摘要
锂电池荷电状态(SOC)的准确估算是制约电动汽车发展的关键技术之一。针对传统Kalman滤波算法因固定的噪声滤波初值不能够跟随工况变化致使SOC估算不准确的问题,基于PNGV模型建立状态空间方程组,将Sage-Husa自适应滤波算法融合到无迹卡尔曼滤波(UKF)算法之中,对噪声进行实时预测和修正,进而提高SOC的估算精度。仿真实验结果表明,AUKF比UKF的估算值更接近于理论参考值,AUKF解决了UKF因固定噪声带来的误差问题,可提高电动汽车启动、巡航、制动等复杂工况下的电池组电流剧烈变化中SOC的估算精度。
One of the key technologies of restricting the development of electric vehicles is the accurate estimation of lithium battery state of charge（SOC）. The traditional Kalman filter algorithm has fixed initial values that cannot follow the changing work conditions, which leads to the inaccurate SOC estimation. The state-space equations based on the PNGV model is established. The Sage-Husa adaptive filter algorithm is integrated into unscented Kalman filter （UKF） algorithm for predicting and correcting the noise in real-time, so as to improve the estimation accuracy of SOC. The simulation results show that the estimation value of the adaptive unscented Kalman filter （AUKF） is closer to the theoretical reference value than UKF, and AUKF solves the problems of deviation caused by fixed noise in UKF, and can improve the SOC estimation accuracy in the dramatic change of the battery pack current under complex conditions such as starting, cruising and braking of the electric vehicle.
出处
《控制工程》
CSCD
北大核心
2017年第8期1611-1616,共6页
Control Engineering of China
基金
国家自然科学基金(61563006)
广西科技攻关项目(桂科攻1598008-2)
广西高校科研项目(KY2015YB165)
广西重点实验室建设项目(14-045-44
14-A-02-05)
广西汽车零部件与整车技术重点实验室开放基金重点项目(2014KFZD01)
广西研究生教育创新计划项目(YCSZ2014199)
