引用本文:梅飞,顾佳琪,裴鑫,郑建勇.基于自适应滚动匹配预测修正模式的光伏区间预测[J].电力自动化设备,2022,42(2):
MEI Fei,GU Jiaqi,PEI Xin,ZHENG Jianyong.Photovoltaic interval prediction based on adaptive rolling matching prediction correction mode[J].Electric Power Automation Equipment,2022,42(2):
【打印本页】   【HTML】   【下载PDF全文】   查看/发表评论  【EndNote】   【RefMan】   【BibTex】
←前一篇|后一篇→ 过刊浏览    高级检索
本文已被:浏览 9250次   下载 1747  
基于自适应滚动匹配预测修正模式的光伏区间预测
梅飞1, 顾佳琪1, 裴鑫1, 郑建勇2
1.河海大学 能源与电气学院,江苏 南京 211100;2.东南大学 电气工程学院,江苏 南京 210096
摘要:
为解决传统点预测方式难以量化光伏发电功率不确定性的问题以及提高预测精度,提出一种基于自适应滚动匹配预测修正模式的光伏区间预测方法。通过结合小波能量的谱聚类方法对历史光伏数据集进行聚类,构建不同类别的模型输入和区间输出并采用宽度学习系统进行训练;建立不同类别、不同置信区间、不同预测功率区间的stable误差分布,并结合优化目标函数找出每个预测功率区间的最优修正分位数点数值;利用滚动匹配预测修正模式进行区间预测。我国无锡某地的2.8 MW光伏电站算例结果表明,所提方法相较于传统的聚类预测方法具有更好的预测效果。
关键词:  光伏区间预测  自适应滚动匹配预测修正模式  谱聚类  宽度学习系统  stable误差分布
DOI:10.16081/j.epae.202201006
分类号:TM615
基金项目:江苏省重点研发计划资助项目(BE2020027)
Photovoltaic interval prediction based on adaptive rolling matching prediction correction mode
MEI Fei1, GU Jiaqi1, PEI Xin1, ZHENG Jianyong2
1.College of Energy and Electrical Engineering, Hohai University, Nanjing 211100, China;2.School of Electrical Engineering, Southeast University, Nanjing 210096, China
Abstract:
In order to solve the problem that the traditional point prediction method is difficult to quantify the uncertainty of photovoltaic generation power and improve the prediction accuracy, a photovoltaic interval prediction method based on adaptive rolling matching prediction correction mode is proposed. The historical photovoltaic data set is clustered by the spectral clustering combined with wavelet energy, and input and interval output of different clusters are constructed and trained by broad learning system. The stable error distribution of different clusters, different confidence intervals, and different prediction power intervals is established, and the optimal modified quantile point value for each prediction power interval is found combined with the optimization objective function. The rolling matching prediction correction mode is used for interval prediction. The case results of a 2.8 MW photovoltaic power station in a region of Wuxi, China show that the proposed method has better prediction effect compared with the traditional clustering prediction method.
Key words:  photovoltaic interval prediction  adaptive rolling matching prediction correction mode  spectral clustering  broad learning system  stable error distribution

用微信扫一扫

用微信扫一扫