滤波(Extend Kalman Filter,EKF)以及无迹卡尔曼滤波算法(Unscented Kalman Filter,UKF)对模型精度要求高,对噪声协方差初值敏感的缺点,并且无需计算复杂的雅克比矩阵。最后通过磷酸铁锂电池和三元电池的不同工况实验对比三种算法对SOC估计的精度,结果表明该算法明显提高了 SOC的估计精度,且对状态突变跟踪能力强,鲁棒性好。在分析了电池SOH定义及影响因素后,以电池的容量作为SOH的标识,对比多种电池容量衰退的模型后,选用幂函数模型来表示电池容量的衰退,并分别采用EKF以及UKF算法来估计电池的容量,研究结果表明UKF算法更能精确的估计电池容量。
[Abstract]:Energy exhaustion and environmental pollution are becoming more and more serious. Electric vehicle is the key way to solve this problem. As the energy storage component of electric vehicle, power battery accounts for 1 / 3 / 1 / 2 of the cost of electric vehicle. The development of its technology still can not fully meet the demand of electric vehicle, which makes it a key bottleneck restricting the development of electric vehicle scale. Therefore, China will power battery special technology into the national "Thirteenth five-year Plan". As a source of energy for electric vehicles, the accurate estimation of the state of charge and the state of health is the prerequisite to ensure the safe and efficient operation of the battery. However, it is difficult to estimate the state accurately due to the factors of temperature, aging and working conditions. In this paper, the accurate estimation of SOC and SOH is studied. Firstly, the performance, advantages and disadvantages of the power battery are summarized. Based on the test platform of the laboratory battery, the test scheme of the power battery is designed. According to the experimental data, the temperature characteristics of the power battery and the characteristics of different discharge rate are analyzed. Hysteresis and aging characteristics were analyzed. After comparing and analyzing the existing battery models, the second order RC equivalent circuit model is established. According to the inherent hysteresis characteristics of the battery, the model parameters of charge and discharge direction are identified respectively. According to the different characteristics of lithium iron phosphate battery and ternary battery, different open circuit voltage functions are selected. The accuracy of the model is verified by experiments, and the cause of the error and the effect of aging on the battery model are analyzed. To overcome the shortcomings of existing SOC estimation methods, this paper uses adaptive strong tracking unscented Kalman filtering algorithm to estimate SOCs. This method can effectively overcome the shortcomings of extended Kalman filter (EKF) and unscented Kalman filter (UKF), which require high accuracy of the model and sensitive to the initial value of noise covariance, and do not need to calculate the complex Jacobian matrix. Finally, the accuracy of the three algorithms for SOC estimation is compared between the lithium iron phosphate battery and the ternary battery under different operating conditions. The results show that the algorithm improves the estimation accuracy of SOC obviously, and has strong ability to track the state mutation and good robustness. After analyzing the definition of battery SOH and its influencing factors, taking the capacity of battery as the mark of SOH, comparing various models of battery capacity decline, the power function model is used to express the decline of battery capacity. EKF and UKF algorithms are used to estimate the battery capacity. The results show that the UKF algorithm can estimate the battery capacity more accurately.
【学位授予单位】：山东大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TM912;U469.72

【相似文献】