基于离散GABOR变换的电网瞬态波形滤噪算法

Study of Noise Reduction for Power System Transient Waves Based on Discrete Gabor Transform

研究了一种基于离散Gabor变换的电网瞬态波形噪声滤除算法。根据电网瞬态波形的Gabor系数在联合时频域分布比较集中 ,而噪声的Gabor系数均匀分布于整个联合时频域的不同特点 ,构造时频掩模函数 ,实现信号和噪声的分离。数值计算结果表明该方法有效去除噪声 ,并保持波形的特征。

A new algorithm of noise reduction for power system transient waves based on discrete Gabor transform is presented in this paper. The Gabor coefficients of power system transient waves span a very small area in the joint time frequency domain, whereas the Gabor coefficients of the noise are different. According to this characteristic, a time-frequency mask function can be designed to separate the transient waves from noise. The numerical results demonstrate that this method can efficiently reduce the noise and keep the original wave unchanged.