基于扩展卡尔曼滤波算法的电网动态状态估计

Dynamic State Estimation in Power System Based on the Extended Kalman Filter

介绍了动态状态估计的卡尔曼滤波原理及扩展卡尔曼滤波(extended Kalman filter,EKF)算法,将其与估计精度和静态状态的加权最小二乘法进行比较,经过MATLAB编程对IEEE9-300节点和某215节点实际系统的测试,发现EKF算法具有明显的优势。为此,针对传统EKF算法存在的主要问题,提出基于时变噪声的改进EKF算法,使动态状态估计在系统正常情况和异常情况下的滤波精度均在理想范围内。电力系统动态状态估计对未来时刻的系统状态、运行轨迹的预测十分重要。

This paper introduces principle of Kalman filter and extended Kalman filter （EKF） algorithm for dynamic state es-timation, and makes a comparison between this dynamic state estimation algorithm with weighted leatt square method for es-timating precision and static state. According to tests on IEEE 9-300 node system and one 215 node system by MATLAB, it finds EKF algorithmhas significant advantages. Therefore, in view of existing problems in traditional EKF algorithm, it proposes an improved EKF algorithm based on time varying noise which can ensure filter tion within a reasonable range in normal state and under abnormal conditions.