使用 SOM 聚类和 ECA 的超短期光伏预测混合模型

IF 1.6 4区 工程技术 Q3 ENGINEERING, ELECTRICAL & ELECTRONIC
Yixin Zhu, Ziyao Wang, Wei Zhang, Yufan Liu, Hao Wu
{"title":"使用 SOM 聚类和 ECA 的超短期光伏预测混合模型","authors":"Yixin Zhu, Ziyao Wang, Wei Zhang, Yufan Liu, Hao Wu","doi":"10.1007/s00202-024-02710-3","DOIUrl":null,"url":null,"abstract":"<p>The precision of the ultra-short-term PV power prediction is crucial for the grid to operate safely and steadily and for PV electricity to be connected on a broad scale. A combination model of ultra-short-term PV prediction based on an attention mechanism is proposed to increase the prediction accuracy of PV output power under various weather circumstances. First, using a Pearson correlation coefficient analysis, important climatic variables closely associated with PV power generation are selected and normalized monthly. The sky condition factor (SCF), a classification index, is computed using a weighted summation. This reduces the dimensionality of the input variables and eliminates seasonal influence on weather classification and the coupling interactions among various meteorological elements. Second, an unsupervised clustering of SCFs using a self-organizing map (SOM) neural network is used to classify three types of weather. After that, convolutional neural networks (CNNs) prediction models are built for each of the three types of weather. The efficient channel attention (ECA) module is then added, allowing the model to focus on key feature information and increase prediction accuracy by adaptively assigning phase weights to each of the multiple channels of feature information that the CNN has extracted. Lastly, the efficacy of the suggested prediction model is verified by simulations run on historical observed data, which demonstrate an improvement in the prediction models accuracy under various weather conditions when compared to the model without the ECA module.</p>","PeriodicalId":50546,"journal":{"name":"Electrical Engineering","volume":null,"pages":null},"PeriodicalIF":1.6000,"publicationDate":"2024-09-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":"{\"title\":\"A hybrid model for ultra-short-term PV prediction using SOM clustering and ECA\",\"authors\":\"Yixin Zhu, Ziyao Wang, Wei Zhang, Yufan Liu, Hao Wu\",\"doi\":\"10.1007/s00202-024-02710-3\",\"DOIUrl\":null,\"url\":null,\"abstract\":\"<p>The precision of the ultra-short-term PV power prediction is crucial for the grid to operate safely and steadily and for PV electricity to be connected on a broad scale. A combination model of ultra-short-term PV prediction based on an attention mechanism is proposed to increase the prediction accuracy of PV output power under various weather circumstances. First, using a Pearson correlation coefficient analysis, important climatic variables closely associated with PV power generation are selected and normalized monthly. The sky condition factor (SCF), a classification index, is computed using a weighted summation. This reduces the dimensionality of the input variables and eliminates seasonal influence on weather classification and the coupling interactions among various meteorological elements. Second, an unsupervised clustering of SCFs using a self-organizing map (SOM) neural network is used to classify three types of weather. After that, convolutional neural networks (CNNs) prediction models are built for each of the three types of weather. The efficient channel attention (ECA) module is then added, allowing the model to focus on key feature information and increase prediction accuracy by adaptively assigning phase weights to each of the multiple channels of feature information that the CNN has extracted. Lastly, the efficacy of the suggested prediction model is verified by simulations run on historical observed data, which demonstrate an improvement in the prediction models accuracy under various weather conditions when compared to the model without the ECA module.</p>\",\"PeriodicalId\":50546,\"journal\":{\"name\":\"Electrical Engineering\",\"volume\":null,\"pages\":null},\"PeriodicalIF\":1.6000,\"publicationDate\":\"2024-09-09\",\"publicationTypes\":\"Journal Article\",\"fieldsOfStudy\":null,\"isOpenAccess\":false,\"openAccessPdf\":\"\",\"citationCount\":\"0\",\"resultStr\":null,\"platform\":\"Semanticscholar\",\"paperid\":null,\"PeriodicalName\":\"Electrical Engineering\",\"FirstCategoryId\":\"5\",\"ListUrlMain\":\"https://doi.org/10.1007/s00202-024-02710-3\",\"RegionNum\":4,\"RegionCategory\":\"工程技术\",\"ArticlePicture\":[],\"TitleCN\":null,\"AbstractTextCN\":null,\"PMCID\":null,\"EPubDate\":\"\",\"PubModel\":\"\",\"JCR\":\"Q3\",\"JCRName\":\"ENGINEERING, ELECTRICAL & ELECTRONIC\",\"Score\":null,\"Total\":0}","platform":"Semanticscholar","paperid":null,"PeriodicalName":"Electrical Engineering","FirstCategoryId":"5","ListUrlMain":"https://doi.org/10.1007/s00202-024-02710-3","RegionNum":4,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q3","JCRName":"ENGINEERING, ELECTRICAL & ELECTRONIC","Score":null,"Total":0}
引用次数: 0

摘要

超短期光伏功率预测的精度对于电网的安全稳定运行和光伏发电的大范围接入至关重要。本文提出了一种基于注意力机制的超短期光伏预测组合模型,以提高各种天气条件下光伏输出功率的预测精度。首先,利用皮尔逊相关系数分析,选择与光伏发电密切相关的重要气候变量,并按月进行归一化处理。天空条件因子(SCF)是一种分类指数,通过加权求和计算得出。这降低了输入变量的维度,消除了季节对天气分类的影响以及各种气象要素之间的耦合相互作用。其次,利用自组织图(SOM)神经网络对 SCF 进行无监督聚类,以对三种天气进行分类。然后,为这三种天气分别建立卷积神经网络(CNN)预测模型。然后添加高效通道关注(ECA)模块,通过自适应地为 CNN 提取的多通道特征信息中的每个通道分配相位权重,使模型能够关注关键特征信息并提高预测精度。最后,通过对历史观测数据进行模拟运行,验证了所建议的预测模型的有效性,结果表明,与没有 ECA 模块的模型相比,预测模型在各种天气条件下的准确性都有所提高。
本文章由计算机程序翻译,如有差异,请以英文原文为准。

A hybrid model for ultra-short-term PV prediction using SOM clustering and ECA

A hybrid model for ultra-short-term PV prediction using SOM clustering and ECA

The precision of the ultra-short-term PV power prediction is crucial for the grid to operate safely and steadily and for PV electricity to be connected on a broad scale. A combination model of ultra-short-term PV prediction based on an attention mechanism is proposed to increase the prediction accuracy of PV output power under various weather circumstances. First, using a Pearson correlation coefficient analysis, important climatic variables closely associated with PV power generation are selected and normalized monthly. The sky condition factor (SCF), a classification index, is computed using a weighted summation. This reduces the dimensionality of the input variables and eliminates seasonal influence on weather classification and the coupling interactions among various meteorological elements. Second, an unsupervised clustering of SCFs using a self-organizing map (SOM) neural network is used to classify three types of weather. After that, convolutional neural networks (CNNs) prediction models are built for each of the three types of weather. The efficient channel attention (ECA) module is then added, allowing the model to focus on key feature information and increase prediction accuracy by adaptively assigning phase weights to each of the multiple channels of feature information that the CNN has extracted. Lastly, the efficacy of the suggested prediction model is verified by simulations run on historical observed data, which demonstrate an improvement in the prediction models accuracy under various weather conditions when compared to the model without the ECA module.

求助全文
通过发布文献求助,成功后即可免费获取论文全文。 去求助
来源期刊
Electrical Engineering
Electrical Engineering 工程技术-工程:电子与电气
CiteScore
3.60
自引率
16.70%
发文量
0
审稿时长
>12 weeks
期刊介绍: The journal “Electrical Engineering” following the long tradition of Archiv für Elektrotechnik publishes original papers of archival value in electrical engineering with a strong focus on electric power systems, smart grid approaches to power transmission and distribution, power system planning, operation and control, electricity markets, renewable power generation, microgrids, power electronics, electrical machines and drives, electric vehicles, railway electrification systems and electric transportation infrastructures, energy storage in electric power systems and vehicles, high voltage engineering, electromagnetic transients in power networks, lightning protection, electrical safety, electrical insulation systems, apparatus, devices, and components. Manuscripts describing theoretical, computer application and experimental research results are welcomed. Electrical Engineering - Archiv für Elektrotechnik is published in agreement with Verband der Elektrotechnik Elektronik Informationstechnik eV (VDE).
×
引用
GB/T 7714-2015
复制
MLA
复制
APA
复制
导出至
BibTeX EndNote RefMan NoteFirst NoteExpress
×
提示
您的信息不完整,为了账户安全,请先补充。
现在去补充
×
提示
您因"违规操作"
具体请查看互助需知
我知道了
×
提示
确定
请完成安全验证×
copy
已复制链接
快去分享给好友吧!
我知道了
右上角分享
点击右上角分享
0
联系我们:info@booksci.cn Book学术提供免费学术资源搜索服务,方便国内外学者检索中英文文献。致力于提供最便捷和优质的服务体验。 Copyright © 2023 布克学术 All rights reserved.
京ICP备2023020795号-1
ghs 京公网安备 11010802042870号
Book学术文献互助
Book学术文献互助群
群 号:481959085
Book学术官方微信