Wind power forecasting based on singular spectrum analysis and a new hybrid Laguerre neural network. (1st February 2020)

• Record Type: Journal Article
• Authors: Wang, Cong; Zhang, Hongli; Ma, Ping
• Is Part Of: Applied energy. Volume 259(2020)
• Publication Date: 2020-02-01
• Subjects: Wind power prediction -- Hybrid Laguerre neural network -- Singular spectrum analysis -- Opposition transition state transition algorithm; Power (Mechanics) -- Periodicals; Energy conservation -- Periodicals; Energy conversion -- Periodicals; 621.042
• DOI: 10.1016/j.apenergy.2019.114139
• Languages: English
• ISSNs: 0306-2619

Abstract

Highlights: SAA and data conversion are used to analyze the wind power series. A new Laguerre orthogonal basis function in (−∞, 0] is proposed and verified. A innovate hybrid model is proposed to forecast multi-step ahead wind power. The OTSTA can increase the prediction accuracy through optimized model parameters. Abstract: Given the intermittency and randomness of wind energy, the mass grid connection of wind power poses great challenges in power system and increases the threat in power system balance. Wind power forecasting can predict the fluctuation of output wind power in wind farms, which can effectively reduce wind power uncertainty. Improving the accuracy of wind power is indispensable for enhancing the efficiency of wind power utilization. To improve the forecasting accuracy, this research proposed a novel wind power forecasting method based on singular spectrum analysis and a new hybrid Laguerre neural network. First, singular spectrum analysis was used to analyze the wind power series, which decomposes the series into two subsequences, namely, trend and harmonic series and noise series. Then, Laguerre neural network and new Laguerre neural network were proposed to build the hybrid forecasting model optimized by the opposition transition state transition algorithm. The two decomposed signals were used for forecasting the future wind power value by using a forecasting model. Finally, the proposed hybrid forecasting method was investigated with respect to the wind farm in Xinjiang, China. Prediction performance results demonstrated that the proposed model has higher accuracy than the Laguerre neural network, hybrid Laguerre neural network, hybrid Laguerre neural network with singular spectrum analysis, hybrid Laguerre neural network with opposition transition state transition algorithm and singular spectrum analysis, and other popular methods.