EN基于改进相空间重构和多通道卷积-FPN网络的光伏功率短期预测
基于改进相空间重构和多通道卷积-FPN网络的光伏功率短期预测
- 太阳能学报 2025年第11期 页码：299-307
- 作者机构：
- 作者简介：
- 基金信息：国家自然科学基金（52275067）;河北省自然科学基金重点项目（F2020203058）;国家重点研发计划（2021YFB3201600）;国网冀北电力有限公司唐山供电公司项目（B30103230016）
DOI：doi:10.19912/j.0254-0096.tynxb.2024-1239
中图分类号： P181-
网络出版：2025-12-04，
纸质出版：2025-12-04
- 稿件说明：
移动端阅览
赵欣, 罗嘉宾, 姜安琦, 等. 基于改进相空间重构和多通道卷积-FPN网络的光伏功率短期预测[J]. 太阳能学报, 2025,(11):299-307.
赵欣, 罗嘉宾, 姜安琦, et al. 基于改进相空间重构和多通道卷积-FPN网络的光伏功率短期预测[J]. 2025, (11): 299-307.
赵欣, 罗嘉宾, 姜安琦, 等. 基于改进相空间重构和多通道卷积-FPN网络的光伏功率短期预测[J]. 太阳能学报, 2025,(11):299-307. DOI： doi:10.19912/j.0254-0096.tynxb.2024-1239.
赵欣, 罗嘉宾, 姜安琦, et al. 基于改进相空间重构和多通道卷积-FPN网络的光伏功率短期预测[J]. 2025, (11): 299-307. DOI： doi:10.19912/j.0254-0096.tynxb.2024-1239.
摘要
针对光伏功率不确定性大导致其预测精度不高的问题
提出一种基于改进相空间重构和多通道卷积-FPN网络的光伏功率短期预测方法。首先改进了相空间重构的方法
通过基于符号分析的极大联合熵求取延迟时间
减少计算量;通过改进Cao算法快速自动确定打开混沌结构的嵌入维数
具有准确性和高效性。再用改进相空间方法重构光伏功率
并证明其有混沌特性
经过多通道卷积神经网络（Mul-CNN）提取混沌特征
同时利用FPN网络
将光伏功率混沌特征与气象等特征进行特征融合作为FPN网络的动态输入特征
充分挖掘数据间的空间特征
最后通过多个全连接层得到预测结果。该方法在国内公开数据集上做应用实验
通过与其他网络模型相比
证明该预测模型的准确度更高。
Abstract
Aiming at the problem that the large uncertainty of PV power leads to its low prediction accuracy
a short-term prediction method of PV power based on improved phase space reconstruction and multi-channel convolution-feature pyramid networks（FPN） is proposed. Firstly
the phase space reconstruction method is improved to reduce the computation by finding the delay time via maximum joint entropy based on symbolic analysis
and the embedding dimension for unfolding the chaotic structure is quickly and automatically determined by the improved Cao algorithm
which has accuracy and efficiency. Then we reconstruct the PV power with the improved phase space method and prove that it has chaotic characteristics
extract the chaotic features by multi-channel convolutional neural network （Mul-CNN）
and at the same time
utilize the FPN network to fuse the chaotic features of the PV power with the meteorological features as the dynamic input features of the FPN network
to fully exploit the chaotic features of the FPN network. network
fully exploiting the spatial features between the data
and finally obtaining the prediction results through multiple fully connected layers. The method is evaluated on a domestic public dataset to do experiments
and compared with other network models
it proves that the prediction model is more accurate.
关键词
Keywords
references
牛威如, 魏凯, 王维洲, 等. 光伏新能源发电负荷不确定性补偿及微电网调度优化研究[J]. 自动化与仪器仪表, 2023(9): 156-160.
BASU M, DAS S.Short-term pumped storage hydrothermal generation scheduling considering uncertainty of load demand and renewable energy sources[J]. Journal of energy storage, 2023, 70: 107933.
孔祥玉, 马玉莹, 艾芊, 等. 新型电力系统多元用户的用电特征建模与用电负荷预测综述[J]. 电力系统自动化, 2023, 47(13): 2-17.
ELTAWIL M A, ZHAO Z M.Grid-connected photovoltaic power systems: technical and potential problems: a review[J]. Renewable and sustainable energy reviews, 2010, 14(1): 112-129.
TSAI W C, TU C S, HONG C M, et al.A review of state-of-the-art and short-term forecasting models for solar PV power generation[J]. Energies, 2023, 16(14): 5436.
贾凌云, 云斯宁, 赵泽妮, 等. 神经网络短期光伏发电预测的应用研究进展[J]. 太阳能学报, 2022, 43(12): 88-97.
秦宇, 许野, 王鑫鹏, 等. 基于改进FCM-LSTM的光伏出力短期预测研究[J]. 太阳能学报, 2024, 45(8): 304-313.
XIAO Z N, HUANG X Q, LIU J, et al.A novel method based on time series ensemble model for hourly photovoltaic power prediction[J]. Energy, 2023, 276: 127542.
吴春华, 董阿龙, 李智华, 等. 基于图相似日和PSO-XGBoost 的光伏功率预测[J]. 高电压技术, 2022, 48(8): 3250-3259.
王育飞, 郝德扬, 薛花, 等. 计及云图和混沌特性的光伏功率组合预测方法[J]. 太阳能学报, 2023, 44(12): 74-81.
SHI J, LEE W J, LIU Y Q, et al.Forecasting power output of photovoltaic systems based on weather classification and support vector machines[J]. IEEE transactions on industry applications, 2012, 48(3): 1064-1069.
WU Z J, CUI N B, GONG D Z, et al.Predicting daily global solar radiation in various climatic regions of China based on hybrid support vector machines with meta-heuristic algorithms[J]. Journal of cleaner production, 2023, 385: 135589.
ZHENG L W, LIU Z K, SHEN J N, et al.Very short-term maximum Lyapunov exponent forecasting tool for distributed photovoltaic output[J]. Applied energy, 2018, 229: 1128-1139.
BAZINE H, MABROUKI M.Chaotic dynamics applied in time prediction of photovoltaic production[J]. Renewable energy, 2019, 136: 1255-1265.
HAN L, ROMERO C E, YAO Z.Wind power forecasting based on principle component phase space reconstruction[J]. Renewable energy, 2015, 81: 737-744.
XU X F, HU S T, SHI P M, et al.Natural phase space reconstruction-based broad learning system for short-term wind speed prediction: case studies of an offshore wind farm[J]. Energy, 2023, 262: 125342.
王海燕, 盛昭瀚. 混沌时间序列相空间重构参数的选取方法[J]. 东南大学学报(自然科学版), 2000, 30(5): 113-117.
张淑清, 李新新, 张立国, 等. 基于符号分析的极大联合熵延迟时间求取方法[J]. 物理学报, 2013, 62(11): 118-123.
岳顺, 李小奇, 翟长治. 基于改进Cao算法确定奇异谱嵌入维数及应用[J]. 测绘工程, 2015, 24(3): 64-68.
黄磊, 舒杰, 姜桂秀, 等. 基于多维时间序列局部支持向量回归的微网光伏发电预测[J]. 电力系统自动化, 2014, 38(5): 19-24.
YAU H T, KUO P H, CAI D Y, et al.Construction of an online machine tool wear prediction system by using a time-delay phase space reconstruction-based dilation convolutional neural network[J]. IEEE sensors journal, 2023, 23(19): 22295-22312.
周飞燕, 金林鹏, 董军. 卷积神经网络研究综述[J]. 计算机学报, 2017, 40(6): 1229-1251.
HE K M, ZHANG X Y, REN S Q, et al.Deep residual learning for image recognition[C]//2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR). Las Vegas, NV, USA, 2016: 770-778.
LIN T Y, DOLLÁR P, GIRSHICK R, et al. Feature pyramid networks for object detection[C]//2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR). Honolulu, HI, USA, 2017: 936-944.
PENG F, MIAO Z, LI F, et al.S-FPN: a shortcut feature pyramid network for sea cucumber detection in underwater images[J]. Expert systems with applications, 2021, 182: 115306.
杨晨洲. 分布式光伏电站发电效率提升策略研究[J]. 光源与照明, 2023(9): 123-125.
0
浏览量
0
下载量
0
CSCD
- Meta-XML下载
- BibTeX下载
- Endnote下载
- RefMan 下载
- NoteExpress下载
- NoteFirst下载
关联资源
相关文章
相关作者
相关机构
AI问答- 技术支持由北京北大方正电子有限公司提供 京ICP备09064830号-19
京公网安备11010802024621 - 本系统建议在Chrome、 IE9+ 以上版本浏览器阅读本站内容，360浏览器请切换至极速模式
- Cookies帮助我们提供服务并提供个性化体验。使用本网站，即表示您同意我们使用Cookies