Short-term regional wind power forecasting for small datasets with input data correction, hybrid neural network, and error analysis

The security of power systems and electrical grids can be affected by the stochastic nature of wind energy. Therefore, reliable techniques for load forecasting and planning must be developed. This paper presents a model for short-term regional wind power forecasting based on small datasets. The mode...

Descripción completa

Guardado en:

Detalles Bibliográficos
Autores principales:	Weichao Dong, Hexu Sun, Jianxin Tan, Zheng Li, Jingxuan Zhang, Yu Yang Zhao
Formato:	article
Lenguaje:	EN
Publicado:	Elsevier 2021
Materias:	K-means clustering Long short-term memory Multiscale mathematical morphological decomposition Short-term regional wind power forecasting Small dataset Stacked multilevel-denoising autoencoder Electrical engineering. Electronics. Nuclear engineering TK1-9971
Acceso en línea:	https://doaj.org/article/ecea42a3e20349f1a36dabc225abd7f2
Etiquetas:	Agregar Etiqueta Sin Etiquetas, Sea el primero en etiquetar este registro!

id	oai:doaj.org-article:ecea42a3e20349f1a36dabc225abd7f2
record_format	dspace
spelling	oai:doaj.org-article:ecea42a3e20349f1a36dabc225abd7f22021-11-24T04:32:12ZShort-term regional wind power forecasting for small datasets with input data correction, hybrid neural network, and error analysis2352-484710.1016/j.egyr.2021.11.021https://doaj.org/article/ecea42a3e20349f1a36dabc225abd7f22021-11-01T00:00:00Zhttp://www.sciencedirect.com/science/article/pii/S2352484721011665https://doaj.org/toc/2352-4847The security of power systems and electrical grids can be affected by the stochastic nature of wind energy. Therefore, reliable techniques for load forecasting and planning must be developed. This paper presents a model for short-term regional wind power forecasting based on small datasets. The model comprises three steps: input data correction, a hybrid neural network, and error analysis. First, regional numerical weather predictions (NWP) are corrected by using a stacked multilevel-denoising autoencoder (SMLDAE) to generate more effective inputs; this is the first study to use SMLDAE for NWP data correction. Second, a neural network-based hybrid model is employed for regional wind power forecasting to predict the wind power in the region. The proposed hybrid model employs three processes: multiscale mathematical morphological decomposition (MMMD), k-means clustering, and a stacked denoising autoencoder. MMMD can decompose the data directly in the time domain, thus, the signal does not need to be transferred from the time domain to the frequency domain to accomplish the decomposition. Third, a long short-term memory network is used for error analysis of the preliminary forecasted data. The preliminary results and error series are aggregated to generate the final forecasting result. For small datasets, we use multi-distribution mega-trend diffusion to augment the dataset. The proposed model was validated using a dataset consisting of data generated by regional wind farms in northern China. The results show that the proposed model enables wind forecasting at both the regional and single-farm level. Moreover, whereas most benchmark models require almost one year of data, the model requires only approximately three months of NWP data to produce reliable forecasting within the next 24 h.Weichao DongHexu SunJianxin TanZheng LiJingxuan ZhangYu Yang ZhaoElsevierarticleK-means clusteringLong short-term memoryMultiscale mathematical morphological decompositionShort-term regional wind power forecastingSmall datasetStacked multilevel-denoising autoencoderElectrical engineering. Electronics. Nuclear engineeringTK1-9971ENEnergy Reports, Vol 7, Iss , Pp 7675-7692 (2021)
institution	DOAJ
collection	DOAJ
language	EN
topic	K-means clustering Long short-term memory Multiscale mathematical morphological decomposition Short-term regional wind power forecasting Small dataset Stacked multilevel-denoising autoencoder Electrical engineering. Electronics. Nuclear engineering TK1-9971
spellingShingle	K-means clustering Long short-term memory Multiscale mathematical morphological decomposition Short-term regional wind power forecasting Small dataset Stacked multilevel-denoising autoencoder Electrical engineering. Electronics. Nuclear engineering TK1-9971 Weichao Dong Hexu Sun Jianxin Tan Zheng Li Jingxuan Zhang Yu Yang Zhao Short-term regional wind power forecasting for small datasets with input data correction, hybrid neural network, and error analysis
description	The security of power systems and electrical grids can be affected by the stochastic nature of wind energy. Therefore, reliable techniques for load forecasting and planning must be developed. This paper presents a model for short-term regional wind power forecasting based on small datasets. The model comprises three steps: input data correction, a hybrid neural network, and error analysis. First, regional numerical weather predictions (NWP) are corrected by using a stacked multilevel-denoising autoencoder (SMLDAE) to generate more effective inputs; this is the first study to use SMLDAE for NWP data correction. Second, a neural network-based hybrid model is employed for regional wind power forecasting to predict the wind power in the region. The proposed hybrid model employs three processes: multiscale mathematical morphological decomposition (MMMD), k-means clustering, and a stacked denoising autoencoder. MMMD can decompose the data directly in the time domain, thus, the signal does not need to be transferred from the time domain to the frequency domain to accomplish the decomposition. Third, a long short-term memory network is used for error analysis of the preliminary forecasted data. The preliminary results and error series are aggregated to generate the final forecasting result. For small datasets, we use multi-distribution mega-trend diffusion to augment the dataset. The proposed model was validated using a dataset consisting of data generated by regional wind farms in northern China. The results show that the proposed model enables wind forecasting at both the regional and single-farm level. Moreover, whereas most benchmark models require almost one year of data, the model requires only approximately three months of NWP data to produce reliable forecasting within the next 24 h.
format	article
author	Weichao Dong Hexu Sun Jianxin Tan Zheng Li Jingxuan Zhang Yu Yang Zhao
author_facet	Weichao Dong Hexu Sun Jianxin Tan Zheng Li Jingxuan Zhang Yu Yang Zhao
author_sort	Weichao Dong
title	Short-term regional wind power forecasting for small datasets with input data correction, hybrid neural network, and error analysis
title_short	Short-term regional wind power forecasting for small datasets with input data correction, hybrid neural network, and error analysis
title_full	Short-term regional wind power forecasting for small datasets with input data correction, hybrid neural network, and error analysis
title_fullStr	Short-term regional wind power forecasting for small datasets with input data correction, hybrid neural network, and error analysis
title_full_unstemmed	Short-term regional wind power forecasting for small datasets with input data correction, hybrid neural network, and error analysis
title_sort	short-term regional wind power forecasting for small datasets with input data correction, hybrid neural network, and error analysis
publisher	Elsevier
publishDate	2021
url	https://doaj.org/article/ecea42a3e20349f1a36dabc225abd7f2
work_keys_str_mv	AT weichaodong shorttermregionalwindpowerforecastingforsmalldatasetswithinputdatacorrectionhybridneuralnetworkanderroranalysis AT hexusun shorttermregionalwindpowerforecastingforsmalldatasetswithinputdatacorrectionhybridneuralnetworkanderroranalysis AT jianxintan shorttermregionalwindpowerforecastingforsmalldatasetswithinputdatacorrectionhybridneuralnetworkanderroranalysis AT zhengli shorttermregionalwindpowerforecastingforsmalldatasetswithinputdatacorrectionhybridneuralnetworkanderroranalysis AT jingxuanzhang shorttermregionalwindpowerforecastingforsmalldatasetswithinputdatacorrectionhybridneuralnetworkanderroranalysis AT yuyangzhao shorttermregionalwindpowerforecastingforsmalldatasetswithinputdatacorrectionhybridneuralnetworkanderroranalysis
_version_	1718416009956687872

Short-term regional wind power forecasting for small datasets with input data correction, hybrid neural network, and error analysis

Ejemplares similares