Short-Term Power Load Prediction of VMD-LSTM Based on ISSA Optimization DOI Creative Commons

Shuai Wu,

Huafeng Cai

Applied Sciences, Journal Year: 2025, Volume and Issue: 15(9), P. 5037 - 5037

Published: May 1, 2025

Accurate short-term power load forecasting (STPLF) is critical for balancing electricity supply–demand and ensuring grid reliability. To address the challenges of fluctuating loads inaccurate predictions by conventional methods, this paper presents a novel hybrid framework combining Variational Mode Decomposition (VMD), Long Short-Term Memory (LSTM), Improved Sparrow Search Algorithm (ISSA). First, series decomposed into intrinsic mode functions (IMFs) via VMD, where optimal decomposition order K determined using permutation entropy (PE). Next, IMFs meteorological covariates are reconstructed feature vectors, which then input LSTM network component-wise forecasting, and, finally, prediction results each component to obtain final result. The (ISSA), integrates piecewise chaotic mapping population initialization augment global exploration capability, employed fine-tune hyperparameters, thereby enhancing precision. Finally, two case studies conducted Australian regional data Detu’an City historical records. experimental indicate that proposed model achieves reductions 73.03% 82.97% compared with VMD-LSTM baseline, validating its superior predictive accuracy cross-domain generalization capability.

Language: Английский

The Review of Time Series Prediction Models and Research on Power Load Forecasting DOI

Z. J. Peng,

Xiaoyang Yang, Shenping Xiao

et al.

Lecture notes in electrical engineering, Journal Year: 2025, Volume and Issue: unknown, P. 183 - 193

Published: Jan. 1, 2025

Language: Английский

Citations

0
An Improved Short-Term Electricity Load Forecasting Method: The VMD–KPCA–xLSTM–Informer Model DOI Creative Commons

Jiaxing You,

Huafeng Cai,

Dongxiao Shi

et al.

Energies, Journal Year: 2025, Volume and Issue: 18(9), P. 2240 - 2240

Published: April 28, 2025

This paper proposes a hybrid forecasting method (VMD–KPCA–xLSTM–Informer) based on variational-mode decomposition (VMD), kernel principal component analysis (KPCA), extended long short-term memory network (xLSTM), and the Informer model. First, decomposes original power load data environmental parameter using VMD to capture their multi-scale characteristics. Next, KPCA extracts nonlinear features reduces dimensionality of decomposed modals eliminate redundant information while retaining key features. The xLSTM then models temporal dependencies enhance model’s capability prediction accuracy. Finally, model processes long-sequence improve efficiency. Experimental results demonstrate that VMD–KPCA–xLSTM–Informer achieves an average absolute percentage error (MAPE) as low 2.432% coefficient determination (R2) 0.9532 dataset I, while, II, it attains MAPE 4.940% R2 0.8897. These confirm significantly improves accuracy stability forecasting, providing robust support for system optimization.

Language: Английский

Citations

0
Research on Prediction Method of Coal Spontaneous Combustion Temperature Based on Spatio-Temporal Graph Attention Mechanism with Time-Frequency Domain Lag Feature Fusion DOI
Ningke Xu, Shuang Li

Published: Jan. 1, 2025

Language: Английский

Citations

0
Short-Term Power Load Prediction of VMD-LSTM Based on ISSA Optimization DOI Creative Commons

Shuai Wu,

Huafeng Cai

Applied Sciences, Journal Year: 2025, Volume and Issue: 15(9), P. 5037 - 5037

Published: May 1, 2025

Accurate short-term power load forecasting (STPLF) is critical for balancing electricity supply–demand and ensuring grid reliability. To address the challenges of fluctuating loads inaccurate predictions by conventional methods, this paper presents a novel hybrid framework combining Variational Mode Decomposition (VMD), Long Short-Term Memory (LSTM), Improved Sparrow Search Algorithm (ISSA). First, series decomposed into intrinsic mode functions (IMFs) via VMD, where optimal decomposition order K determined using permutation entropy (PE). Next, IMFs meteorological covariates are reconstructed feature vectors, which then input LSTM network component-wise forecasting, and, finally, prediction results each component to obtain final result. The (ISSA), integrates piecewise chaotic mapping population initialization augment global exploration capability, employed fine-tune hyperparameters, thereby enhancing precision. Finally, two case studies conducted Australian regional data Detu’an City historical records. experimental indicate that proposed model achieves reductions 73.03% 82.97% compared with VMD-LSTM baseline, validating its superior predictive accuracy cross-domain generalization capability.

Language: Английский

Citations

0