电动汽车动力电池模型参数在线辨识及SOC估计

发布时间：2018-03-16 00:27

本文选题：SOC估计　切入点：参数辨识　出处：《吉林大学》2014年硕士论文　论文类型：学位论文

【摘要】：近年来，电动汽车发展迅速，作为其核心部件，动力电池成为全球研究的热点，而准确地估计电池荷电状态（State of Charge, SOC）一直是电动汽车电池研究领域的重点和难点。本文围绕提高动力电池SOC估计的精度问题主要做了以下几方面的工作：首先明确了研究动力电池的意义，接着比较了几种常用的动力电池，指出了锂离子动力电池的优越性，讨论了SOC的定义及其影响因素，并对电池SOC估计的常用方法进行了较为全面地介绍和分析。接下来研究了锂离子电池模型，分析比较了几种常用的电化学模型和等效电路模型，综合各模型优缺点，建立了便于工程实现的二阶RC网络等效电路模型，并推导出了其状态空间模型。测取了锂离子电池的开路电压（Open Curcit Valtage,OCV）OCV-SOC关系，对二阶等效电路模型的参数进行离线辨识，并对模型进行了验证。采用适合实时运算的遗忘因子递推最小二乘法在线辨识锂离子电池等效电路模型的参数，并在Matlab/Simulink环境下编写辨识程序，同时封装成为可以在线运行的S-Function形式。在自定义电流放电情况下辨识了等效电路模型各参数，结合锂离子电池特性分析实验结果，证明了辨识方法的正确性和合理性。提出用联合参数辨识的结合卡尔曼滤波的开路电压法（简称“联合估计算法”）来估计锂离子电池的SOC，即使用遗忘因子最小二乘法对模型参数在线辨识、修正，利用卡尔曼滤波算法估计电池的开路电压，再根据开路电压法来估算电池SOC。在这部分中，首先，，讨论了实验结果的参考值；其次，搭建了易于实时在线运行的联合估计算法Simulink模型；最后，在模拟工况下验证了联合估计算法估计SOC的准确性。
[Abstract]:In recent years, electric vehicle (EV) has developed rapidly, as its core component, power battery has become a global research hotspot, and accurate estimation of the state of charge (SOCC) has been the focus and difficulty in the field of EV battery research. The main work of this paper is to improve the accuracy of SOC estimation of power battery. In this paper, the significance of studying power battery is clarified, then several kinds of commonly used battery are compared, the superiority of lithium ion battery is pointed out, and the definition of SOC and its influencing factors are discussed. The common methods of battery SOC estimation are introduced and analyzed. Then, the model of lithium ion battery is studied, and several common electrochemical models and equivalent circuit models are analyzed and compared. Combining the advantages and disadvantages of each model, the equivalent circuit model of second-order RC network is established, which is convenient for engineering implementation. The open circuit voltage open Curcit value OCVOCV-SOC relationship of the lithium ion battery is measured, and the parameters of the second order equivalent circuit model are identified off-line, and the model is verified. The parameters of the equivalent circuit model of lithium ion battery are on-line identified by the recursive least square method of forgetting factor, which is suitable for real-time operation, and the identification program is written under the environment of Matlab/Simulink. At the same time, the encapsulation becomes the S-Function form which can be run online. The parameters of the equivalent circuit model are identified in the case of self-defined current discharge, and the correctness and rationality of the identification method are proved by analyzing the experimental results of the characteristics of the lithium-ion battery. An open-circuit voltage method ("joint estimation algorithm") based on combined parameter identification and Kalman filter is proposed to estimate SOCof lithium-ion batteries, even though the forgetting factor least square method is used to identify and modify the model parameters online. The open circuit voltage of the battery is estimated by Kalman filter algorithm, and the SOC of the battery is estimated according to the open circuit voltage method. In this part, the reference value of the experimental results is discussed firstly; secondly, the reference value of the experimental results is discussed. The Simulink model of joint estimation algorithm is built, which is easy to run on line in real time. Finally, the accuracy of joint estimation algorithm is verified under the simulation condition.
【学位授予单位】：吉林大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：U469.72;TM912

【参考文献】