基于小波-卡尔曼滤波的基波分量提取

为满足继电保护准确动作的要求,国内外学者对卡尔曼滤波进行了广泛研究。在系统发生故障期间,电压电流信号中包含大量的非周期分量与谐波分量,这将极大地延迟卡尔曼滤波器的收敛速度。为此,文中结合卡尔曼滤波和小波变换,提出一种用于故障检测及提取基波分量的方法。即运用sym4小波对基波分量估计值进行分解重构,利用高频分量模极大值来检测故障信号,从而更新增益系数和误差协方差矩阵,减少了卡尔曼滤波器的响应时间。结果表明,该算法可精确捕捉故障信号发生时刻,不因干扰噪声影响而产生误差,能快速、准确地提取故障暂态信号中的基波分量,且对频率波动有一定的鲁棒性。

Wavelet-based Kalman Filter for Fundamental Component

In order to satisfy accurate action of relay protection,Kalman filter has been researched by domestic and foreign scholars in recent years.For that Kalman filter has response delay for fault,a novel method that wavelet transform and Kalman filter were mixed to extract the fundamental signal was proposed.The fault could be detected by using sym4 wavelet to transform and code the estimate value of fundamental signal.At the moment,the gain coefficient and the error covariance matrix would change so that the response time of Kalman filter is reduced.The simulation results demonstrate that the approach could exactly extract the fundamental signal of fault signal and detect the time of fault occurring.The approach can also exactly estimate the fundamental signal and was robust enough to system frequency variation within a certain range.