基于EKPF的机电暂态过程动态状态估计  

EKPF-Based Dynamic State Estimation of Electromechanical Transient Process

作　　者：杨振瑞 桂建忠 邵帅[1] 赵祥[1] 郑存龙 

机构地区：[1]东北电力大学电气工程学院,吉林吉林132012

出　　处：《中国电力》2015年第11期94-98,共5页Electric Power

摘　　要：根据暂态期间发电机的转子动态特性建立了动态状态估计(DSE)模型,针对扩展卡尔曼滤波(EKF)一阶线性化导致的滤波精度下降甚至滤波发散问题,结合粒子滤波(PF)提出了一种基于扩展卡尔曼粒子滤波(EKPF)的机电暂态过程动态估计方法,采用重采样策略选择粗糙采样以防止样本退化。利用WSCC三机九节点系统实现了EKF、无迹卡尔曼滤波(UKF)和EKPF 3种算法的DSE,仿真结果说明了EKPF算法的有效性,且暂态期间其估计效果明显优于另2种方法。A dynamic state estimation model （DSE） is built according to the dynamic characteristics of generator rotor during e]ectromeehanical transient process. Considering the fact that the Extended Kalman Filter （EKF） has poor tracking accuracy and even filter divergence because of the first-order linearization, a transient process dynamic state estimator is proposed in this paper, which combines the Particle Filter （PF） with the choosing rough sampling strategies to prevent sample degradation. Finally, the EKF-based dynamic state estimation, unscented Kalman filter （UKF） and EKPF are achieved respectively on the three-machine nine-bus system of American West Grid （WSCC）. Sinmlatiou has proved the effectiveness of the EKPF algorithm, and that the filtering performance during electromeehanieal transient process is obviously superior to the other two methods.

关 键 词：扩展卡尔曼粒子滤波(EKPF) 样本退化 机电暂态过程 重采样策略 

分 类 号：TM31[电气工程—电机]