二维表示的电能质量故障数据压缩

Compression of 2-D representation of power quality event data

为处理大量的电能质量监测数据,应用图像压缩中的嵌入式零树小波(EZW)算法,压缩二维表示的电能质量故障数据.故障数据按整周期倍数进行截断和重组,一维电能质量故障数据变换为二维;二维小波多分辨率分析分解此二维数据得到小波系数;应用嵌入式零树小波编码完成压缩.数据的二维表示使嵌入式零树小波编码得以直接应用,无需改变算法来适应原一维数据.仿真实验表明,此方法具有压缩率高、速度快和特征不变的特点.压缩时采用7级量化,既节省了67.5%的数据存储空间,又不影响故障数据的特征.

Embedded zerotree wavelet （EZW） algorithm used in image compression was applied to compress the 2-D representation of power quality event data in order to handle mass data from power quality monitoring. 1-D power quality event data were transformed to be 2-D after event data were cut off and reconstructed according to an integer multiple of data period; 2-D wavelet multiresolution analysis was performed to decompose the 2-D data to an array of wavelet coefficients; and EZW was applied to complete compression. 2-D representation allows EZW to be directly applied; so it is unnecessary to change EZW to adapt the original 1-D data. Simulation and test results show that this method has high compressibility, high speed, and invariable characters. When seven-stage quantization was adopted, 67. 5% data storage space was saved and the characters of power quality event data were kept unchanged.