Multi-step wind speed forecasting based on a hybrid forecasting   architecture and an improved bat algorithm

doi:10.1016/j.enconman.2017.04.012

CORC > 北京大学 > 工学院

	Multi-step wind speed forecasting based on a hybrid forecasting architecture and an improved bat algorithm
	Xiao, Liye ; Qian, Feng ; Shao, Wei
刊名	ENERGY CONVERSION AND MANAGEMENT
	2017
关键词	Hybrid forecasting architecture Improved bat algorithm Singular spectrum analysis, wind speed forecasting ARTIFICIAL NEURAL-NETWORKS EMPIRICAL MODE DECOMPOSITION SUPPORT VECTOR MACHINES MULTIOBJECTIVE OPTIMIZATION INSPIRED ALGORITHM ELECTRICAL-POWER WAVELET PACKET TIME-SERIES PREDICTION CLASSIFICATION
DOI	10.1016/j.enconman.2017.04.012
英文摘要	As one of the most promising sustainable energy sources, wind energy plays an important role in energy development because of its cleanliness without causing pollution. Generally, wind speed forecasting, which has an essential influence on wind power systems, is regarded as a challenging task. Analyses based on single-step wind speed forecasting have been widely used, but their results are insufficient in ensuring the reliability and controllability of wind power systems. In this paper, a new forecasting architecture based on decomposing algorithms and modified neural networks is successfully developed for multi-step wind speed forecasting. Four different hybrid models are contained in this architecture, and to further improve the forecasting performance, a modified bat algorithm (BA) with the conjugate gradient (CG) method is developed to optimize the initial weights between layers and thresholds of the hidden layer of neural networks. To investigate the forecasting abilities of the four models, the wind speed data collected from four different wind power stations in Penglai, China, were used as a case study. The numerical experiments showed that the hybrid model including the singular spectrum analysis and general regression neural network with CG-BA (SSA-CG-BA-GRNN) achieved the most accurate forecasting results in one-step to three-step wind speed forecasting. (C) 2017 Elsevier Ltd. All rights reserved.; National Natural Science Foundation of China [61331007, 61471105]; 973 Project [613273]; SCI(E); ARTICLE; 410-430; 143
语种	英语
内容类型	期刊论文
源URL	[http://ir.pku.edu.cn/handle/20.500.11897/472577]
专题	工学院
推荐引用方式 GB/T 7714	Xiao, Liye,Qian, Feng,Shao, Wei. Multi-step wind speed forecasting based on a hybrid forecasting architecture and an improved bat algorithm[J]. ENERGY CONVERSION AND MANAGEMENT,2017.
APA	Xiao, Liye,Qian, Feng,&Shao, Wei.(2017).Multi-step wind speed forecasting based on a hybrid forecasting architecture and an improved bat algorithm.ENERGY CONVERSION AND MANAGEMENT.
MLA	Xiao, Liye,et al."Multi-step wind speed forecasting based on a hybrid forecasting architecture and an improved bat algorithm".ENERGY CONVERSION AND MANAGEMENT (2017).