基于改进的粒子群优化扩展卡尔曼滤波算法的锂电池模型参数辨识与荷电状态估计

为解决锂电池荷电状态（SOC）难以精确估计的问题，提出了基于改进的粒子群优化扩展卡尔曼滤波（IPSO-EKF）算法预测电池SOC。为减小参数非线性特性影响，重新构建了EKF算法电池状态空间方程，以辨识出的电池模型参数为基础，获得SOC最优估计。采用IPSO算法优化EKF算法噪声方差矩阵，解决系统状态误差协方差矩阵和测量噪声协方差矩阵最优解获取难题，进一步提高SOC的估计精度。计算结果表明：IPSO-EKF算法能够精确地辨识电池模型参数和SOC值，并能够很好地修正状态变量初始误差。

Estimation of Model Parameters and SOC of Lithium Batteries Based on IPSO-EKF

An extended Kalman filter (EKF) which is optimized by the improved particle swarm optimization (IPSO) algorithm is proposed to estimate the state-of-charge (SOC) of battery. A new state space equation applied to EKF algorithm is constituted to reduce the influence of non-linear characteristics of parameters, and the optimal estimation of SOC is obtained based on the real-time identification of battery model parameters. IPSO algorithm is applied to optimize the system state error covariance matrix and measurement noise covariance matrix to improve the estimation accuracy of SOC by solving the problems in achieving the optimal solutions of these covariance matrixes. The results show that the IPSO-EKF algorithm can estimate the model parameters and SOC of battery accurately, and correct the state variable initial error.