Abstract: To overcome the problem that real wavelet can not extract the phase information and the extraction of high-frequency coefficients is not very satisfactory in power quality disturbance detection, an orthogonal compactly supported complex wavelet is proposed and applied in power quality disturbance detection. The new algorithm combines Hilbert transform with wavelets transform together, constructs a set of complex values filters, then analyzes the imaginary and real part respectively to extract the phase, amplitude and appropriate combination information. Simulation results show that using such new complex wavelets can extract phase information, thus can detect the mutation more accurately in the power quality disturbance. Comparing with real wavelet and other complex wavelet, the new algorithm can distinguish the type of disturbance and has higher SNR with the same threshold condition in denoising of power quality signal.

Key words: complex wavelets, power quality, Hilbert transform, denoising

摘要： 针对实小波在电能质量扰动检测时不能提取相位信息以及对高频系数提取效果不理想的缺点，提出了正交紧支撑Daubechies复小波的构造方法。即将希尔伯特变换和小波变换结合起来，构造一组复值的滤波器组，然后将其虚部和实部分别进行分析，实现了相位、幅值和适当组合信息等信息的提取。实验仿真结果表明利用该类复小波可提取相位信息，从而更准确地检测出电能质量扰动信号中的突变点，相对实小波能分辨出其扰动的类型，在相同阈值的条件下对电能质量信号进行去噪时，相对实小波及其他复小波去噪后的信噪比有明显的提高。

关键词: 复小波, 电能质量, 希尔伯特变换, 去噪