Wind Power Short-Term Prediction Method Based on Time-Domain Dual-Channel Adaptive Learning Model

Driven by dual carbon targets, the scale of wind power integration has surged dramatically. However, its strong volatility causes insufficient short-term prediction accuracy, severely constraining grid security and economic dispatch. To address three key challenges in extracting temporal characteris...

Full description

Saved in:
Bibliographic Details
Main Authors: Haotian Guo, Keng-Weng Lao, Junkun Hao, Xiaorui Hu
Format: Article
Language:English
Published: MDPI AG 2025-07-01
Series:Energies
Subjects:
Online Access:https://www.mdpi.com/1996-1073/18/14/3722
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1849732868539940864
author Haotian Guo
Keng-Weng Lao
Junkun Hao
Xiaorui Hu
author_facet Haotian Guo
Keng-Weng Lao
Junkun Hao
Xiaorui Hu
author_sort Haotian Guo
collection DOAJ
description Driven by dual carbon targets, the scale of wind power integration has surged dramatically. However, its strong volatility causes insufficient short-term prediction accuracy, severely constraining grid security and economic dispatch. To address three key challenges in extracting temporal characteristics of strong volatility, adaptive fusion of multi-source features, and enhancing model interpretability, this paper proposes a Time-Domain Dual-Channel Adaptive Learning Model (TDDCALM). The model employs dual-channel feature decoupling: one Transformer encoder layer captures global dependencies while the raw state layer preserves local temporal features. After TCN-based feature compression, an adaptive weighted early fusion mechanism dynamically optimizes channel weights. The ACON adaptive activation function autonomously learns optimal activation patterns, with fused features visualized through visualization techniques. Validation on two wind farm datasets (A/B) demonstrates that the proposed method reduces RMSE by at least 8.89% compared to the best deep learning baseline, exhibits low sensitivity to time window sizes, and establishes a novel paradigm for forecasting highly volatile renewable energy power.
format Article
id doaj-art-1f91dfbb5d624f048001d9111a661ecb
institution DOAJ
issn 1996-1073
language English
publishDate 2025-07-01
publisher MDPI AG
record_format Article
series Energies
spelling doaj-art-1f91dfbb5d624f048001d9111a661ecb2025-08-20T03:08:12ZengMDPI AGEnergies1996-10732025-07-011814372210.3390/en18143722Wind Power Short-Term Prediction Method Based on Time-Domain Dual-Channel Adaptive Learning ModelHaotian Guo0Keng-Weng Lao1Junkun Hao2Xiaorui Hu3State Key Laboratory of Internet of Things for Smart City, University of Macau, Macao 999078, ChinaState Key Laboratory of Internet of Things for Smart City, University of Macau, Macao 999078, ChinaSchool of Electrical Engineering, Dalian University of Technology, Dalian 116024, ChinaState Key Laboratory of Internet of Things for Smart City, University of Macau, Macao 999078, ChinaDriven by dual carbon targets, the scale of wind power integration has surged dramatically. However, its strong volatility causes insufficient short-term prediction accuracy, severely constraining grid security and economic dispatch. To address three key challenges in extracting temporal characteristics of strong volatility, adaptive fusion of multi-source features, and enhancing model interpretability, this paper proposes a Time-Domain Dual-Channel Adaptive Learning Model (TDDCALM). The model employs dual-channel feature decoupling: one Transformer encoder layer captures global dependencies while the raw state layer preserves local temporal features. After TCN-based feature compression, an adaptive weighted early fusion mechanism dynamically optimizes channel weights. The ACON adaptive activation function autonomously learns optimal activation patterns, with fused features visualized through visualization techniques. Validation on two wind farm datasets (A/B) demonstrates that the proposed method reduces RMSE by at least 8.89% compared to the best deep learning baseline, exhibits low sensitivity to time window sizes, and establishes a novel paradigm for forecasting highly volatile renewable energy power.https://www.mdpi.com/1996-1073/18/14/3722short-term forecast of wind powertime-domain dual-channel adaptive learning modelACONadaptive weighted early fusion
spellingShingle Haotian Guo
Keng-Weng Lao
Junkun Hao
Xiaorui Hu
Wind Power Short-Term Prediction Method Based on Time-Domain Dual-Channel Adaptive Learning Model
Energies
short-term forecast of wind power
time-domain dual-channel adaptive learning model
ACON
adaptive weighted early fusion
title Wind Power Short-Term Prediction Method Based on Time-Domain Dual-Channel Adaptive Learning Model
title_full Wind Power Short-Term Prediction Method Based on Time-Domain Dual-Channel Adaptive Learning Model
title_fullStr Wind Power Short-Term Prediction Method Based on Time-Domain Dual-Channel Adaptive Learning Model
title_full_unstemmed Wind Power Short-Term Prediction Method Based on Time-Domain Dual-Channel Adaptive Learning Model
title_short Wind Power Short-Term Prediction Method Based on Time-Domain Dual-Channel Adaptive Learning Model
title_sort wind power short term prediction method based on time domain dual channel adaptive learning model
topic short-term forecast of wind power
time-domain dual-channel adaptive learning model
ACON
adaptive weighted early fusion
url https://www.mdpi.com/1996-1073/18/14/3722
work_keys_str_mv AT haotianguo windpowershorttermpredictionmethodbasedontimedomaindualchanneladaptivelearningmodel
AT kengwenglao windpowershorttermpredictionmethodbasedontimedomaindualchanneladaptivelearningmodel
AT junkunhao windpowershorttermpredictionmethodbasedontimedomaindualchanneladaptivelearningmodel
AT xiaoruihu windpowershorttermpredictionmethodbasedontimedomaindualchanneladaptivelearningmodel