Power Data Compression and High-precision Reconstruction Based on Residual Dual Attention Mechanism Network

Under the background of the rapid development of 5G and the gradual deepening of the construction of a strong smart grid, in order to solve the problem of the mismatch between the power system data storage capacity and the data collection and transmission capacity, this paper proposes a power data compression and high-precision reconstruction based on the residual dual attention mechanism network method. In the data compression mechanism, a power data image construction method is designed, and the power data is compressed by Gaussian filtering down-sampling. In the reconstruction mechanism, the channel attention mechanism and the spatial attention mechanism are combined to construct a residual double attention mechanism network, so as to achieve high-precision reconstruction of compressed data. Through the simulation test on the I-BLEND data set, it is verified that the proposed method can effectively compress the power data, reduce the data storage pressure, and achieve more accurate reconstruction results than other super-resolution reconstruction methods.