摘要
提出一种基于经验模式分解(EMD)、加权马尔可夫链与分位数回归(quantile regression,QR)的风电功率概率区间预测方法。由于风功率数据与风速显著相关,首先对历史风速进行经验模式分解,得到不同频率段的风速,再以不同频率段的风速为样本,分别对其进行加权马尔可夫链预测,相加得到最终预测风速。最后将所得的预测风速代入QR预测模型,得到一定置信水平下的风电功率概率区间的上下限。以区间覆盖率和区间平均带宽为评价指标,与马尔可夫链下的QR法和加权马尔可夫链下的QR法的对比仿真表明,提出的基于经验模式分解与加权马尔可夫链下的QR法具有风电功率概率预区间预测的覆盖率更高,平均带宽更窄,精度更好的预测效果。
The study of uncertainties in wind power is critical to power system planning and operational decision making. A method of wind power probability interval prediction based on empirical mode decomposition(EMD),weighted Markov chain and quantile regression(QR)is proposed. Because the wind power data is significantly related to the wind speed,the empirical mode decomposition method is applied to historical wind speed. Then the wind speed of different frequency segments is obtained. We take the wind speed of different frequency segments as a sample,and the weighted Markov chain prediction is carried out respectively. Add the results to get the final predicted wind speed. Finally,the predicted wind speed is applied to the QR prediction model to obtain the upper and lower limits of the wind power probability interval at a certain confidence level. The interval coverage and interval average bandwidth were used as evaluation indexes. Compared with the QR method under the Markov chain and the QR method under the weighted Markov chain,it is shown that the proposed QR method based on the EMD and the weighted Markov chain has the higher prediction interval coverage,the narrower average bandwidth,and the better accuracy.
作者
杨锡运
马雪
张洋
张璜
耿娜
Yang Xiyun;Ma Xue;Zhang Yang;Zhang Huang;Geng Na(School of Control and Computer Engineering,North China Electric Power University,Beijing 102206,China;Jilin Electric Power Research Institute,Changchun 130021,China)
出处
《太阳能学报》
EI
CAS
CSCD
北大核心
2020年第2期66-72,共7页
Acta Energiae Solaris Sinica
基金
国家自然科学基金(51677067)
中央高校基本科研业务费专项资金(2015MS32)。
