中文摘要：

无迹卡尔曼滤波（unscented Kalman filter，UKV）是一种非线性滤波方法。由于假设系统噪声的方差为常数，UKF的估计结果会受到未知系统噪声的影响。为减小未知系统噪声对动态状态估计的影响，提出了一种改进的自适应UKF（adaptive unscented Kalman filter，AUKF）算法。该方法通过在UKF中引入渐消记忆指数加权的Sage-Husa噪声统计估值器，能够估计时变系统噪声的均值和方差。利用IEEE57和IEEE118测试系统，在典型日负荷条件下对AUKF方法的有效性进行仿真验证，结果表明所提出的AUKF算法与传统UKF方法相比，在不增加计算复杂度的同时，能够提高状态估计精度。

英文摘要：

The estimation results of unscented Kalman filter （UKF）, a nonlinear filtering method, are usually affected by the unknown system noise because the assumption of system noise variance is constant. This paper proposed an improved adaptive unscented Kalman filter （AUKF） algorithm used fading memory exponential weighted Sage-Husa filter in order to reduce the effect of unknown system noise in the dynamic state estimate process. The AUKF method can estimate the mean and variance of the system noise. It is verified the effectiveness of the AUKF method for an IEEE 57 and 118 buses test systems in the condition of typicaldaily load, and the simulating results showed that the AUKF delivered better estimation accuracy in the same calculation complexity than the UKF.