Power System Fault Data Compression Using the Wavelet Transform and Vector Quantification

In practical application of fault analysis in power system, not all information of frequency bands is needed for signal processing. So according to the application characteristics of fault analysis of power system, this paper uses lossless compression for the useful frequency band signal and lossy compression for the unimportant frequency band signal. A data compression method is proposed for fault transient signal according to the special requirements of power system. Firstly, put up the simulation circuit, gather the fault data and then dispersed wavelet transform is employed to decompose the fault signal into information of various frequency bands. Then different quantization and codes are used to frequency bands respectively, such as the method of Huffman code is employed to compress low frequency band, threshold quantification is used for high frequency bands. The simulation in MATLAB6.5 shows that the method proposed has a good compression rate and wide potential in application.