基于数学形态滤波的电压闪变检测研究

Voltage Flicker Detection Based on Mathematical Morphology Filtering

电压闪变信号中经常混叠有随机噪声,针对Prony近似频谱估计算法对噪声较为敏感的缺陷,将数学形态学滤波和Prony分析相结合应用于电压闪变检测研究。首先利用数学形态学滤波法对含噪闪变信号进行去噪处理,再通过Prony算法和Hilbert变换分析并提取去噪后的特征信息。仿真结果显示,该方法有利于提高Prony算法对闪变信号的精确分析,同时验证了其可行性与准确性。

Voltage flicker signal often aliases random noise. For prony approximate spectrum estimation algorithm is very sensitive to noise, the combination of mathematical morphology filtering and prony analysis is applied in the study of voltage flicker detection. The mathematical morphology filtering method is adopted to denoise the flicker signals with noise. The prony algorithm and the Hilbert transform are then used to analyze and extract the feature information. Simulation results show that this method can effectively improve the accurate analysis of the flicker signal by using prony algorithm, and verified its feasibility and accuracy.