• Infrared and Laser Engineering
  • Vol. 51, Issue 12, 20220097 (2022)
Xin Xia1, Chuanliang He1, Yingjie Lv2, Shouzhi Wang2, Bo Zhang2, Chen Chen3, Haipeng Chen4、*, and Meixuan Li5、*
Author Affiliations
  • 1State Grid Laboratory of Power Line Communication Application Technology, Beijing Smart-Chip Microelectronics Techno1ogy Co., Ltd, Beijing 102200, China
  • 2Beijing Electric Power Science & Smart Chip Technology Company Limited, Beijing 100192, China
  • 3College of Instrumentation and Electrical Engineering, Jilin University, Changchun 130026, China
  • 4Department of Electrical Engineering, Northeast Electric Power University, Jilin 132012, China
  • 5Institute for Interdisciplinary Quantum Information Technology, Jilin Engineering Normal University, Changchun 130052, China
  • show less
    DOI: 10.3788/IRLA20220097 Cite this Article
    Xin Xia, Chuanliang He, Yingjie Lv, Shouzhi Wang, Bo Zhang, Chen Chen, Haipeng Chen, Meixuan Li. Image data compression technology of smart grid operation based on deep learning[J]. Infrared and Laser Engineering, 2022, 51(12): 20220097 Copy Citation Text show less
    Correlation of time series data
    Fig. 1. Correlation of time series data
    Structure of a convolutional neural network
    Fig. 2. Structure of a convolutional neural network
    Flow chart of data compression of power grid based on CNN
    Fig. 3. Flow chart of data compression of power grid based on CNN
    Results of image compression in power grid before and after processing based on CNN
    Fig. 4. Results of image compression in power grid before and after processing based on CNN
    Parameter typeValue
    OptimizerAdam
    Batch size64
    Number of convolutional layers3
    Number of convolution kernels(32, 64, 64)
    Learning rate0.4
    Table 1. Optimized hyperparameters results of CNN
    N=0.3 N=0.4 N=0.5 N=0.6 N=0.7 N=0.8
    Compression ratio87.38%84.22%80.34%77.45%60.23%55.31%
    Average precision66.33%68.6274.16%90.35%91.73%93.56%
    Table 2. Data compression accuracy
    IndexANNDBNCNN
    Average precision82.32%87.71%91.73%
    Training time/s168215331271
    Table 3. Comparison results of different models
    Xin Xia, Chuanliang He, Yingjie Lv, Shouzhi Wang, Bo Zhang, Chen Chen, Haipeng Chen, Meixuan Li. Image data compression technology of smart grid operation based on deep learning[J]. Infrared and Laser Engineering, 2022, 51(12): 20220097
    Download Citation