Interpretable wind speed prediction with multivariate time series and temporal fusion transformers

Wind power has been utilized well in power systems, so steady and successful wind speed forecasting is crucial to security management power grid market economy. To date, most researchers have often discounted the interpretability of prediction models, leading to obscure forecasts. This study puts fo...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Veröffentlicht in:Energy (Oxford) Jg. 252; S. 123990
Hauptverfasser: Wu, Binrong, Wang, Lin, Zeng, Yu-Rong
Format: Journal Article
Sprache:Englisch
Veröffentlicht: Oxford Elsevier Ltd 01.08.2022
Elsevier BV
Schlagworte:
Algorithms
Canada
Decomposition
Deep learning
Economic forecasting
Empirical analysis
energy
Evolutionary computation
Forecasting
Interpretable forecasting
Machine learning
Market economies
market economy
Mathematical models
Meteorological data
Multisource data
Optimization
Performance measurement
prediction
Prediction models
Security management
Sequences
temporal variation
time series analysis
Variational mode decomposition
Wind power
Wind speed
Wind speed forecasting
Canada
Deep learning
Variational mode decomposition
Interpretable forecasting
Wind speed forecasting
Multisource data
ISSN:0360-5442, 1873-6785
Online-Zugang:Volltext
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
Beschreibung
Zusammenfassung:Wind power has been utilized well in power systems, so steady and successful wind speed forecasting is crucial to security management power grid market economy. To date, most researchers have often discounted the interpretability of prediction models, leading to obscure forecasts. This study puts forward a unique forecasting methodology that incorporates notable decomposition techniques, multifactor interpretable forecasting models, and optimization algorithms. In the proposed model, variational mode decomposition is employed to break down the raw wind speed sequence into a set of intrinsic mode functions. Adaptive differential evolution is then used for optimizing several parameters of temporal fusion transformers (TFT) to achieve satisfactory forecasting performance. TFT is a new attention-based deep learning model that puts together high-performance multi-horizon prediction and interpretable insights into temporal dynamics. Empirical studies using eight real-world 1-h wind speed data sets in Albert, Canada, and Five Points, USA demonstrate that the system using the proposed model outperforms those employing other comparable models in nearly all performance metrics. Examples of TFT's interpretable outputs are the importance ranking of the decomposed wind speed sub-sequences and meteorological data and attention analysis of different step lengths. The findings signify substantial progress for wind speed prediction and aid policymakers. •Propose a novel interpretable wind speed prediction methodology named VMD-ADE-TFT.•Impacts of influencing factors are analyzed by TFT's interpretability capability.•VMD-ADE-TFT is used to achieve better forecasting accuracy for eight cases.
Bibliographie:ObjectType-Article-1
SourceType-Scholarly Journals-1
ObjectType-Feature-2
content type line 14
content type line 23
ISSN:0360-5442
1873-6785
DOI:10.1016/j.energy.2022.123990