| 基于波动特性挖掘的短期光伏功率预测 |
| |
| 引用本文： | 吉锌格,李慧,叶林,王丽婕. 基于波动特性挖掘的短期光伏功率预测[J]. 太阳能学报, 2022, 43(5): 146-155. DOI: 10.19912/j.0254-0096.tynxb.2020-0961 |
| |
| 作者姓名： | 吉锌格 李慧 叶林 王丽婕 |
| |
| 作者单位： | 1.北京信息科技大学自动化学院,北京 100192; 2.中国农业大学信息与电气工程学院,北京 100083 |
| |
| 基金项目： | 国家重点研发计划（2018YFB0904200）; 国家自然科学基金（51607009）; 北京市自然科学基金（3172015）; 国家电网有限公司配套科技项目（SGLNDKOOKJJS1800266） |
| |
| 摘 要： | 综合考虑光伏功率受气象因素影响所呈现出的规律性和波动性,对光伏功率波动类型进行划分与聚类识别提出一种基于波动特性挖掘的短期光伏功率预测方法,。在此基础上,利用数值天气预报和基于互信息熵的相关性分析法提取各类功率波动对应的天气波动特征及其强相关气象因子,建立基于波动特性挖掘的长短期记忆网络组合预测模型,挖掘天气波动与光伏功率波动之间的潜在映射规律。最后,识别出待测日天气波动类型与预测模型之间的匹配关系,利用组合预测模型实现光伏功率预测。通过对中国西北地区某光伏电站的预测分析,验证了所提预测方法的有效性。
|
| 关 键 词： | 光伏发电 功率预测 数据挖掘 波动 浓度学习 信息熵 |
| 收稿时间： | 2020-09-10 |
SHORT-TERM PHOTOVOLTAIC POWER FORECASTING BASED ON FLUCTUATION CHARACTERISTICS MINING |
| |
| Ji Xin,#x,ge,Li Hui,Ye Lin,Wang Lijie. SHORT-TERM PHOTOVOLTAIC POWER FORECASTING BASED ON FLUCTUATION CHARACTERISTICS MINING[J]. Acta Energiae Solaris Sinica, 2022, 43(5): 146-155. DOI: 10.19912/j.0254-0096.tynxb.2020-0961 |
| |
| Authors: | Ji Xin&#x ge Li Hui Ye Lin Wang Lijie |
| |
| Affiliation: | 1. College of Automation, Beijing Information Science and Technology University, Beijing 100192, China; 2. College of Information and Electrical Engineering, China Agricultural University, Beijing 100083, China |
| |
| Abstract: | A short-term photovoltaic power forecasting method based on fluctuation characteristics mining is proposed in this paper. Firstly, the classification method and cluster identification method of photovoltaic power fluctuation are presented, considering the regularity and volatility of photovoltaic power affected by meteorological factors. Secondly,the Numerical Weather Prediction and the correlation analysis based on mutual information entropy are used to extract the weather fluctuation characteristics and highly correlated meteorological factors corresponding to various power fluctuations. Thirdly,the combined model of the long-short term memory network is put forward to mine the potential mapping relationship between the weather fluctuation and photovoltaic power fluctuation. Finally,after the types of weather fluctuations on the tested day are identified,its photovoltaic powers are predicted by using the combined method. The results of a photovoltaic power station in Northwest China show that the proposed model is effective. |
| |
| Keywords: | photovoltaic power generation power forecasting data mining fluctuations deep learning information entropy |
|
| 点击此处可从《太阳能学报》浏览原始摘要信息 |
|
点击此处可从《太阳能学报》下载免费的PDF全文 |
|
