摘要
为了改善传统卡尔曼滤波算法估计SOC时量测噪声的影响,提出了将传统卡尔曼滤波算法与模糊控制相结合的动力电池SOC的自适应估计方法。通过实时监控量测噪声实际方差与理论方差之间的差值,实现对量测噪声协方差矩阵的实时在线调整,提高算法在实际应用中的鲁棒性。通过基于联邦城市行驶工况(FUDS)验证混合算法的有效性。结果表明,基于模糊卡尔曼滤波算法的SOC估计最大误差仅为0.21%,高于传统卡尔曼滤波估计精度最大误差0.53%。仿真结果表明,该方法可以有效解决传统卡尔曼滤波算法估计不准和累计误差的问题。
In order to improve the measurement noise effect of SOC estimation presented in traditional Kalman filtering algorithm, a novel adaptive method for the SOC estimation was proposed by combined the traditional EKF with fuzzy logic algorithm. The difference between real-time and theoretical variance was monitored to realize the online adjust for the measurement noise covariance matrix and improve the robustness in practical application. The FUDS driving cycle was used to validate the effectiveness of hybrid algorithm. The verification result shows that the maximum error based on the fuzzy-AEKF is 0.21%, which is better than the accuracy of 0.53% estimated by traditional EKF. The simulation result shows that this method can effectively solve the problem of inaccuracy estimation and accumulative error in traditional EKF algorithm.
出处
《电源技术》
CAS
CSCD
北大核心
2016年第9期1836-1839,1883,共5页
Chinese Journal of Power Sources
基金
国家"863"项目(2011AA11A223)
关键词
动力电池
荷电状态
模糊控制器
自适应卡尔曼滤波
电动汽车
lithium-ion battery
state of charge
fuzzy logic controller
adaptive Kalman filtering
electric vehicles
