摘要
由于同步相量测量单元(phasor measurement unit,PMU)测得数据中存在误差和噪声,无法直接作为调度和控制的参考数据。提出一种基于无迹变换强跟踪滤波(unscented transformation strong tracking filter,UTSTF)的发电机动态状态估计。该方法利用对称采样策略进行sigma点采样,通过引入渐消因子来修正预测协方差矩阵,在线调整增益矩阵,滤波得到动态过程中发电机状态变量的估计值。算例结果表明,UTSTF无论在跟踪速度、精度以及对噪声的鲁棒性能上较无迹卡尔曼滤波和强跟踪滤波均有所提高。
The phasor measurement unit（PMU） data cannot be used directly by controlling and scheduling due to the errors and noises. This paper proposes a novel dynamic state estimation for synchronous machines, based on unscented transformation of strong tracking filter（UTSTF）. The proposed method uses symmetric sampling strategy and introduces a fading factor to adjust the gain matrix online. The simulation shows that the UTSTF performs well in the accuracy of estimation and the tracking speed, as well as robustness about measurement noise, compared with the unscented kalman filter and strong tracking filter.
出处
《中国电机工程学报》
EI
CSCD
北大核心
2016年第3期615-623,共9页
Proceedings of the CSEE
基金
国家自然科学基金项目(51277052
51107032
1104045)~~
关键词
动态状态估计
发电机
机电暂态
无迹变换强跟踪滤波
鲁棒性
dynamic state estimation（DSE）
synchronous machine
electromechanical transient process
unscented transformation strong tracking filter（UTSTF）
robustness
