期刊文献+

基于EWT-ESN的短期风电功率预测研究 被引量:12

SHORT-TERM WIND POWER PREDICTION BASED ON EWT-ESN
原文传递
导出
摘要 针对风电功率具有强非线性的特点,提出一种经验小波变换(EWT)和回声状态网络(ESN)的短期风电功率组合预测方法。EWT吸取了经验模态分解(EMD)和小波分析各自的优点,核心思想是根据信号中包含的频谱信息建立基于傅里叶支持的小波滤波器组。首先,通过提取频域极大值点对信号的Fourier谱进行自适应的划分,然后根据信号频谱特性构造正交小波滤波器组来提取信号中包含的具有紧支撑Fourier频谱特性的调幅.调频(AM-FM)分量。EwT是在小波框架下建立的自适应信号分解方法,相比于EMD,其具有理论性强,计算量小,分解的模态数量少的优点。该文采用EWT将原始风电功率序列分解为具有特征差异的的不同分量,采用ESN对各分量分别预测并叠加来得到最终的预测结果;最后,将EWT-ESN方法应用在国内外两个短期风电功率实例中,实验结果表明,该文方法可有效提高风电功率预测的精度。 According to the characteristics of strong nonlinear exist in wind power, a kind of combined short-term wind power forecasting method which based on empirical wavelet transform (EWT)and echo state network (ESN)is proposed. EWT inherits the advantages of empirical mode decomposition (EMD) and wavelet transform, the key idea is to build a wavelet filter bank based on Fourier supports detected from the information contained in the processed signal spectrum. Firstly, the Fourier spectrum was segmented adaptively by extracting the maxima point in the frequency domain, and then the orthogonal wa;~elet filter banks was constructed according to the signal spectrum characteristics so as to extract different AM-FM component with compact support Fourier spectrum. As an adaptive signal decomposition method, EWT is a self-adaptive analysis approach which was established under the wavelet framework, compared with the EMD, EWT takes advantages of strong theoretical, small amount of computations and less decomposed modes. The original wind power sequence is decomposed into different components with different the characteristics by using EWT, the ESN is adopted to forecast each component and the ultimate forecasting results can be obtained by the superposition of the corresponding forecasting model. Finally, the proposed method is applied to the instance of short-term wind power forecasting at home and abroad, experiment results confirm that the proposed method can effectively improve the accuracy of wind power prediction.
作者 王新友 李青 郑少鹏 Wang Xinyou1, Li Qing2, Zheng Shaopeng2(1. College of Science And Technology, Gansu Radio & TV University, Lanzhou 730030, China ; 2. Grid Technology Center, Xinjiang Electric Power Research Institute of the State Grid, Urumqi 830011, Chin)
出处 《太阳能学报》 EI CAS CSCD 北大核心 2018年第3期633-642,共10页 Acta Energiae Solaris Sinica
基金 甘肃广播电视大学科研基金(2014-ZD-01)
关键词 经验小波变换 预测 小波分析 回声状态网络 风电功率 empirical wavelet transform forecasting wavelet analysis echo state network wind power
  • 相关文献

参考文献5

二级参考文献106

共引文献663

同被引文献175

引证文献12

二级引证文献127

相关作者

内容加载中请稍等...

相关机构

内容加载中请稍等...

相关主题

内容加载中请稍等...

浏览历史

内容加载中请稍等...
;
使用帮助 返回顶部