Compression of power quality event data using 2D representation

Abstract

This paper introduces the energy compaction efficiency of a novel two-dimensional (2D) representation for compression of recorded power quality event data. The 2D representation enables to exploit redundancy across vertical samples that correspond to samples which are far from each other in time. In one-dimension (1D), compression algorithms are unable to depict the correlation between such far away samples. However, the 2D representation renders statistically related far away samples near to each other in the vertical dimension, hence several transform domain techniques are able to compress such rendered data efficiently. In this paper, using real life sampled power quality event data, wavelet transform energy compaction efficiencies are compared through a basis restriction error analysis and the results are justified by simple and commercially available 1D and 2D wavelet based coders. The preliminary results indicate that the 2D representation provides a significant energy compaction efficiency