Optimizing Power Quality Signal Compression: Harnessing Compressed Sensing and Reconstruction Techniques for Big Data Measurement

The following research proposes a compression technique that combines traditional lossy compression methods with newer ones to identify properties of power quality signals. The data collected undergoes biorthogonal wavelet transformation and filter integration to remove the ripple added to the signal. The system utilizes Matching Pursuit to create an orthogonal dictionary, achieving compression ratios of 846:1. The quality indicators achieved are Percentage of Retained Energy (RTE) = 0.9969, Normalized Mean Squared Error NMSE = 0.0030, and Correlation (COR) = 0.9969, demonstrating the technique’s efficiency. This research’s results surpass the most relevant papers in Q1 journals.