动力电池SOC估算研究与实现

发布时间：2018-04-29 18:10

本文选题：SOC + 三阶RC等效电路模型　；参考：《桂林电子科技大学》2014年硕士论文

【摘要】：电荷状态(State of Charge，SOC)是电池管理系统(BMS)中的重要参数，准确估算SOC，可保证电池维持在合理的电压范围内，防止由于过充或深放对电池的损伤，延长电池寿命。该文建立了三阶RC等效电路模型，通过双卡尔曼滤波(DKF)方法在线辨识电池模型所有参数，全面准确地反映电池的动态特性，并在MATLAB/Simulink环境下搭建了仿真模型以验证其有效性。针对状态变量较多、耦合性较强、噪声随机性强和可能出现野值的电池系统，运用叠加原理，分解测量方程，将状态变量分开估算，消弱了它们之间的耦合关系，提出了一种基于扩展卡尔曼滤波(EKF)的新型SOC估算方法，即New-EKF算法。在所设计的DKF基础上，运用EKF与所提出的方法在不同电流工况下对SOC进行估算，其结果表明EKF在恒流工况下，SOC估算精度较好，但在变电流工况下，其SOC的估算效果较差，甚至发散，而所提出的新型SOC估算方法在不同电流工况下，SOC估算精度较高，特别适合用于电流剧烈变化的电动汽车用动力电池的SOC估算中，同时验证了所建立电池模型、DKF参数辨识方法及所提出的SOC估算方法的有效性和可行性。为了进一步提高SOC估算精度，运用EKF估算SOC时，加入了自适应滤波方法，即自适应扩展卡尔曼滤波（AEKF）算法，，同时在使用DKF在线辨识模型参数过程中，为了防止由计算舍入误差的影响，导致估算误差矩阵失去非负定性所产生的滤波发散现象，将UD分解方法用于滤波时间更新和状态更新中，增强了算法稳定性，降低了计算复杂度，所提出方法称为UD-DKF方法。基于UD-DKF，运用AEKF算法在不同电流工况下对SOC进行估算，实验结果表明AEKF算法能准确估算SOC，即使在不同SOC初值误差下，也能较快地收敛到真值，具有良好的鲁棒性和收敛性，对噪声有较强的抑制作用。
[Abstract]:The charge state (State of Charge, SOC) is an important parameter in the battery management system (BMS). The accurate estimation of SOC can ensure that the battery is maintained in a reasonable range of voltage and prevents the battery from overcharging or deeply damaging the battery and prolonging the battery life.
The three order RC equivalent circuit model is established in this paper. The two Calman filter (DKF) method is used to identify all the parameters of the battery model online. The dynamic characteristics of the battery are fully and accurately reflected, and a simulation model is built in the MATLAB/Simulink environment to verify its effectiveness.
For the battery system with more state variables, stronger coupling, strong noise randomness and possible field value, the superposition principle is used to decompose measurement equations, to estimate the state variables separately and to weaken the coupling relationship between them. A new SOC estimation method based on extended Calman filter (EKF), that is, New-EKF algorithm, is put forward. On the basis of DKF, EKF and the proposed method are used to estimate the SOC under different current conditions. The results show that the SOC estimation precision is better under the constant current condition, but the estimation effect of SOC is poor and even diverges under the condition of variable current, and the proposed new SOC estimation method has a higher estimation precision of SOC under different current conditions. It is particularly suitable for the SOC estimation of power batteries for electric vehicles with intense current changes, and the validity and feasibility of the established battery model, the DKF parameter identification method and the proposed SOC estimation method are also verified.
In order to further improve the accuracy of SOC estimation, the adaptive filtering method, adaptive extended Calman filter (AEKF) algorithm is added to the estimation of SOC by EKF, and in the process of identifying the model parameters in the DKF online, in order to prevent the influence of the calculation rounding error, the estimation error matrix loses the filtering divergence produced by non negative qualitative analysis. The UD decomposition method is used in filtering time updating and state updating, which enhances the stability of the algorithm and reduces the computational complexity. The proposed method is called the UD-DKF method. Based on UD-DKF, the AEKF algorithm is used to estimate the SOC under different current conditions. The experimental results show that the AEKF algorithm can accurately estimate SOC, even if the initial value of SOC is mistaken. It also converges to the true value quickly, and has good robustness and convergence, and has a strong inhibitory effect on noise.

【学位授予单位】：桂林电子科技大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：TM912

【参考文献】