Research on Power Data Imaging and Compression-Reconstruction Algorithm
Xin Yuan, Qingshan Xu, Yin Wu, Guiyuan Xue · 2021 IEEE Sustainable Power and Energy Conference (iSPEC) · 2021
With the further construction of strong smart grid, a large amount of power data has been generated, which needs to be stored and visualized to facilitate the subsequent analysis of the power system. Aiming at the problems that the traditional power data visualization method is not suitable for the emerging deep learning model and the power system data storage capacity does not match the data acquisition and transmission capability, power data imaging and power data image compression-and-reconstruction method based on embedded zero-tree wavelet image coding is proposed. In power data imaging, HSV color space is used to construct two-dimensional pixel images of power data; In the image compression-and-reconstruction mechanism, the embedded zero-tree wavelet coding is used to compress and decompress the power data image by combining the zero-tree structure with SAQ. Case studies show that the proposed method can effectively visualize power load data, realize effective data compression, reduce data storage pressure, and achieve accurate image reconstruction.