TY - GEN
T1 - Comparison of recurrent neural networks for wind power forecasting
AU - López, Erick
AU - Valle, Carlos
AU - Allende-Cid, Héctor
AU - Allende, Héctor
N1 - Publisher Copyright:
© Springer Nature Switzerland AG 2020.
Copyright:
Copyright 2020 Elsevier B.V., All rights reserved.
PY - 2020
Y1 - 2020
N2 - Integrating wind power to the electrical grid is complicated due to the stochastic nature of the wind, which makes its prediction a challenging task. Then, it is important to devise forecasting tools to support this task. For example, a network that integrates an Echo State Network architecture and Long Short-Term Memory blocks as hidden units (ESN+LSTM) has been proposed, showing good performance against a physical model. This paper proposes to compare this network versus Echo State Network (ESN) and Long Short-Term Memory (LSTM), to forecast wind power from 1 to 24 h ahead. Results show than the ESN+LSTM model outperforms the performance reached for ESN and LSTM, in terms of MSE, MAE, and the metrics used in the Taylor diagram. In addition, we observe that the advantage of this network is statistically significant during the first moments of the forecast horizon, in terms of T-test and Wilcoxon-test.
AB - Integrating wind power to the electrical grid is complicated due to the stochastic nature of the wind, which makes its prediction a challenging task. Then, it is important to devise forecasting tools to support this task. For example, a network that integrates an Echo State Network architecture and Long Short-Term Memory blocks as hidden units (ESN+LSTM) has been proposed, showing good performance against a physical model. This paper proposes to compare this network versus Echo State Network (ESN) and Long Short-Term Memory (LSTM), to forecast wind power from 1 to 24 h ahead. Results show than the ESN+LSTM model outperforms the performance reached for ESN and LSTM, in terms of MSE, MAE, and the metrics used in the Taylor diagram. In addition, we observe that the advantage of this network is statistically significant during the first moments of the forecast horizon, in terms of T-test and Wilcoxon-test.
KW - Echo State Network
KW - Long Short-Term Memory
KW - Multivariate time series
KW - Recurrent Neural Networks
KW - Wind power forecasting
UR - http://www.scopus.com/inward/record.url?scp=85087274753&partnerID=8YFLogxK
U2 - 10.1007/978-3-030-49076-8_3
DO - 10.1007/978-3-030-49076-8_3
M3 - Conference contribution
AN - SCOPUS:85087274753
SN - 9783030490751
T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
SP - 25
EP - 34
BT - Pattern Recognition - 12th Mexican Conference, MCPR 2020, Proceedings
A2 - Figueroa Mora, Karina Mariela
A2 - Anzurez Marín, Juan
A2 - Cerda, Jaime
A2 - Carrasco-Ochoa, Jesús Ariel
A2 - Martínez-Trinidad, José Francisco
A2 - Olvera-López, José Arturo
PB - Springer
T2 - 12th Mexican Conference on Pattern Recognition, MCPR 2020
Y2 - 24 June 2020 through 27 June 2020
ER -