基于MPA-DKF的单液流锌镍电池荷电状态估计  

State-of-Charge Estimation of Single-Flow Zinc-Nickel Battery Based on MPA-DKF

作　　者：宋春宁[1] 郑少耿 凌旗金 苏有平 SONG Chun-ning;ZHENG Shao-geng;LING Qi-jin;SU You-ping(College of Electrical Engineering,Guangxi University,Nanning Guangxi 530004,China)

机构地区：[1]广西大学电气工程学院,广西南宁530004

出　　处：《计算机仿真》2023年第1期94-98,124,共6页Computer Simulation

基　　金：国家自然科学基金资助项目(51767005);广西自然科学基金项目(2016GXNSFAA380328)。

摘　　要：基于单液流锌镍电池二阶RC等效电路模型,采用双卡尔曼滤波算法(DKF)联合估计该电池荷电状态(SOC)及欧姆内阻参数,以降低电池模型参数时变对SOC估计精度的影响。同时,考虑噪声统计特性(过程噪声协方差、量测噪声协方差)及状态误差协方差初值对DKF算法估计性能的影响,提出采用海洋捕食者优化(MPA)算法对其进行寻优以更好地发挥DKF算法的估计性能。仿真结果表明,所提出的MPA-DKF算法相较于扩展卡尔曼滤波(EKF)算法及DKF算法在估算单液流锌镍电池SOC时具有更高的估算精度及更强的鲁棒性,其最大估算误差不超过1%,且在SOC初值错误的情况下仍能快速修正至准确值附近。Based on the second-order RC equivalent circuit model of the single-flow zinc-nickel battery, the dual Kalman filter algorithm(DKF) is used to jointly estimate the battery state of charge(SOC) and ohmic resistance parameters to reduce the impact of time-varying model parameters on the accuracy of SOC estimation. At the same time, considering the impact of noise statistical characteristics(process noise covariance, measurement noise covariance) and the initial value of the state error covariance on the estimation performance of the DKF algorithm, it is proposed to use the marine predator optimization(MPA) algorithm to optimize it to enhance the estimated performance of the DKF algorithm. The simulation results show that the proposed MPA-DKF algorithm has higher estimation accuracy and stronger robustness in estimating the SOC of a single-flow zinc-nickel battery than the extended Kalman filter(EKF) algorithm and the DKF algorithm. The maximum estimation error does not exceed 1%,and it can be quickly corrected to near the accurate value when the initial SOC value is wrong.

关 键 词：单液流锌镍电池 双卡尔曼滤波算法 荷电状态 海洋捕食者优化算法 

分 类 号：TP391.9[自动化与计算机技术—计算机应用技术]