SlideShare a Scribd company logo
Min-Woo Choi
Network Science Lab
Dept. of Artificial Intelligence
The Catholic University of Korea
E-mail: choimin1231@catholic.ac.kr
2023. 02. 14 Publish: 2022
Journal: Energy (IF: 8.85)
1
οƒ˜ Introduction
β€’ Limitation
β€’ Purpose
β€’ Main contribution
οƒ˜ Methodology
β€’ Data description
β€’ Experiment setup
οƒ˜ Results
οƒ˜ Conclusions
οƒ˜ Limitation and Future work
2
1. Introduction
Limitation
β€’ CNN-based spatio-temporal prediction models require wind information that nodes at regular
intervals or nodes in a square array.
β€’ But actual wind farms are not necessarily arranged regularly.
Related study
β€’ Fu et al. (2019): Spatiotemporal attention network (STAN)
- Using Multi head self attention to extract the spatial correlation.
β€’ Khodayar et al. (2019): Graph convolution deep learning architecture (GCDLA)
- To capture the deep spatiotemporal information of wind speed.
1. Develop a relatively simple multi-node forecasting model by taking advantage of the various
architectures of the above methods.
2. Proposed Spatial-Temporal Graph Transformer Network (STGTN) model to improve short-
term wind speed prediction performance.
Purpose of study
3
1. Introduction
Main contribution
β€’ External attention mechanism is incorporated into the forecasting model
οƒΌ It can capture dynamic spatial information and reduce the complexity of the network.
β€’ A transformer model with graph convolution is proposed
οƒΌ To learn spatial correlation based on the Euclidean distance between wind farms.
4
2. Methodology
Data description
β€’ Location: Danish offshore wind farms
β€’ Period : February 6, 2014 to June 6, 2014
β€’ Number of data: 111 wind turbine node (14,400 points)
β€’ Resolution: 10 min
β€’ Range of data:
β€’ Training/Validation/Test set: 3:1:1
β€’ Using historical data for each 12 points, the wind speed
is predicted for 10 min to 1 h.
5
2. Methodology
Problem formulation
β€’ Position information between wind turbines is represented as a graph
𝐺 = 𝑉, 𝐸, 𝐴
𝑉 = π‘›π‘œπ‘‘π‘’π‘  π‘œπ‘“ 𝑀𝑖𝑛𝑑 𝑑𝑒𝑏𝑖𝑛𝑒𝑠
𝐸 = π‘π‘œπ‘›π‘›π‘’π‘π‘‘π‘–π‘œπ‘›π‘  𝑏𝑒𝑑𝑀𝑒𝑒𝑛 π‘‘β„Žπ‘’ π‘›π‘œπ‘‘π‘’π‘ 
𝐴 = π‘Žπ‘‘π‘—π‘Žπ‘π‘’π‘›π‘π‘¦ π‘šπ‘Žπ‘‘π‘Ÿπ‘–π‘₯ π‘€π‘–π‘‘β„Ž πΈπ‘’π‘π‘™π‘–π‘‘π‘’π‘Žπ‘› π‘‘π‘–π‘ π‘‘π‘Žπ‘›π‘π‘’ 𝑏𝑒𝑑𝑀𝑒𝑒𝑛 π‘›π‘œπ‘‘π‘’π‘ 
Spatialtemporal model of Short-term wind speed forecasting
Experiment setup
β€’ Batch size: 64
β€’ learning rate: 0.01
β€’ Optimization: Adam
β€’ Decay rate of 0.7 after every 10 epochs
𝑉
𝐸 𝐴
1 2 3
2
3
𝑑 𝑑
6
2. Methodology
Graph transformer
Structure of the graph transformer
 The fusion of spatial and
temporal information
 it considered to improve the robustness of
extracting spatial features of wind speed
 higher-level features can be extracted in the
spatial correlation based on Euclidean distance
External attention mechanism
7
2. Methodology
Architecture of the proposed model
Aggregate
spatiotemporal
features
Extract
temporal
feature
Extract
spatial
feature
Form input
through
residual
connection
Fig. 1. Illustration of the proposed spatial-temporal graph transformer network (STGTN).
Main contribution
8
2. Methodology
Benchmark models
β€’ SVR
β€’ DL-STF
β€’ STAN
β€’ STGTN-T (Transformer)
β€’ STGTN (proposed)
Not include spatial information
Based on spatial-temporal information (previous study)
Transformer and MLP
9
3. Results
10 min to 1 h prediction
β€’ The results show that STGTN dominates all
considered methods in terms of the lowest RMSE
values.
β€’ STGTN with MLP outperforms STGTN-T with
transformer.
β€’ Forecasting models based on spatiotemporal
information can track changes of wind speed faster.
β€’ The performance of the STGTN model is more stable
compared with the DL-STF and STAN models.
0 min to 2000 min prediction
Consider the all wind turbine node
10
3. Results
May 20, performance
β€’ These results show that spatial information may degrades the forecasting performance when the
standard deviation of wind speed is small.
May 22, performance Wind speed Condition: fluctuation is small
β€’ SVR is better than DL-STF
β€’ SVR is better than STGTN-T
11
3. Results
June 4, performance
May 21, performance
June 4, performance Wind speed Condition: fluctuation is large
β€’ The MLP and Transformer performs similarly as the feature extractor when large-scale wind speed
fluctuations exist.
 On the whole, STGTN provides more stable and accurate wind speed forecasts compared with
other methods under different wind speed conditions.
12
4. Conclusions
β€’ The results indicate that the proposed model can effectively utilize the spatiotemporal
information to generate more accurate wind speed forecasts.
β€’ Wind speed data of adjacent nodes is sufficiently used to correct the wrong predictions
caused by outliers in the historical data of individual nodes.
β€’ The forecasting results confirms that the proposed model can yield stable wind speed
forecasts regardless of different scales of wind speed fluctuations.
13
5. Limitation and Future work
β€’ Seasonal factors are not considered. (just consider May & June)
β€’ The utilization of wind direction information is not discussed.
14
Thank you!

More Related Content

Similar to NS-CUK Joint Journal Club: Minwoo Choi, Review on "Short-term wind speed forecasting based on spatial-temporal graph transformer networks", Energy 2021

Predicting the Wind - wind farm prospecting with GIS
Predicting the Wind - wind farm prospecting with GISPredicting the Wind - wind farm prospecting with GIS
Predicting the Wind - wind farm prospecting with GIS
Kenex Ltd
Β 
NS-CUK Seminar: S.T.Nguyen, Review on "Weather-Aware Fiber-Wireless Traffic P...
NS-CUK Seminar: S.T.Nguyen, Review on "Weather-Aware Fiber-Wireless Traffic P...NS-CUK Seminar: S.T.Nguyen, Review on "Weather-Aware Fiber-Wireless Traffic P...
NS-CUK Seminar: S.T.Nguyen, Review on "Weather-Aware Fiber-Wireless Traffic P...
ssuser4b1f48
Β 
Presentation: Wind Speed Prediction using Radial Basis Function Neural Network
Presentation: Wind Speed Prediction using Radial Basis Function Neural NetworkPresentation: Wind Speed Prediction using Radial Basis Function Neural Network
Presentation: Wind Speed Prediction using Radial Basis Function Neural Network
Arzam Muzaffar Kotriwala
Β 
Analysis of Time Diversity Gain for Satellite Communication Link based on Ku-...
Analysis of Time Diversity Gain for Satellite Communication Link based on Ku-...Analysis of Time Diversity Gain for Satellite Communication Link based on Ku-...
Analysis of Time Diversity Gain for Satellite Communication Link based on Ku-...
IJECEIAES
Β 
Wind power forecasting: A Case Study in Terrain using Artificial Intelligence
Wind power forecasting: A Case Study in Terrain using Artificial IntelligenceWind power forecasting: A Case Study in Terrain using Artificial Intelligence
Wind power forecasting: A Case Study in Terrain using Artificial Intelligence
IRJET Journal
Β 
Short-term wind speed forecasting system using deep learning for wind turbine...
Short-term wind speed forecasting system using deep learning for wind turbine...Short-term wind speed forecasting system using deep learning for wind turbine...
Short-term wind speed forecasting system using deep learning for wind turbine...
IJECEIAES
Β 
09 huld presentation_61853_4_a
09 huld presentation_61853_4_a09 huld presentation_61853_4_a
IEEE International Conference Presentation
IEEE International Conference PresentationIEEE International Conference Presentation
IEEE International Conference Presentation
Anmol Dwivedi
Β 
KDDCup2022_Teletraan.pdf
KDDCup2022_Teletraan.pdfKDDCup2022_Teletraan.pdf
KDDCup2022_Teletraan.pdf
YueTan7
Β 
Wind Turbines: Will they ever become economically feasible?
Wind Turbines: Will they ever become economically feasible? Wind Turbines: Will they ever become economically feasible?
Wind Turbines: Will they ever become economically feasible?
Jeffrey Funk
Β 
sensors-23-04512-v3.pdf
sensors-23-04512-v3.pdfsensors-23-04512-v3.pdf
sensors-23-04512-v3.pdf
SekharSankuri1
Β 
Very-Short Term Wind Power Forecasting through Wavelet Based ANFIS
Very-Short Term Wind Power Forecasting through Wavelet Based ANFISVery-Short Term Wind Power Forecasting through Wavelet Based ANFIS
Very-Short Term Wind Power Forecasting through Wavelet Based ANFIS
International Journal of Power Electronics and Drive Systems
Β 
Test different neural networks models for forecasting of wind,solar and energ...
Test different neural networks models for forecasting of wind,solar and energ...Test different neural networks models for forecasting of wind,solar and energ...
Test different neural networks models for forecasting of wind,solar and energ...
Tonmoy Ibne Arif
Β 
Wind power prediction using a nonlinear autoregressive exogenous model netwo...
Wind power prediction using a nonlinear autoregressive  exogenous model netwo...Wind power prediction using a nonlinear autoregressive  exogenous model netwo...
Wind power prediction using a nonlinear autoregressive exogenous model netwo...
IJECEIAES
Β 
Program on Mathematical and Statistical Methods for Climate and the Earth Sys...
Program on Mathematical and Statistical Methods for Climate and the Earth Sys...Program on Mathematical and Statistical Methods for Climate and the Earth Sys...
Program on Mathematical and Statistical Methods for Climate and the Earth Sys...
The Statistical and Applied Mathematical Sciences Institute
Β 
Future guidelines the meteorological view - Isabel MartΓ­nez (AEMet)
Future guidelines the meteorological view - Isabel MartΓ­nez (AEMet)Future guidelines the meteorological view - Isabel MartΓ­nez (AEMet)
Future guidelines the meteorological view - Isabel MartΓ­nez (AEMet)
IrSOLaV Pomares
Β 
[20240422_LabSeminar_Huy]Taming_Effect.pptx
[20240422_LabSeminar_Huy]Taming_Effect.pptx[20240422_LabSeminar_Huy]Taming_Effect.pptx
[20240422_LabSeminar_Huy]Taming_Effect.pptx
thanhdowork
Β 
IRJET- Sink Mobility based Energy Efficient Routing Protocol for Wireless Sen...
IRJET- Sink Mobility based Energy Efficient Routing Protocol for Wireless Sen...IRJET- Sink Mobility based Energy Efficient Routing Protocol for Wireless Sen...
IRJET- Sink Mobility based Energy Efficient Routing Protocol for Wireless Sen...
IRJET Journal
Β 
Upcoming Datasets: Global wind map, Jake Badger ( Risoe DTU)
Upcoming Datasets: Global wind map, Jake Badger ( Risoe DTU)Upcoming Datasets: Global wind map, Jake Badger ( Risoe DTU)
Upcoming Datasets: Global wind map, Jake Badger ( Risoe DTU)
IRENA Global Atlas
Β 
115 0115 joshi
115 0115 joshi115 0115 joshi

Similar to NS-CUK Joint Journal Club: Minwoo Choi, Review on "Short-term wind speed forecasting based on spatial-temporal graph transformer networks", Energy 2021 (20)

Predicting the Wind - wind farm prospecting with GIS
Predicting the Wind - wind farm prospecting with GISPredicting the Wind - wind farm prospecting with GIS
Predicting the Wind - wind farm prospecting with GIS
Β 
NS-CUK Seminar: S.T.Nguyen, Review on "Weather-Aware Fiber-Wireless Traffic P...
NS-CUK Seminar: S.T.Nguyen, Review on "Weather-Aware Fiber-Wireless Traffic P...NS-CUK Seminar: S.T.Nguyen, Review on "Weather-Aware Fiber-Wireless Traffic P...
NS-CUK Seminar: S.T.Nguyen, Review on "Weather-Aware Fiber-Wireless Traffic P...
Β 
Presentation: Wind Speed Prediction using Radial Basis Function Neural Network
Presentation: Wind Speed Prediction using Radial Basis Function Neural NetworkPresentation: Wind Speed Prediction using Radial Basis Function Neural Network
Presentation: Wind Speed Prediction using Radial Basis Function Neural Network
Β 
Analysis of Time Diversity Gain for Satellite Communication Link based on Ku-...
Analysis of Time Diversity Gain for Satellite Communication Link based on Ku-...Analysis of Time Diversity Gain for Satellite Communication Link based on Ku-...
Analysis of Time Diversity Gain for Satellite Communication Link based on Ku-...
Β 
Wind power forecasting: A Case Study in Terrain using Artificial Intelligence
Wind power forecasting: A Case Study in Terrain using Artificial IntelligenceWind power forecasting: A Case Study in Terrain using Artificial Intelligence
Wind power forecasting: A Case Study in Terrain using Artificial Intelligence
Β 
Short-term wind speed forecasting system using deep learning for wind turbine...
Short-term wind speed forecasting system using deep learning for wind turbine...Short-term wind speed forecasting system using deep learning for wind turbine...
Short-term wind speed forecasting system using deep learning for wind turbine...
Β 
09 huld presentation_61853_4_a
09 huld presentation_61853_4_a09 huld presentation_61853_4_a
09 huld presentation_61853_4_a
Β 
IEEE International Conference Presentation
IEEE International Conference PresentationIEEE International Conference Presentation
IEEE International Conference Presentation
Β 
KDDCup2022_Teletraan.pdf
KDDCup2022_Teletraan.pdfKDDCup2022_Teletraan.pdf
KDDCup2022_Teletraan.pdf
Β 
Wind Turbines: Will they ever become economically feasible?
Wind Turbines: Will they ever become economically feasible? Wind Turbines: Will they ever become economically feasible?
Wind Turbines: Will they ever become economically feasible?
Β 
sensors-23-04512-v3.pdf
sensors-23-04512-v3.pdfsensors-23-04512-v3.pdf
sensors-23-04512-v3.pdf
Β 
Very-Short Term Wind Power Forecasting through Wavelet Based ANFIS
Very-Short Term Wind Power Forecasting through Wavelet Based ANFISVery-Short Term Wind Power Forecasting through Wavelet Based ANFIS
Very-Short Term Wind Power Forecasting through Wavelet Based ANFIS
Β 
Test different neural networks models for forecasting of wind,solar and energ...
Test different neural networks models for forecasting of wind,solar and energ...Test different neural networks models for forecasting of wind,solar and energ...
Test different neural networks models for forecasting of wind,solar and energ...
Β 
Wind power prediction using a nonlinear autoregressive exogenous model netwo...
Wind power prediction using a nonlinear autoregressive  exogenous model netwo...Wind power prediction using a nonlinear autoregressive  exogenous model netwo...
Wind power prediction using a nonlinear autoregressive exogenous model netwo...
Β 
Program on Mathematical and Statistical Methods for Climate and the Earth Sys...
Program on Mathematical and Statistical Methods for Climate and the Earth Sys...Program on Mathematical and Statistical Methods for Climate and the Earth Sys...
Program on Mathematical and Statistical Methods for Climate and the Earth Sys...
Β 
Future guidelines the meteorological view - Isabel MartΓ­nez (AEMet)
Future guidelines the meteorological view - Isabel MartΓ­nez (AEMet)Future guidelines the meteorological view - Isabel MartΓ­nez (AEMet)
Future guidelines the meteorological view - Isabel MartΓ­nez (AEMet)
Β 
[20240422_LabSeminar_Huy]Taming_Effect.pptx
[20240422_LabSeminar_Huy]Taming_Effect.pptx[20240422_LabSeminar_Huy]Taming_Effect.pptx
[20240422_LabSeminar_Huy]Taming_Effect.pptx
Β 
IRJET- Sink Mobility based Energy Efficient Routing Protocol for Wireless Sen...
IRJET- Sink Mobility based Energy Efficient Routing Protocol for Wireless Sen...IRJET- Sink Mobility based Energy Efficient Routing Protocol for Wireless Sen...
IRJET- Sink Mobility based Energy Efficient Routing Protocol for Wireless Sen...
Β 
Upcoming Datasets: Global wind map, Jake Badger ( Risoe DTU)
Upcoming Datasets: Global wind map, Jake Badger ( Risoe DTU)Upcoming Datasets: Global wind map, Jake Badger ( Risoe DTU)
Upcoming Datasets: Global wind map, Jake Badger ( Risoe DTU)
Β 
115 0115 joshi
115 0115 joshi115 0115 joshi
115 0115 joshi
Β 

More from ssuser4b1f48

NS-CUK Seminar: V.T.Hoang, Review on "GOAT: A Global Transformer on Large-sca...
NS-CUK Seminar: V.T.Hoang, Review on "GOAT: A Global Transformer on Large-sca...NS-CUK Seminar: V.T.Hoang, Review on "GOAT: A Global Transformer on Large-sca...
NS-CUK Seminar: V.T.Hoang, Review on "GOAT: A Global Transformer on Large-sca...
ssuser4b1f48
Β 
NS-CUK Seminar: J.H.Lee, Review on "Graph Propagation Transformer for Graph R...
NS-CUK Seminar: J.H.Lee, Review on "Graph Propagation Transformer for Graph R...NS-CUK Seminar: J.H.Lee, Review on "Graph Propagation Transformer for Graph R...
NS-CUK Seminar: J.H.Lee, Review on "Graph Propagation Transformer for Graph R...
ssuser4b1f48
Β 
NS-CUK Seminar: H.B.Kim, Review on "Cluster-GCN: An Efficient Algorithm for ...
NS-CUK Seminar: H.B.Kim,  Review on "Cluster-GCN: An Efficient Algorithm for ...NS-CUK Seminar: H.B.Kim,  Review on "Cluster-GCN: An Efficient Algorithm for ...
NS-CUK Seminar: H.B.Kim, Review on "Cluster-GCN: An Efficient Algorithm for ...
ssuser4b1f48
Β 
NS-CUK Seminar: H.E.Lee, Review on "Weisfeiler and Leman Go Neural: Higher-O...
NS-CUK Seminar: H.E.Lee,  Review on "Weisfeiler and Leman Go Neural: Higher-O...NS-CUK Seminar: H.E.Lee,  Review on "Weisfeiler and Leman Go Neural: Higher-O...
NS-CUK Seminar: H.E.Lee, Review on "Weisfeiler and Leman Go Neural: Higher-O...
ssuser4b1f48
Β 
NS-CUK Seminar:V.T.Hoang, Review on "GRPE: Relative Positional Encoding for G...
NS-CUK Seminar:V.T.Hoang, Review on "GRPE: Relative Positional Encoding for G...NS-CUK Seminar:V.T.Hoang, Review on "GRPE: Relative Positional Encoding for G...
NS-CUK Seminar:V.T.Hoang, Review on "GRPE: Relative Positional Encoding for G...
ssuser4b1f48
Β 
NS-CUK Seminar: J.H.Lee, Review on "Learnable Structural Semantic Readout for...
NS-CUK Seminar: J.H.Lee, Review on "Learnable Structural Semantic Readout for...NS-CUK Seminar: J.H.Lee, Review on "Learnable Structural Semantic Readout for...
NS-CUK Seminar: J.H.Lee, Review on "Learnable Structural Semantic Readout for...
ssuser4b1f48
Β 
Aug 22nd, 2023: Case Studies - The Art and Science of Animation Production)
Aug 22nd, 2023: Case Studies - The Art and Science of Animation Production)Aug 22nd, 2023: Case Studies - The Art and Science of Animation Production)
Aug 22nd, 2023: Case Studies - The Art and Science of Animation Production)
ssuser4b1f48
Β 
Aug 17th, 2023: Case Studies - Examining Gamification through Virtual/Augment...
Aug 17th, 2023: Case Studies - Examining Gamification through Virtual/Augment...Aug 17th, 2023: Case Studies - Examining Gamification through Virtual/Augment...
Aug 17th, 2023: Case Studies - Examining Gamification through Virtual/Augment...
ssuser4b1f48
Β 
Aug 10th, 2023: Case Studies - The Power of eXtended Reality (XR) with 360Β°
Aug 10th, 2023: Case Studies - The Power of eXtended Reality (XR) with 360Β°Aug 10th, 2023: Case Studies - The Power of eXtended Reality (XR) with 360Β°
Aug 10th, 2023: Case Studies - The Power of eXtended Reality (XR) with 360Β°
ssuser4b1f48
Β 
Aug 8th, 2023: Case Studies - Utilizing eXtended Reality (XR) in Drones)
Aug 8th, 2023: Case Studies - Utilizing eXtended Reality (XR) in Drones)Aug 8th, 2023: Case Studies - Utilizing eXtended Reality (XR) in Drones)
Aug 8th, 2023: Case Studies - Utilizing eXtended Reality (XR) in Drones)
ssuser4b1f48
Β 
NS-CUK Seminar: J.H.Lee, Review on "Learnable Structural Semantic Readout for...
NS-CUK Seminar: J.H.Lee, Review on "Learnable Structural Semantic Readout for...NS-CUK Seminar: J.H.Lee, Review on "Learnable Structural Semantic Readout for...
NS-CUK Seminar: J.H.Lee, Review on "Learnable Structural Semantic Readout for...
ssuser4b1f48
Β 
NS-CUK Seminar: H.E.Lee, Review on "Gated Graph Sequence Neural Networks", I...
NS-CUK Seminar: H.E.Lee,  Review on "Gated Graph Sequence Neural Networks", I...NS-CUK Seminar: H.E.Lee,  Review on "Gated Graph Sequence Neural Networks", I...
NS-CUK Seminar: H.E.Lee, Review on "Gated Graph Sequence Neural Networks", I...
ssuser4b1f48
Β 
NS-CUK Seminar:V.T.Hoang, Review on "Augmentation-Free Self-Supervised Learni...
NS-CUK Seminar:V.T.Hoang, Review on "Augmentation-Free Self-Supervised Learni...NS-CUK Seminar:V.T.Hoang, Review on "Augmentation-Free Self-Supervised Learni...
NS-CUK Seminar:V.T.Hoang, Review on "Augmentation-Free Self-Supervised Learni...
ssuser4b1f48
Β 
NS-CUK Journal club: H.E.Lee, Review on " A biomedical knowledge graph-based ...
NS-CUK Journal club: H.E.Lee, Review on " A biomedical knowledge graph-based ...NS-CUK Journal club: H.E.Lee, Review on " A biomedical knowledge graph-based ...
NS-CUK Journal club: H.E.Lee, Review on " A biomedical knowledge graph-based ...
ssuser4b1f48
Β 
NS-CUK Seminar: H.E.Lee, Review on "PTE: Predictive Text Embedding through L...
NS-CUK Seminar: H.E.Lee,  Review on "PTE: Predictive Text Embedding through L...NS-CUK Seminar: H.E.Lee,  Review on "PTE: Predictive Text Embedding through L...
NS-CUK Seminar: H.E.Lee, Review on "PTE: Predictive Text Embedding through L...
ssuser4b1f48
Β 
NS-CUK Seminar: H.B.Kim, Review on "Inductive Representation Learning on Lar...
NS-CUK Seminar: H.B.Kim,  Review on "Inductive Representation Learning on Lar...NS-CUK Seminar: H.B.Kim,  Review on "Inductive Representation Learning on Lar...
NS-CUK Seminar: H.B.Kim, Review on "Inductive Representation Learning on Lar...
ssuser4b1f48
Β 
NS-CUK Seminar: H.E.Lee, Review on "PTE: Predictive Text Embedding through L...
NS-CUK Seminar: H.E.Lee,  Review on "PTE: Predictive Text Embedding through L...NS-CUK Seminar: H.E.Lee,  Review on "PTE: Predictive Text Embedding through L...
NS-CUK Seminar: H.E.Lee, Review on "PTE: Predictive Text Embedding through L...
ssuser4b1f48
Β 
NS-CUK Seminar: J.H.Lee, Review on "Relational Self-Supervised Learning on Gr...
NS-CUK Seminar: J.H.Lee, Review on "Relational Self-Supervised Learning on Gr...NS-CUK Seminar: J.H.Lee, Review on "Relational Self-Supervised Learning on Gr...
NS-CUK Seminar: J.H.Lee, Review on "Relational Self-Supervised Learning on Gr...
ssuser4b1f48
Β 
NS-CUK Seminar: H.B.Kim, Review on "metapath2vec: Scalable representation le...
NS-CUK Seminar: H.B.Kim,  Review on "metapath2vec: Scalable representation le...NS-CUK Seminar: H.B.Kim,  Review on "metapath2vec: Scalable representation le...
NS-CUK Seminar: H.B.Kim, Review on "metapath2vec: Scalable representation le...
ssuser4b1f48
Β 
NS-CUK Seminar: H.E.Lee, Review on "Graph Star Net for Generalized Multi-Tas...
NS-CUK Seminar: H.E.Lee,  Review on "Graph Star Net for Generalized Multi-Tas...NS-CUK Seminar: H.E.Lee,  Review on "Graph Star Net for Generalized Multi-Tas...
NS-CUK Seminar: H.E.Lee, Review on "Graph Star Net for Generalized Multi-Tas...
ssuser4b1f48
Β 

More from ssuser4b1f48 (20)

NS-CUK Seminar: V.T.Hoang, Review on "GOAT: A Global Transformer on Large-sca...
NS-CUK Seminar: V.T.Hoang, Review on "GOAT: A Global Transformer on Large-sca...NS-CUK Seminar: V.T.Hoang, Review on "GOAT: A Global Transformer on Large-sca...
NS-CUK Seminar: V.T.Hoang, Review on "GOAT: A Global Transformer on Large-sca...
Β 
NS-CUK Seminar: J.H.Lee, Review on "Graph Propagation Transformer for Graph R...
NS-CUK Seminar: J.H.Lee, Review on "Graph Propagation Transformer for Graph R...NS-CUK Seminar: J.H.Lee, Review on "Graph Propagation Transformer for Graph R...
NS-CUK Seminar: J.H.Lee, Review on "Graph Propagation Transformer for Graph R...
Β 
NS-CUK Seminar: H.B.Kim, Review on "Cluster-GCN: An Efficient Algorithm for ...
NS-CUK Seminar: H.B.Kim,  Review on "Cluster-GCN: An Efficient Algorithm for ...NS-CUK Seminar: H.B.Kim,  Review on "Cluster-GCN: An Efficient Algorithm for ...
NS-CUK Seminar: H.B.Kim, Review on "Cluster-GCN: An Efficient Algorithm for ...
Β 
NS-CUK Seminar: H.E.Lee, Review on "Weisfeiler and Leman Go Neural: Higher-O...
NS-CUK Seminar: H.E.Lee,  Review on "Weisfeiler and Leman Go Neural: Higher-O...NS-CUK Seminar: H.E.Lee,  Review on "Weisfeiler and Leman Go Neural: Higher-O...
NS-CUK Seminar: H.E.Lee, Review on "Weisfeiler and Leman Go Neural: Higher-O...
Β 
NS-CUK Seminar:V.T.Hoang, Review on "GRPE: Relative Positional Encoding for G...
NS-CUK Seminar:V.T.Hoang, Review on "GRPE: Relative Positional Encoding for G...NS-CUK Seminar:V.T.Hoang, Review on "GRPE: Relative Positional Encoding for G...
NS-CUK Seminar:V.T.Hoang, Review on "GRPE: Relative Positional Encoding for G...
Β 
NS-CUK Seminar: J.H.Lee, Review on "Learnable Structural Semantic Readout for...
NS-CUK Seminar: J.H.Lee, Review on "Learnable Structural Semantic Readout for...NS-CUK Seminar: J.H.Lee, Review on "Learnable Structural Semantic Readout for...
NS-CUK Seminar: J.H.Lee, Review on "Learnable Structural Semantic Readout for...
Β 
Aug 22nd, 2023: Case Studies - The Art and Science of Animation Production)
Aug 22nd, 2023: Case Studies - The Art and Science of Animation Production)Aug 22nd, 2023: Case Studies - The Art and Science of Animation Production)
Aug 22nd, 2023: Case Studies - The Art and Science of Animation Production)
Β 
Aug 17th, 2023: Case Studies - Examining Gamification through Virtual/Augment...
Aug 17th, 2023: Case Studies - Examining Gamification through Virtual/Augment...Aug 17th, 2023: Case Studies - Examining Gamification through Virtual/Augment...
Aug 17th, 2023: Case Studies - Examining Gamification through Virtual/Augment...
Β 
Aug 10th, 2023: Case Studies - The Power of eXtended Reality (XR) with 360Β°
Aug 10th, 2023: Case Studies - The Power of eXtended Reality (XR) with 360Β°Aug 10th, 2023: Case Studies - The Power of eXtended Reality (XR) with 360Β°
Aug 10th, 2023: Case Studies - The Power of eXtended Reality (XR) with 360Β°
Β 
Aug 8th, 2023: Case Studies - Utilizing eXtended Reality (XR) in Drones)
Aug 8th, 2023: Case Studies - Utilizing eXtended Reality (XR) in Drones)Aug 8th, 2023: Case Studies - Utilizing eXtended Reality (XR) in Drones)
Aug 8th, 2023: Case Studies - Utilizing eXtended Reality (XR) in Drones)
Β 
NS-CUK Seminar: J.H.Lee, Review on "Learnable Structural Semantic Readout for...
NS-CUK Seminar: J.H.Lee, Review on "Learnable Structural Semantic Readout for...NS-CUK Seminar: J.H.Lee, Review on "Learnable Structural Semantic Readout for...
NS-CUK Seminar: J.H.Lee, Review on "Learnable Structural Semantic Readout for...
Β 
NS-CUK Seminar: H.E.Lee, Review on "Gated Graph Sequence Neural Networks", I...
NS-CUK Seminar: H.E.Lee,  Review on "Gated Graph Sequence Neural Networks", I...NS-CUK Seminar: H.E.Lee,  Review on "Gated Graph Sequence Neural Networks", I...
NS-CUK Seminar: H.E.Lee, Review on "Gated Graph Sequence Neural Networks", I...
Β 
NS-CUK Seminar:V.T.Hoang, Review on "Augmentation-Free Self-Supervised Learni...
NS-CUK Seminar:V.T.Hoang, Review on "Augmentation-Free Self-Supervised Learni...NS-CUK Seminar:V.T.Hoang, Review on "Augmentation-Free Self-Supervised Learni...
NS-CUK Seminar:V.T.Hoang, Review on "Augmentation-Free Self-Supervised Learni...
Β 
NS-CUK Journal club: H.E.Lee, Review on " A biomedical knowledge graph-based ...
NS-CUK Journal club: H.E.Lee, Review on " A biomedical knowledge graph-based ...NS-CUK Journal club: H.E.Lee, Review on " A biomedical knowledge graph-based ...
NS-CUK Journal club: H.E.Lee, Review on " A biomedical knowledge graph-based ...
Β 
NS-CUK Seminar: H.E.Lee, Review on "PTE: Predictive Text Embedding through L...
NS-CUK Seminar: H.E.Lee,  Review on "PTE: Predictive Text Embedding through L...NS-CUK Seminar: H.E.Lee,  Review on "PTE: Predictive Text Embedding through L...
NS-CUK Seminar: H.E.Lee, Review on "PTE: Predictive Text Embedding through L...
Β 
NS-CUK Seminar: H.B.Kim, Review on "Inductive Representation Learning on Lar...
NS-CUK Seminar: H.B.Kim,  Review on "Inductive Representation Learning on Lar...NS-CUK Seminar: H.B.Kim,  Review on "Inductive Representation Learning on Lar...
NS-CUK Seminar: H.B.Kim, Review on "Inductive Representation Learning on Lar...
Β 
NS-CUK Seminar: H.E.Lee, Review on "PTE: Predictive Text Embedding through L...
NS-CUK Seminar: H.E.Lee,  Review on "PTE: Predictive Text Embedding through L...NS-CUK Seminar: H.E.Lee,  Review on "PTE: Predictive Text Embedding through L...
NS-CUK Seminar: H.E.Lee, Review on "PTE: Predictive Text Embedding through L...
Β 
NS-CUK Seminar: J.H.Lee, Review on "Relational Self-Supervised Learning on Gr...
NS-CUK Seminar: J.H.Lee, Review on "Relational Self-Supervised Learning on Gr...NS-CUK Seminar: J.H.Lee, Review on "Relational Self-Supervised Learning on Gr...
NS-CUK Seminar: J.H.Lee, Review on "Relational Self-Supervised Learning on Gr...
Β 
NS-CUK Seminar: H.B.Kim, Review on "metapath2vec: Scalable representation le...
NS-CUK Seminar: H.B.Kim,  Review on "metapath2vec: Scalable representation le...NS-CUK Seminar: H.B.Kim,  Review on "metapath2vec: Scalable representation le...
NS-CUK Seminar: H.B.Kim, Review on "metapath2vec: Scalable representation le...
Β 
NS-CUK Seminar: H.E.Lee, Review on "Graph Star Net for Generalized Multi-Tas...
NS-CUK Seminar: H.E.Lee,  Review on "Graph Star Net for Generalized Multi-Tas...NS-CUK Seminar: H.E.Lee,  Review on "Graph Star Net for Generalized Multi-Tas...
NS-CUK Seminar: H.E.Lee, Review on "Graph Star Net for Generalized Multi-Tas...
Β 

Recently uploaded

June Patch Tuesday
June Patch TuesdayJune Patch Tuesday
June Patch Tuesday
Ivanti
Β 
Building Production Ready Search Pipelines with Spark and Milvus
Building Production Ready Search Pipelines with Spark and MilvusBuilding Production Ready Search Pipelines with Spark and Milvus
Building Production Ready Search Pipelines with Spark and Milvus
Zilliz
Β 
Your One-Stop Shop for Python Success: Top 10 US Python Development Providers
Your One-Stop Shop for Python Success: Top 10 US Python Development ProvidersYour One-Stop Shop for Python Success: Top 10 US Python Development Providers
Your One-Stop Shop for Python Success: Top 10 US Python Development Providers
akankshawande
Β 
Main news related to the CCS TSI 2023 (2023/1695)
Main news related to the CCS TSI 2023 (2023/1695)Main news related to the CCS TSI 2023 (2023/1695)
Main news related to the CCS TSI 2023 (2023/1695)
Jakub Marek
Β 
Operating System Used by Users in day-to-day life.pptx
Operating System Used by Users in day-to-day life.pptxOperating System Used by Users in day-to-day life.pptx
Operating System Used by Users in day-to-day life.pptx
Pravash Chandra Das
Β 
Azure API Management to expose backend services securely
Azure API Management to expose backend services securelyAzure API Management to expose backend services securely
Azure API Management to expose backend services securely
Dinusha Kumarasiri
Β 
Monitoring and Managing Anomaly Detection on OpenShift.pdf
Monitoring and Managing Anomaly Detection on OpenShift.pdfMonitoring and Managing Anomaly Detection on OpenShift.pdf
Monitoring and Managing Anomaly Detection on OpenShift.pdf
Tosin Akinosho
Β 
Nunit vs XUnit vs MSTest Differences Between These Unit Testing Frameworks.pdf
Nunit vs XUnit vs MSTest Differences Between These Unit Testing Frameworks.pdfNunit vs XUnit vs MSTest Differences Between These Unit Testing Frameworks.pdf
Nunit vs XUnit vs MSTest Differences Between These Unit Testing Frameworks.pdf
flufftailshop
Β 
GenAI Pilot Implementation in the organizations
GenAI Pilot Implementation in the organizationsGenAI Pilot Implementation in the organizations
GenAI Pilot Implementation in the organizations
kumardaparthi1024
Β 
5th LF Energy Power Grid Model Meet-up Slides
5th LF Energy Power Grid Model Meet-up Slides5th LF Energy Power Grid Model Meet-up Slides
5th LF Energy Power Grid Model Meet-up Slides
DanBrown980551
Β 
Introduction of Cybersecurity with OSS at Code Europe 2024
Introduction of Cybersecurity with OSS  at Code Europe 2024Introduction of Cybersecurity with OSS  at Code Europe 2024
Introduction of Cybersecurity with OSS at Code Europe 2024
Hiroshi SHIBATA
Β 
Choosing The Best AWS Service For Your Website + API.pptx
Choosing The Best AWS Service For Your Website + API.pptxChoosing The Best AWS Service For Your Website + API.pptx
Choosing The Best AWS Service For Your Website + API.pptx
Brandon Minnick, MBA
Β 
Presentation of the OECD Artificial Intelligence Review of Germany
Presentation of the OECD Artificial Intelligence Review of GermanyPresentation of the OECD Artificial Intelligence Review of Germany
Presentation of the OECD Artificial Intelligence Review of Germany
innovationoecd
Β 
System Design Case Study: Building a Scalable E-Commerce Platform - Hiike
System Design Case Study: Building a Scalable E-Commerce Platform - HiikeSystem Design Case Study: Building a Scalable E-Commerce Platform - Hiike
System Design Case Study: Building a Scalable E-Commerce Platform - Hiike
Hiike
Β 
Deep Dive: AI-Powered Marketing to Get More Leads and Customers with HyperGro...
Deep Dive: AI-Powered Marketing to Get More Leads and Customers with HyperGro...Deep Dive: AI-Powered Marketing to Get More Leads and Customers with HyperGro...
Deep Dive: AI-Powered Marketing to Get More Leads and Customers with HyperGro...
saastr
Β 
Unlock the Future of Search with MongoDB Atlas_ Vector Search Unleashed.pdf
Unlock the Future of Search with MongoDB Atlas_ Vector Search Unleashed.pdfUnlock the Future of Search with MongoDB Atlas_ Vector Search Unleashed.pdf
Unlock the Future of Search with MongoDB Atlas_ Vector Search Unleashed.pdf
Malak Abu Hammad
Β 
Serial Arm Control in Real Time Presentation
Serial Arm Control in Real Time PresentationSerial Arm Control in Real Time Presentation
Serial Arm Control in Real Time Presentation
tolgahangng
Β 
Nordic Marketo Engage User Group_June 13_ 2024.pptx
Nordic Marketo Engage User Group_June 13_ 2024.pptxNordic Marketo Engage User Group_June 13_ 2024.pptx
Nordic Marketo Engage User Group_June 13_ 2024.pptx
MichaelKnudsen27
Β 
Fueling AI with Great Data with Airbyte Webinar
Fueling AI with Great Data with Airbyte WebinarFueling AI with Great Data with Airbyte Webinar
Fueling AI with Great Data with Airbyte Webinar
Zilliz
Β 
Columbus Data & Analytics Wednesdays - June 2024
Columbus Data & Analytics Wednesdays - June 2024Columbus Data & Analytics Wednesdays - June 2024
Columbus Data & Analytics Wednesdays - June 2024
Jason Packer
Β 

Recently uploaded (20)

June Patch Tuesday
June Patch TuesdayJune Patch Tuesday
June Patch Tuesday
Β 
Building Production Ready Search Pipelines with Spark and Milvus
Building Production Ready Search Pipelines with Spark and MilvusBuilding Production Ready Search Pipelines with Spark and Milvus
Building Production Ready Search Pipelines with Spark and Milvus
Β 
Your One-Stop Shop for Python Success: Top 10 US Python Development Providers
Your One-Stop Shop for Python Success: Top 10 US Python Development ProvidersYour One-Stop Shop for Python Success: Top 10 US Python Development Providers
Your One-Stop Shop for Python Success: Top 10 US Python Development Providers
Β 
Main news related to the CCS TSI 2023 (2023/1695)
Main news related to the CCS TSI 2023 (2023/1695)Main news related to the CCS TSI 2023 (2023/1695)
Main news related to the CCS TSI 2023 (2023/1695)
Β 
Operating System Used by Users in day-to-day life.pptx
Operating System Used by Users in day-to-day life.pptxOperating System Used by Users in day-to-day life.pptx
Operating System Used by Users in day-to-day life.pptx
Β 
Azure API Management to expose backend services securely
Azure API Management to expose backend services securelyAzure API Management to expose backend services securely
Azure API Management to expose backend services securely
Β 
Monitoring and Managing Anomaly Detection on OpenShift.pdf
Monitoring and Managing Anomaly Detection on OpenShift.pdfMonitoring and Managing Anomaly Detection on OpenShift.pdf
Monitoring and Managing Anomaly Detection on OpenShift.pdf
Β 
Nunit vs XUnit vs MSTest Differences Between These Unit Testing Frameworks.pdf
Nunit vs XUnit vs MSTest Differences Between These Unit Testing Frameworks.pdfNunit vs XUnit vs MSTest Differences Between These Unit Testing Frameworks.pdf
Nunit vs XUnit vs MSTest Differences Between These Unit Testing Frameworks.pdf
Β 
GenAI Pilot Implementation in the organizations
GenAI Pilot Implementation in the organizationsGenAI Pilot Implementation in the organizations
GenAI Pilot Implementation in the organizations
Β 
5th LF Energy Power Grid Model Meet-up Slides
5th LF Energy Power Grid Model Meet-up Slides5th LF Energy Power Grid Model Meet-up Slides
5th LF Energy Power Grid Model Meet-up Slides
Β 
Introduction of Cybersecurity with OSS at Code Europe 2024
Introduction of Cybersecurity with OSS  at Code Europe 2024Introduction of Cybersecurity with OSS  at Code Europe 2024
Introduction of Cybersecurity with OSS at Code Europe 2024
Β 
Choosing The Best AWS Service For Your Website + API.pptx
Choosing The Best AWS Service For Your Website + API.pptxChoosing The Best AWS Service For Your Website + API.pptx
Choosing The Best AWS Service For Your Website + API.pptx
Β 
Presentation of the OECD Artificial Intelligence Review of Germany
Presentation of the OECD Artificial Intelligence Review of GermanyPresentation of the OECD Artificial Intelligence Review of Germany
Presentation of the OECD Artificial Intelligence Review of Germany
Β 
System Design Case Study: Building a Scalable E-Commerce Platform - Hiike
System Design Case Study: Building a Scalable E-Commerce Platform - HiikeSystem Design Case Study: Building a Scalable E-Commerce Platform - Hiike
System Design Case Study: Building a Scalable E-Commerce Platform - Hiike
Β 
Deep Dive: AI-Powered Marketing to Get More Leads and Customers with HyperGro...
Deep Dive: AI-Powered Marketing to Get More Leads and Customers with HyperGro...Deep Dive: AI-Powered Marketing to Get More Leads and Customers with HyperGro...
Deep Dive: AI-Powered Marketing to Get More Leads and Customers with HyperGro...
Β 
Unlock the Future of Search with MongoDB Atlas_ Vector Search Unleashed.pdf
Unlock the Future of Search with MongoDB Atlas_ Vector Search Unleashed.pdfUnlock the Future of Search with MongoDB Atlas_ Vector Search Unleashed.pdf
Unlock the Future of Search with MongoDB Atlas_ Vector Search Unleashed.pdf
Β 
Serial Arm Control in Real Time Presentation
Serial Arm Control in Real Time PresentationSerial Arm Control in Real Time Presentation
Serial Arm Control in Real Time Presentation
Β 
Nordic Marketo Engage User Group_June 13_ 2024.pptx
Nordic Marketo Engage User Group_June 13_ 2024.pptxNordic Marketo Engage User Group_June 13_ 2024.pptx
Nordic Marketo Engage User Group_June 13_ 2024.pptx
Β 
Fueling AI with Great Data with Airbyte Webinar
Fueling AI with Great Data with Airbyte WebinarFueling AI with Great Data with Airbyte Webinar
Fueling AI with Great Data with Airbyte Webinar
Β 
Columbus Data & Analytics Wednesdays - June 2024
Columbus Data & Analytics Wednesdays - June 2024Columbus Data & Analytics Wednesdays - June 2024
Columbus Data & Analytics Wednesdays - June 2024
Β 

NS-CUK Joint Journal Club: Minwoo Choi, Review on "Short-term wind speed forecasting based on spatial-temporal graph transformer networks", Energy 2021

  • 1. Min-Woo Choi Network Science Lab Dept. of Artificial Intelligence The Catholic University of Korea E-mail: choimin1231@catholic.ac.kr 2023. 02. 14 Publish: 2022 Journal: Energy (IF: 8.85)
  • 2. 1 οƒ˜ Introduction β€’ Limitation β€’ Purpose β€’ Main contribution οƒ˜ Methodology β€’ Data description β€’ Experiment setup οƒ˜ Results οƒ˜ Conclusions οƒ˜ Limitation and Future work
  • 3. 2 1. Introduction Limitation β€’ CNN-based spatio-temporal prediction models require wind information that nodes at regular intervals or nodes in a square array. β€’ But actual wind farms are not necessarily arranged regularly. Related study β€’ Fu et al. (2019): Spatiotemporal attention network (STAN) - Using Multi head self attention to extract the spatial correlation. β€’ Khodayar et al. (2019): Graph convolution deep learning architecture (GCDLA) - To capture the deep spatiotemporal information of wind speed. 1. Develop a relatively simple multi-node forecasting model by taking advantage of the various architectures of the above methods. 2. Proposed Spatial-Temporal Graph Transformer Network (STGTN) model to improve short- term wind speed prediction performance. Purpose of study
  • 4. 3 1. Introduction Main contribution β€’ External attention mechanism is incorporated into the forecasting model οƒΌ It can capture dynamic spatial information and reduce the complexity of the network. β€’ A transformer model with graph convolution is proposed οƒΌ To learn spatial correlation based on the Euclidean distance between wind farms.
  • 5. 4 2. Methodology Data description β€’ Location: Danish offshore wind farms β€’ Period : February 6, 2014 to June 6, 2014 β€’ Number of data: 111 wind turbine node (14,400 points) β€’ Resolution: 10 min β€’ Range of data: β€’ Training/Validation/Test set: 3:1:1 β€’ Using historical data for each 12 points, the wind speed is predicted for 10 min to 1 h.
  • 6. 5 2. Methodology Problem formulation β€’ Position information between wind turbines is represented as a graph 𝐺 = 𝑉, 𝐸, 𝐴 𝑉 = π‘›π‘œπ‘‘π‘’π‘  π‘œπ‘“ 𝑀𝑖𝑛𝑑 𝑑𝑒𝑏𝑖𝑛𝑒𝑠 𝐸 = π‘π‘œπ‘›π‘›π‘’π‘π‘‘π‘–π‘œπ‘›π‘  𝑏𝑒𝑑𝑀𝑒𝑒𝑛 π‘‘β„Žπ‘’ π‘›π‘œπ‘‘π‘’π‘  𝐴 = π‘Žπ‘‘π‘—π‘Žπ‘π‘’π‘›π‘π‘¦ π‘šπ‘Žπ‘‘π‘Ÿπ‘–π‘₯ π‘€π‘–π‘‘β„Ž πΈπ‘’π‘π‘™π‘–π‘‘π‘’π‘Žπ‘› π‘‘π‘–π‘ π‘‘π‘Žπ‘›π‘π‘’ 𝑏𝑒𝑑𝑀𝑒𝑒𝑛 π‘›π‘œπ‘‘π‘’π‘  Spatialtemporal model of Short-term wind speed forecasting Experiment setup β€’ Batch size: 64 β€’ learning rate: 0.01 β€’ Optimization: Adam β€’ Decay rate of 0.7 after every 10 epochs 𝑉 𝐸 𝐴 1 2 3 2 3 𝑑 𝑑
  • 7. 6 2. Methodology Graph transformer Structure of the graph transformer  The fusion of spatial and temporal information  it considered to improve the robustness of extracting spatial features of wind speed  higher-level features can be extracted in the spatial correlation based on Euclidean distance External attention mechanism
  • 8. 7 2. Methodology Architecture of the proposed model Aggregate spatiotemporal features Extract temporal feature Extract spatial feature Form input through residual connection Fig. 1. Illustration of the proposed spatial-temporal graph transformer network (STGTN). Main contribution
  • 9. 8 2. Methodology Benchmark models β€’ SVR β€’ DL-STF β€’ STAN β€’ STGTN-T (Transformer) β€’ STGTN (proposed) Not include spatial information Based on spatial-temporal information (previous study) Transformer and MLP
  • 10. 9 3. Results 10 min to 1 h prediction β€’ The results show that STGTN dominates all considered methods in terms of the lowest RMSE values. β€’ STGTN with MLP outperforms STGTN-T with transformer. β€’ Forecasting models based on spatiotemporal information can track changes of wind speed faster. β€’ The performance of the STGTN model is more stable compared with the DL-STF and STAN models. 0 min to 2000 min prediction Consider the all wind turbine node
  • 11. 10 3. Results May 20, performance β€’ These results show that spatial information may degrades the forecasting performance when the standard deviation of wind speed is small. May 22, performance Wind speed Condition: fluctuation is small β€’ SVR is better than DL-STF β€’ SVR is better than STGTN-T
  • 12. 11 3. Results June 4, performance May 21, performance June 4, performance Wind speed Condition: fluctuation is large β€’ The MLP and Transformer performs similarly as the feature extractor when large-scale wind speed fluctuations exist.  On the whole, STGTN provides more stable and accurate wind speed forecasts compared with other methods under different wind speed conditions.
  • 13. 12 4. Conclusions β€’ The results indicate that the proposed model can effectively utilize the spatiotemporal information to generate more accurate wind speed forecasts. β€’ Wind speed data of adjacent nodes is sufficiently used to correct the wrong predictions caused by outliers in the historical data of individual nodes. β€’ The forecasting results confirms that the proposed model can yield stable wind speed forecasts regardless of different scales of wind speed fluctuations.
  • 14. 13 5. Limitation and Future work β€’ Seasonal factors are not considered. (just consider May & June) β€’ The utilization of wind direction information is not discussed.

Editor's Notes

  1. μ•ˆλ…•ν•˜μ‹­λ‹ˆκΉŒ, μ €λŠ” κΈ°μƒν™˜κ²½ κ΄€λ ¨ 연ꡬλ₯Ό μˆ˜ν–‰ν–ˆλ˜ 졜민우라고 ν•˜κ΅¬μš”. μ €μ˜ 전곡지식을 ν† λŒ€λ‘œ μ΄μ˜€μ€€ κ΅μˆ˜λ‹˜κ³Ό μ•žμœΌλ‘œ 연ꡬλ₯Ό 진행해 λ‚˜κ°ˆ μ˜ˆμ •μž…λ‹ˆλ‹€. μ•žμœΌλ‘œ μ˜μ–΄ μ‹€λ ₯의 ν–₯상을 μœ„ν•΄ μ˜μ–΄λ‘œ λ°œν‘œν•˜κ² μœΌλ‚˜. 아직은 λŒ€λ³Έμ„ λŒ€λΆ€λΆ„ μ°Έμ‘°ν•œλ‹€λŠ” 점 μ–‘ν•΄λΆ€νƒλ“œλ¦½λ‹ˆλ‹€. The topic of my presentation is Short-term wind speed forecasting based on spatial-temporal graph transformer networks.
  2. The order of contents is as follows:
  3. First, the limitations of previous studies in wind speed prediction are as follows. When we use a spatiotemporal prediction model to predict wind speed based on CNN, a square array or nodes with regular intervals are required. Like that grid. However, in actually it is not composed of regular data. Therefore, as a related study, Fu extracted spatial correlation using the STAN model. and Kyodayar extracted spatiotemporal information of wind speed using the GCDLA model. There are few related studies, but taking advantage of related studies, this study proposed STGTN to predict wind speed.
  4. The main contribution of this paper, First~ Second~
  5. The data used were wind farms located in Danish offshore, and data were collected at 10-minute intervals with 111 wind turbine nodes. The data set was divided into 3:1:1, and the wind speed was predicted using the historical data of each 12 points.
  6. The experiment setup is as written, and the information of wind turbines is constructed as follows. V is ~ E is ~ A is ~
  7. First, this is a structure of the graph transformer model. First Attention mechanism cannot capture time series and spatial information compared with the recurrent and convolutional structure. Therefore, the spatial-temporal position in embedding layer, temporal and spatial information is injected into the input sequence before performing subsequent operations. In External attention mechanism ~ And Graph convolution extract the higher level features in spatial correlation
  8. It is a configuration of the proposed model as STGTN. first, spatial features are extracted from the Graph transformer, and second, inputs are formed through residual connections, moved to MLP, TEMPORAL features are extracted. In the last convolutional layer, the wind speed is predicted by aggregate.
  9. Benchmark models were constructed to validate the proposed model. First, SVR model not include the spatial information, DL-STF and STAN models based on spatial-temporal information and the STGTN-T model using Transformer was used instead of MLP model.
  10. Next is the results part. Looking at the table, the proposed model performed better than the benchmarking models. What is noteworthy here is that STGTN showed better performance than STGTN-T, which means that the MLP used instead of the transformer improved the prediction accuracy, It has the advantage of reducing the overhead of model training and hyperparameter tuning with a simple structure.
  11. Next, when the days of wind speed fluctuation is small, look at the upper table SVR is better than DL-STF and under table SVR is better than STGTN-T. These results show that spatial information may degrades the forecasting performance when the Flutuation is small.
  12. Conversely, STGTN-T and STGTN were found to be effective when the fluctuations is large. The proposed models all showed good predictive performance regardless of wind speed conditions.
  13. The proposed model is stable and has excellent spatio-temporal predictability.
  14. Anyone have any questions?