针对EKF算法对锂离子SOC估计时,由于系统噪声导致结果不精确的问题,提出了SageHusa-AUKF算法,针对其传统模型结构采用定阶Thevenin/RC模型,存在模型固定不精确的问题,提出AIC智能多阶模型选择方法。通过AIC模型阶数选择的Sage-Husa-AUKF算法,可以很好消除模型误差,提高SOC估计精度,将控制误差在1%以内,具有良好的鲁棒性与收敛性,实验验证了所提方法的有效性,适用于复杂工况下的锂离子SOC估计。

Focused on lithium-ion SOC estimation for EKF algorithm,the result is not accurate due to system noise problem,Sage-Husa AUKF algorithm was proposed,aiming at its traditional model structure by using the first-order Thevenin/RC model,exist the problem of large voltage fluctuation,AIC smart order model selection method is put forward.Through the AIC model order selection of Sage-Husa-AUKF algorithm,error can be eliminated well model,improve the SOC estimation precision,the error is within 1%,has good robustness and convergence,the effectiveness of the proposed method is verified by the experiments and it is suitable for complex conditions of lithium-ion SOC estimation.