中文摘要：

针对暂态电能质量扰动的实际检测过程中,存在着较强的脉冲噪声和白噪声干扰,影响暂态信息准确提取的问题,设计了有效的滤波算法以在保留信号特征的前提下最大限度地抑制噪声干扰影响,该法是将基于数学形态学的广义形态滤波器作为复数小波变换的前置滤波单元,形成的一种新型形态-复小波变换综合检测算法。仿真结果表明,基于该算法的滤波器不仅可很好地解决电能质量扰动分析中滤除随机噪声和脉冲噪声的困难,还可较好地保持扰动信号的形状和特征。另外用Daubechies实小波构造了相应的正交紧支对称复小波,由其提供的复合信息可准确地检测出扰动并进行时间定位。分别用电压暂降、暂态振荡、短时谐波畸变及微小扰动对所提方法进行了数字仿真验证,结果证实了基于形态-复小波变换综合检测方法的正确性和有效性。

英文摘要：

A variety of impulses and white noises exist in the actual detection of transient power quality disturbances. These noises are obstructive to the accurate extraction of the transient disturbance signals of power quality. So, the effective filter ought to be designed to restrain these noises before the detection and location of transient power quality disturbances. In the way, the characteristics of the original signal can be well retained and the noise interferences are suppressed. Based on Mathematical Morphology theory, the generalized morphological filter is used as the preposed unit of the complex wavelet transform, which are combined as a novel approach of the morphology-complex wavelet detection algorithm. Simulations show that it can effectively filter white noise and impulse noise with the Signal-to-Noise-Ratio （SNR） increased and the Mean Square Error （MSE） decreased. In the paper, the complex wavelet derived by Daubechies real wavelet and its compound information are applied to detect and locate for the filtered results of power quality disturbances with noises. The voltage sag, high frequency oscillation, short-time harmonic and slight amplitude sag signals with noises are used to verify the validity of the filter-location approach respectively. Numerical simulation results show that the proposed detection approach based on morphology-complex wavelet transform is valid and effective.