Power load is characterized by nonlinearity, non-stationarity, and strong variability. Traditional time series forecasting methods are not sufficient to capture its complex features and evolving patterns. To improve the accuracy of short-term power load forecasting, an improved VMD-EEMD-LSTM hybrid model was proposed. First, historical power load data were decomposed by variational mode decomposition (VMD), and intrinsic mode components and a residual term at different frequencies were obtained. Then, the residual term generated by VMD was further decomposed by ensemble empirical mode decomposition (EEMD). After decomposition, the intrinsic mode functions (IMFs) obtained from VMD and EEMD were used as inputs to a long short-term memory (LSTM) network for forecasting. The final forecasting result was obtained by combining the predicted values of all components. Comparative experiments were carried out using actual power load data from a region in Jiangxi Province. The results show that, compared with the VMD-LSTM model, the root mean square error (RMSE) and mean absolute error (MAE) of the proposed model are reduced by 56.57% and 56.03%, respectively. Compared with the EEMD-LSTM model, the RMSE and MAE are reduced by 10.06% and 9.01%, respectively. The results indicate that the proposed model achieves higher forecasting accuracy.
LIU Wei.
Short-term Power Load Forecasting Based on Improved VMD-LSTM Model with Residual Correction[J].
Science Technology and Engineering, 2026, 26(20): 8684-8691 DOI:10.12404/j.issn.1671-1815.2501169
ChenZhenyu, LiuJinbo, LiChen, et al. Ultra short term power load forecasting based on LSTM and XGBoost combination model[J]. Grid Technology, 2020, 44(2): 614-620.
MuChenyu, XueWenbin, MuXianying, et al. Research on short term load forecasting based on VMD-LSTM Attention model[J]. Modern Electronic Technology, 2023, 46(17): 174-178.
ZhangLin, GaoShengqiang, SongYu, et al. Short-term combined forecasting model for power load based on ALIF-VMD secondary decomposition and NGO-CNN-LSTM[J]. Science Technology and Engineering, 2025, 25(11): 4583-4597.
PanYanxia, LiuGuorui, RenJianjing, et al. Multi-task short-term power grid load forecasting oriented to cross-quarter multi-period feature bidirectional clustering and time series migration[J]. Power System Technology, 2025, 49(4): 1479-1490.
AiXin, ZhouZhiyu, WeiYanping, et al. Transferable load bidding strategy based on autoregressive integral moving average model[J]. Power System Automation, 2017, 41(20): 26-31, 104.
AnYingkun, ZhuYongdan. Prediction of power load based on linear regression and exponential smoothing[J]. Power Equipment Management, 2021(5): 177-179.
[17]
AdebiyiA A, AdewumiA O, AyoC K. Comparison of ARIMA and artificial neural networks models for stock price prediction[J]. Journal of Applied Mathematics, 2014, 2014: 614342.
[18]
ParkD C, El-SharkawiM A, MarksR J, et al. Electric load forecasting using an artificial neural network[J]. IEEE Transactions on Power Systems, 1991, 6(2): 442-449.
LiangZhi, SunGuoqiang, LiHucheng, et al. Short term load forecasting based on VMD and PSO optimized deep belief networks[J]. Grid Technology, 2018, 42(2): 598-606
[21]
YangY, YangY. Hybrid method for short-term time series forecasting based on EEMD[J]. IEEE Access, 2020, 8: 61915-61928.
[22]
SongC, ChenX, XiaW, et al. Application of a novel signal decomposition prediction model in minute sea level prediction[J]. Ocean Engineering, 2022, 260: 111961.
[23]
WangC, LiuZ, WeiH, et al. Hybrid deep learning model for short-term wind speed forecasting based on time series decomposition and gated recurrent unit[J]. Complex System Modeling and Simulation, 2021, 1(4): 308-321.
[24]
ChenH, LuT, HeX, et al. An improved DFVMD-LSTM hybrid time series model for predicting total solar irradiance[J]. Acta Geodynamica et Geomaterialia, 2023, 20: 139-153.
[25]
ChenH, LuT, HuangJ, et al. An improved VMD-LSTM model for time-varying GNSS time series prediction with temporally correlated noise: 14[J]. Remote Sensing, 2023, 15(14): 3694.
[26]
WangL, LiuY, LiT, et al. Short-term PV power prediction based on optimized VMD and LSTM[J]. IEEE Access, 2020, 8: 165849-165862.
[27]
HuangY, YanL, ChengY, et al. Coal thickness prediction method based on VMD and LSTM[J]. Electronics, 2022, 11(2): 232.
LiuJie, CongLanmei, XiaYuanyang, et al. Short term power load prediction based on DBO-VMD and IWOA BILSTM neural network combination model[J]. Power System Protection and Control, 2024, 52 (8): 123-133.
[30]
ChenH, LuT, HuangJ, et al. Animproved VMD-EEMD-LSTM time series hybrid prediction model for sea surface height derived from satellite altimetry data[J]. Journal of Marine Science and Engineering, 2023, 11(12): 2386.
[31]
TejaK, TiwariR, MohantyS. Adaptive denoising of ECG using EMD, EEMD and CEEMDAN signal processing techniques[J]. Journal of Physics: Conference Series, 2020, 1706(1): 012077.
[32]
DragomiretskiyK, ZossoD. Variational mode decomposition[J]. IEEE Transactions on Signal Processing, 2014, 62(3): 531-544.
[33]
LuukkoP J J, HelskeJ, RäsänenE. Introducing libeemd: a program package for performing the ensemble empirical mode decomposition[J]. Computational Statistics, 2016, 31(2): 545-557.
YueYoujun, LiuYinghan, ZhaoHui, et al. Short term power load forecasting based on CEEMDAN-SE and DBN[J]. Electrical Measurement and Instrumentation, 2020, 57(17): 59-65.
[36]
SagheerA, KotbM. Time series forecasting of petroleum production using deep LSTM recurrent networks[J]. Neurocomputing, 2019, 323: 203-213.
[37]
LiC, WuY, LinH, et al. ECG denoising method based on an improved VMD algorithm[J]. IEEE Sensors Journal, 2022, 22(23): 22725-22733.
[38]
LiY, WeiH, HanZ, et al. Deep learning-based safety helmet detection in engineering management based on convolutional neural networks[J]. Advances in Civil Engineering, 2020, 2020: 1-10.