Wind Speed Forecasting Method Using EEMD and the Combination Forecasting Method Based on GPR and LSTM
Short-term wind speed prediction is of cardinal significance for maximization of wind power utilization. However, the strong intermittency and volatility of wind speed pose a challenge to the wind speed prediction model. To improve the accuracy of wind speed prediction, a novel model using the ensem...
Main Authors: | , , |
---|---|
Format: | Article |
Language: | English |
Published: |
MDPI AG
2018-10-01
|
Series: | Sustainability |
Subjects: | |
Online Access: | http://www.mdpi.com/2071-1050/10/10/3693 |
id |
doaj-224a3fd142b34fe79a548e0faace7c8c |
---|---|
record_format |
Article |
spelling |
doaj-224a3fd142b34fe79a548e0faace7c8c2020-11-25T02:29:16ZengMDPI AGSustainability2071-10502018-10-011010369310.3390/su10103693su10103693Wind Speed Forecasting Method Using EEMD and the Combination Forecasting Method Based on GPR and LSTMYuansheng Huang0Shijian Liu1Lei Yang2Department of Economics and Management, North China Electric Power University, Baoding 071003, ChinaDepartment of Economics and Management, North China Electric Power University, Baoding 071003, ChinaDepartment of Economics and Management, North China Electric Power University, Baoding 071003, ChinaShort-term wind speed prediction is of cardinal significance for maximization of wind power utilization. However, the strong intermittency and volatility of wind speed pose a challenge to the wind speed prediction model. To improve the accuracy of wind speed prediction, a novel model using the ensemble empirical mode decomposition (EEMD) method and the combination forecasting method for Gaussian process regression (GPR) and the long short-term memory (LSTM) neural network based on the variance-covariance method is proposed. In the proposed model, the EEMD method is employed to decompose the original data of wind speed series into several intrinsic mode functions (IMFs). Then, the LSTM neural network and the GPR method are utilized to predict the IMFs, respectively. Lastly, based on the IMFs’ prediction results with the two forecasting methods, the variance-covariance method can determine the weight of the two forecasting methods and offer a combination forecasting result. The experimental results from two forecasting cases in Zhangjiakou, China, indicate that the proposed approach outperforms other compared wind speed forecasting methods.http://www.mdpi.com/2071-1050/10/10/3693short-term wind speed forecastingensemble empirical mode decompositionthe combination forecasting methodlong short-term memory neural networkGaussian process regression |
collection |
DOAJ |
language |
English |
format |
Article |
sources |
DOAJ |
author |
Yuansheng Huang Shijian Liu Lei Yang |
spellingShingle |
Yuansheng Huang Shijian Liu Lei Yang Wind Speed Forecasting Method Using EEMD and the Combination Forecasting Method Based on GPR and LSTM Sustainability short-term wind speed forecasting ensemble empirical mode decomposition the combination forecasting method long short-term memory neural network Gaussian process regression |
author_facet |
Yuansheng Huang Shijian Liu Lei Yang |
author_sort |
Yuansheng Huang |
title |
Wind Speed Forecasting Method Using EEMD and the Combination Forecasting Method Based on GPR and LSTM |
title_short |
Wind Speed Forecasting Method Using EEMD and the Combination Forecasting Method Based on GPR and LSTM |
title_full |
Wind Speed Forecasting Method Using EEMD and the Combination Forecasting Method Based on GPR and LSTM |
title_fullStr |
Wind Speed Forecasting Method Using EEMD and the Combination Forecasting Method Based on GPR and LSTM |
title_full_unstemmed |
Wind Speed Forecasting Method Using EEMD and the Combination Forecasting Method Based on GPR and LSTM |
title_sort |
wind speed forecasting method using eemd and the combination forecasting method based on gpr and lstm |
publisher |
MDPI AG |
series |
Sustainability |
issn |
2071-1050 |
publishDate |
2018-10-01 |
description |
Short-term wind speed prediction is of cardinal significance for maximization of wind power utilization. However, the strong intermittency and volatility of wind speed pose a challenge to the wind speed prediction model. To improve the accuracy of wind speed prediction, a novel model using the ensemble empirical mode decomposition (EEMD) method and the combination forecasting method for Gaussian process regression (GPR) and the long short-term memory (LSTM) neural network based on the variance-covariance method is proposed. In the proposed model, the EEMD method is employed to decompose the original data of wind speed series into several intrinsic mode functions (IMFs). Then, the LSTM neural network and the GPR method are utilized to predict the IMFs, respectively. Lastly, based on the IMFs’ prediction results with the two forecasting methods, the variance-covariance method can determine the weight of the two forecasting methods and offer a combination forecasting result. The experimental results from two forecasting cases in Zhangjiakou, China, indicate that the proposed approach outperforms other compared wind speed forecasting methods. |
topic |
short-term wind speed forecasting ensemble empirical mode decomposition the combination forecasting method long short-term memory neural network Gaussian process regression |
url |
http://www.mdpi.com/2071-1050/10/10/3693 |
work_keys_str_mv |
AT yuanshenghuang windspeedforecastingmethodusingeemdandthecombinationforecastingmethodbasedongprandlstm AT shijianliu windspeedforecastingmethodusingeemdandthecombinationforecastingmethodbasedongprandlstm AT leiyang windspeedforecastingmethodusingeemdandthecombinationforecastingmethodbasedongprandlstm |
_version_ |
1724834089899393024 |