Abstract: In order to improve the quality of the de-noised power quality signal, an efficient adaptive power quality signal de-noising algorithm based on the Double-Density Discrete Wavelet Transform（DD DWT） is proposed. DD DWT has approximately shift invariance and can accurately describe the real characteristics of signals. Meanwhile, the Bivariate Shrinkage Function（BSF） considers the relationship of the inter-level and intra-level coefficients. So, BSF with local variance estimation is adopted to processing wavelet coefficients. Then the signals are synthesized using the wavelet coefficients processed. The experimental results indicate that, the proposed algorithm can remove the noise more efficiently and keep original power quality signal characters, and that the visual quality of the denoised signal is improved.

Key words: Double-Density Discrete Wavelet Transform（DD DWT）, Bivariate Shrinkage Function（BSF）, power quality signal, Signal Noise Rotio（SNR）

摘要： 为改善电能质量信号的去噪效果，提出一种基于双密度小波变换的自适应电能质量信号去噪算法。双密度小波变换具有近似的平移不变性，能更准确地描述信号的真实特征。而双变量收缩函数充分考虑小波系数的层内层间关系，对小波系数采用结合局部方差估计的双变量收缩函数进行去噪处理，并用收缩后的小波系数重构信号。实验结果表明：该算法在有效滤除噪声的同时，能够更好地保留电能质量信号的特征信息，使去噪信号的视觉信息有较大改善。

关键词: 双密度小波变换, 双变量收缩函数, 电能质量信号, 信噪比