摘要
在电力系统中风电装机容量增长的背景下,高精度的超短期风功率预测是保证系统可靠运行的重要基础。为此,提出一种以复数据经验模态分解的噪声辅助信号分解法(NACEMD)和Elman神经网络为基础的超短期风功率组合预测方法。在风功率序列中添加白噪声,使用NACEMD将其按照不同波动尺度逐级分解,得到不同时频特性的分量,然后利用Elman神经网络对各分量建立预测模型,以各分量的不同时频特性为基准对预测结果进行叠加,得到风功率预测值。实例分析表明,提出的组合预测法既可进一步减轻现有方法中存在的模态混叠现象,具备较高的预测精度。研究成果可为风功率预测提供参考。
With the occupancy of wind power generation increasing in power system, the high-accuracy of ultra-shortterm wind power prediction was the foundation of maintaining the power system operating stably. This paper presented a combination method based on the noise assisted complex empirical mode decomposition （NACEMD） and Elman neural network to forecast the ultra-short-term wind power. The white noise series were adding in the raw wind power sequence to set up complex data series. Then the wind power series were decomposed by the NACEMD according to the different fluctuation scale, and a set of relatively stable components that had different time-frequency characteristics were obtained. All of the components were used to establish the Elman neural network prediction model. Finally, according to the differ- ent time-frequency characteristics of the components, the superposition of the predicted results of each component was taken as the ultimate forecasting value. The practical calculation example shows that this method can not only reduce the mode mixing issues in common method, but also has high prediction accuracy. The research results can provide the refer- ence for wind power prediction.
作者
杨楠
叶迪
周峥
鄢晶
黄禹
董邦天
YANG Nan1 ,YE Di1 ,ZHOU Zheng1 ,YAN Jing2 , HUANG Yu1 ,DONG Bang-tian1(1. Hubei Provincial Collaborative Center for New Energy Microgrid, Yichang 443002, China; 2. Economic ＆ Technology Research Institute, State Grid Hubei Electric Power Company, Wuhan 430077, Chin)
出处
《水电能源科学》
北大核心
2018年第9期209-211,171,共4页
Water Resources and Power
基金
国家自然科学基金项目(51607104)
三峡大学学位论文培优基金项目(2018SSPY078)
