• Laser & Optoelectronics Progress
  • Vol. 58, Issue 24, 2415005 (2021)
Qi He1, Zeyu Hu1, Huifang Xu1、2、*, Wei Song1, and YanLin Du1
Author Affiliations
  • 1College of Information Technology, Shanghai Ocean University, Shanghai 201306, China
  • 2College of Information Technology, Shanghai Jian Qiao University, Shanghai 201306, China
  • show less
    DOI: 10.3788/LOP202158.2415005 Cite this Article Set citation alerts
    Qi He, Zeyu Hu, Huifang Xu, Wei Song, YanLin Du. Sea Surface Temperature Prediction Method Based on Empirical Mode Decomposition-Gated Recurrent Unit Model[J]. Laser & Optoelectronics Progress, 2021, 58(24): 2415005 Copy Citation Text show less
    Structure of proposed prediction model
    Fig. 1. Structure of proposed prediction model
    Structure of GRU neuron
    Fig. 2. Structure of GRU neuron
    SST sequence data of five sites
    Fig. 3. SST sequence data of five sites
    Sequences of five SST sequences after EMD decomposition
    Fig. 4. Sequences of five SST sequences after EMD decomposition
    Comparison between complexity of original sequence and multi-scale complexity after decomposition
    Fig. 5. Comparison between complexity of original sequence and multi-scale complexity after decomposition
    Comparison of prediction results of five experimental data in 3 neural networks
    Fig. 6. Comparison of prediction results of five experimental data in 3 neural networks
    Comparison of experimental results between EMD based on machine learning model and direct machine learning models
    Fig. 7. Comparison of experimental results between EMD based on machine learning model and direct machine learning models
    Comparison between predicted value and real value under 4 different prediction models
    Fig. 8. Comparison between predicted value and real value under 4 different prediction models
    Comparison of supplementary experimental results of SST5
    Fig. 9. Comparison of supplementary experimental results of SST5
    SSTParameterLSTMEMD-LSTMGRUEMD-GRU
    SST1MSE0.19630.10770.17210.0706
    MAE0.32690.20400.32030.1965
    SST2MSE0.23550.14630.20920.0692
    MAE0.36660.28040.35830.1978
    SST3MSE0.48960.30400.29240.0964
    MAE0.37330.23710.38980.2333
    SST4MSE0.38960.26560.45370.2609
    MAE0.47590.39670.50670.3891
    SST5MSE0.87240.38100.87440.2366
    MAE0.65070.44190.69110.3710
    Table 1. Comparison between prediction results of EMD based on machine learning model and those of direct machine learning models
    In-outParameterGRUEMD-GRU
    7--1MSE0.25690.1748
    MAE0.37530.3282
    10--5MSE0.87440.2366
    MAE0.69110.3710
    15--7MSE1.04590.6309
    MAE0.77670.6062
    Table 2. Supplementary experimental results
    Qi He, Zeyu Hu, Huifang Xu, Wei Song, YanLin Du. Sea Surface Temperature Prediction Method Based on Empirical Mode Decomposition-Gated Recurrent Unit Model[J]. Laser & Optoelectronics Progress, 2021, 58(24): 2415005
    Download Citation