Surface Temperature Prediction of East China Sea Based on Variational Mode Decomposition-Long-Short Term Memory-Broad Learning System Hybrid Model

Table 1. Algorithm parameters
View table
Table 1. Algorithm parameters
Model Parameter Value
VMD k 3
Alpha 1
Tol 10^-6
LSTM Layers 3
Neurous ｛64，32，16｝
Dropout P 0.2
BLS S 0.9
C 10^-4
N1，N2，N3 ｛16，30，280｝
Dense Layers 2
Neurous ｛10，1｝
Batch_size 128
Epoch 300

Table 2. Accuracy comparison with existing models

View table

Table 2. Accuracy comparison with existing models

Model	Metrics	T1	T2	T3	T4	T5
STL-LSTM	MAE	0.2694	0.2772	0.2909	0.2870	0.2647
STL-LSTM	RMSE	0.3518	0.3588	0.3678	0.3345	0.3240
EMD-LSTM	MAE	0.2012	0.2130	0.1988	0.2136	0.1823
EMD-LSTM	RMSE	0.2491	0.2612	0.2439	0.2564	0.2349
MC-LSTM	MAE	0.2575	0.2645	0.2472	0.2578	0.2524
MC-LSTM	RMSE	0.3215	0.3264	0.3124	0.3278	0.3164
VMD-LSTM-BLS	MAE	0.0727	0.0784	0.0676	0.0826	0.0766
VMD-LSTM-BLS	RMSE	0.0957	0.1089	0.0864	0.1032	0.1166

Table 3. Comparison of prediction accuracy of models with different time steps

View table

Table 3. Comparison of prediction accuracy of models with different time steps

Model	MAE（1，5，7）			RMSE（1，5，7）
LSTM	0.2545	0.3309	0.3956	0.2845	0.4351	0.4956
GRU	0.2589	0.3504	0.4015	0.3104	0.4527	0.5046
SVR	0.3145	0.4908	0.6804	0.4112	0.6096	0.8063
EMD-LSTM	0.0865	0.2012	0.2735	0.1187	0.2491	0.3045
STL-LSTM	0.1364	0.2694	0.2954	0.1854	0.3518	0.3845
MC-LSTM	0.2036	0.2575	0.2978	0.2342	0.3215	0.3521
VMD-LSTM	0.0627	0.1166	0.1709	0.0854	0.1436	0.1963
LSTM-BLS	0.2366	0.2785	0.3587	0.2688	0.3847	0.4423
VMD-LSTM-BLS	0.0457	0.0727	0.0669	0.0742	0.0957	0.0885

Tools

Get Citation

Copy Citation Text

Ying Han, Kaiqiang Sun, jianing Yan, Changming Dong. Surface Temperature Prediction of East China Sea Based on Variational Mode Decomposition-Long-Short Term Memory-Broad Learning System Hybrid Model[J]. Laser & Optoelectronics Progress, 2023, 60(7): 0701001

Download Citation

EndNote(RIS)BibTex Plain Text

Set citation alerts for article

Save article for my favorites