Application and Contrast Analysis of BP and RBF Neural Network in Short-Term Wind Power Prediction

Abstract:

Wind power prediction is very important to maintain the power balance and economic operation of power system. The BP and RBF neural network were respectively used to predict one wind turbines’ output power, in 4 hours, on a wind farm in Shandong Province. The results show that the BP model, with 6-13-1 net structure and considering the meteorological factors, exhibits the best prediction accuracy (MAPE is 3.59%, NRMSE is 1.58%). The most important factor in the meteorological information for power prediction is temperature, followed by air pressure, relative humidity finally. BP model is slightly better than RBF model, but the latter is much better in the learning speed and stability. Dynamic-BP neural network, combined with the dynamical weight adjustment method, is better than BP neural network in solving the weight problem. These methods are feasible to the wind power prediction.

• Periodical: Applied Mechanics and Materials (Volume 492)
• Pages: 544-549
• DOI: https://doi.org/10.4028/www.scientific.net/AMM.492.544
• Online since: January 2014
• Authors: Wen Hua Li, Qian Xiao, Jin Long Liu, Hui Qiao Liu
• Keywords: BP Neural Network (BPNN), RBF Neural Network, Wind Power Prediction

References

[1] Han Shuang. Research on short-team prediction methods of wind power in wind farm[D]. Beijing: North China Electric Power University, 2008: 14-17.

[2] Ma Lei, Luan Shiyan. A review on the forecasting of wind speed and generated power[J]. Renewable and Sustainable Energy Reviews, 2009(13): 915-920.

DOI: 10.1016/j.rser.2008.02.002

[3] Fan Gaofeng, Wang Weisheng, Liu Chun, et al. Wind power prediction based on artificial neural network[J]. Proceedings of the CSEE, 2008, 28(34): 118-123.

[4] Meng Yangyang, Lu Jiping, Sun Huali, et al. short-team prediction of wind power based on similar days and artificial neural network [J]. Power System Technology, 2010, 34(12): 163-167.

[5] WenXin, ZhouLu, Danli, et al. Application of neural network based on Matlab[M]. Beijing: SciencePress, (2001).

[6] Du Ying, Lu Jiping, Li Qing, et al. short-team prediction of wind speed in wind farm based on least squares support vector machine [J]. Power System Technology, 2008, 32(15): 62-66.

[7] Peng Huaiwu, Liu Fangrui, Yang Xiaofeng. Research on short-team prediction of wind power based on artificial neural network [J]. East China Electric Power, 2009, 37(11): 62-66.

[8] Yang Xiuyuan, Xiao Yang, Chen Shuyong. Wind speed and generated power forecasting in wind farm[J]. Proceedings of the CSEE, 2005, 25(11): 1-5.

[9] Zhang Yao, Xi Yunhua, Hu Jinlei. Medium and long-term load forecasting method based on PCA and RBF network [J]. Electrotechnical Application, 2008, 21(2): 61-64.