基于FCM-SENet-TCN的低压台区光伏超短期功率预测方法

doi:10.11930/j.issn.1004-9649.202409031

摘要/Abstract

摘要：

现有光伏功率预测的方法在应对低压台区分布式光伏时，存在初始数据过于冗余、预测特征提取困难，进而导致预测精度不足的问题。提出一种基于FCM-SENet-TCN的低压台区光伏超短期功率预测方法。首先，利用模糊C均值聚类算法（fuzzy cmeans，FCM）充分挖掘多源气象环境数据，将初始数据集以不同天气进行聚类，降低初始数据冗余度；其次，将压缩和激励网络（squeeze-and-excitation networks，SENet）融入时间卷积网络（temporal convolutional network，TCN），高效提取复杂特征并提高预测精度；最后，应用平均绝对百分比误差和均方根误差作为评价指标，对预测结果进行评估。仿真结果表明：所提预测方法可以充分利用初始气象数据，能够针对低压台区分布式光伏发电机组出力特点，做出更为精确的超短期功率预测。

关键词: 低压台区, 光伏功率预测, 模糊C均值聚类

Abstract:

The existing methods for predicting photovoltaic power face problems such as excessive initial data redundancy and difficulty in extracting predictive features when facing distributed photovoltaics in low-voltage substations, resulting in insufficient prediction accuracy. Therefore, this article proposes a low-voltage photovoltaic ultra short term power prediction method based on FCM-SENet TCN. Firstly, the fuzzy C-means clustering algorithm (FCM) is used to fully explore multi-source meteorological environment data, clustering the initial dataset with different weather conditions to reduce initial data redundancy; Secondly, the Squeeze and Excitation Networks (SENet) will be integrated into the Temporal Convolutional Network (TCN) to efficiently extract complex features and improve prediction accuracy; Finally, the average absolute percentage error and root mean square error are used as evaluation indicators to assess the prediction results. The simulation results show that the proposed prediction method can fully utilize the initial meteorological data and make more accurate ultra short term power predictions based on the output characteristics of distributed photovoltaic generators in low-voltage substations.

Key words: low voltage substation area, photovoltaic power prediction, fuzzy C-means clustering

魏伟, 余鹤, 叶利, 汪应春. 基于FCM-SENet-TCN的低压台区光伏超短期功率预测方法[J]. 中国电力, 2025, 58(6): 172-179.

WEI Wei, YU He, YE Li, WANG Yingchun. Low Voltage Substation Photovoltaic Ultra Short Term Power Prediction Method Based on FCM-SENet-TCN[J]. Electric Power, 2025, 58(6): 172-179.

导出引用管理器 EndNote|Ris|BibTeX

链接本文: https://www.electricpower.com.cn/CN/10.11930/j.issn.1004-9649.202409031

https://www.electricpower.com.cn/CN/Y2025/V58/I6/172

图/表 16

表 1 皮尔逊相关系数

Table 1 Pearson correlation coefficient

气象因素		相关系数
有功功率		1.000
温度		0.426
相对湿度		–0.405
太阳辐射度		0.912
风速		0.541
降雨量		–0.075
风向		–0.040

图 1 TCN模型结构

Fig.1 Model structure of TCN

图 2 残差连接结构

Fig.2 Residual connections' structure

图 3 FCM-SENet-TCN 模型预测流程

Fig.3 FCM-SENet TCN model prediction process

表 2 光伏系统数据

Table 2 Photovoltaic system data

项目		数据
逆变器尺寸/类型		4×6 kW；SMA SMC 6000 A
跟踪器类型		DEGE Renergie 5000 NT；双轴
阵列区域/m²		4×38.37
面板类型		天合TSM-195 DC01 A
面板数量/个		4×30
面板额定值/W		195
阵列评级/kW		23.4

图 4 春季预测结果误差

Fig.4 Spring forecast results error

图 5 夏季预测结果误差

Fig.5 Summer forecast results error

图 6 秋季预测结果误差

Fig.6 Autumn forecast results error

图 7 冬季预测结果误差

Fig.7 Winter forecast results error

表 3 不同季节的 RMSE值比较

Table 3 Comparison of RMSE values for different seasons

模型	春	夏	秋	冬	平均
FCM-LSTM	0.8479	1.3762	0.7325	0.4235	0.8450
FCM-TCN	0.6051	0.4352	0.6986	0.3995	0.5346
FCM-SENet-TCN	0.5389	0.4689	0.5096	0.3245	0.4604

表 4 不同季节的 MAPE值比较

Table 4 Comparison of MAE values for different seasons

模型	春	夏	秋	冬	平均
FCM-LSTM	0.7254	1.1702	0.5368	0.3368	0.6932
FCM-TCN	0.4589	0.4263	0.5076	0.2963	0.4223
FCM-SENet-TCN	0.4422	0.3692	0.4039	0.2896	0.3762

图 8 晴天预测结果误差

Fig.8 Sunny forecast results error

图 9 多云预测结果误差

Fig.9 Cloudy forecast results error

图 10 雨天预测结果误差

Fig.10 Rain forecast results error

表 5 不同天气条件下 RMSE值的比较

Table 5 Comparison of RMSE values for different weather conditions

模型	晴天	多云	雨天	平均
FCM-LSTM	0.8635	1.5468	1.6793	1.3632
FCM-TCN	0.4625	1.0956	1.6489	1.0691
FCM-SENet-TCN	0.4686	0.8896	1.4165	0.9249

表 6 不同天气条件下MAPE值的比较

Table 6 Comparison of MAE values for different weather conditions

模型	晴天	多云	雨天	平均
FCM-LSTM	0.7654	1.1232	1.1256	1.0047
FCM-TCN	0.3475	0.9032	1.2186	0.8231
FCM-SENet-TCN	0.3465	0.7365	0.9322	0.6717

参考文献 29

1	孙通, 张沈习, 曹毅, 等. 计及5G基站可调特性的配电网分布式光伏准入容量鲁棒优化[J]. 中国电力, 2025, 58 (2): 140- 146. DOI
	SUN Tong, ZHANG Shenxi, CAO Yi, et al. Robust optimization of hosting capacity of distributed photovoltaics in distribution network considering adjustable characteristics of 5G base station[J]. Electric Power, 2025, 58 (2): 140- 146. DOI
2	周洋, 黄德志, 李培栋, 等. 考虑平衡端点相位不对称及光伏接入的低压配电网三相潮流模型[J]. 中国电力, 2024, 57 (10): 190- 198.
	ZHOU Yang, HUANG Dezhi, LI Peidong, et al. A three-phase power flow model for low-voltage distribution networks considering balanced bus phase asymmetry and photovoltaic access[J]. Electric Power, 2024, 57 (10): 190- 198.
3	董强, 徐君, 方东平, 等. 基于光伏出力特性的分布式光储系统优化调度策略[J]. 综合智慧能源, 2024, 46 (4): 17- 23. DOI
	DONG Qiang, XU Jun, FANG Dongping, et al. Optimal scheduling strategy of distributed PV-energy storage systems based on PV output characteristics[J]. Integrated Intelligent Energy, 2024, 46 (4): 17- 23. DOI
4	葛亚明, 戴上, 梁文腾, 等. 基于气象融合与深度学习的分布式光伏出力区间预测[J]. 电网与清洁能源, 2024, 40 (8): 112- 120. DOI
	GE Yaming, DAI Shang, LIANG Wenteng, et al. Prediction of distributed photovoltaic output interval based on meteorological fusion and deep learning[J]. Power System and Clean Energy, 2024, 40 (8): 112- 120. DOI
5	宗龙, 赵亚典. 基于光伏建筑一体化的综合能源管控系统设计[J]. 内蒙古电力技术, 2023, 41 (6): 68- 74.
	ZONG Long, ZHAO Yadian. Design of comprehensive energy management and control system based on building integrated photovoltaics[J]. Inner Mongolia Electric Power, 2023, 41 (6): 68- 74.
6	陆毅, 薛枫, 唐小波, 等. 基于余弦相似度和TSO-BP的短期光伏预测方法[J]. 浙江电力, 2024, 43 (6): 22- 30.
	LU Yi, XUE Feng, TANG Xiaobo, et al. A short-term PV power forecasting method based on cosine similarity and TSO-BP neural network[J]. Zhejiang Electric Power, 2024, 43 (6): 22- 30.
7	朱涛, 杨欢红, 肖峰, 等. 基于改进灰狼算法和TCN-QRF的超短期光伏出力概率预测[J]. 浙江电力, 2024, 43 (8): 85- 93.
	ZHU Tao, YANG Huanhong, XIAO Feng, et al. Probabilistic forecasting of ultra-short-term PV output using the improved GWO and TCN-QRF[J]. Zhejiang Electric Power, 2024, 43 (8): 85- 93.
8	杨彪, 李佳蓉, 魏子轲. 美国分布式能源发展对我国的启示[J]. 中国电力企业管理, 2024, (25): 94- 96.
9	赵耀, 高少炜, 李东东, 等. 基于天气相似聚类与QRNN的短期光伏功率区间概率预测[J]. 电力系统自动化, 2023, 47 (23): 152- 161. DOI
	ZHAO Yao, GAO Shaowei, LI Dongdong, et al. Short-term interval probability prediction of photovoltaic power based on weather similarity clustering and quantile regression neural network[J]. Automation of Electric Power Systems, 2023, 47 (23): 152- 161. DOI
10	杨彪, 颜伟, 莫静山. 考虑源荷功率随机性和相关性的主导节点选择与无功分区方法[J]. 电力系统自动化, 2021, 45 (11): 61- 67.
	YANG Biao, YAN Wei, MO Jingshan. Pilot-bus selection and network partitioning method considering randomness and correlation of source-load power[J]. Automation of Electric Power Systems, 2021, 45 (11): 61- 67.
11	马成廉, 李闯, 薛冰, 等. 含高比例光伏配电网分区电压协调控制策略[J]. 东北电力大学学报, 2024, 44 (4): 77- 85.
	MA Chenglian, LI Chuang, XUE Bing, et al. Voltage coordination control strategy for zoning in distribution network with high proportion of photovoltaic[J]. Journal of Northeast Electric Power University, 2024, 44 (4): 77- 85.
12	杨康, 李蓝青, 李艺丰, 等. 一种新型分布式光伏出力区间预测方法[J]. 发电技术, 2024, 45 (4): 684- 695. DOI
	YANG Kang, LI Lanqing, LI Yifeng, et al. A novel distributed photovoltaic output interval prediction method[J]. Power Generation Technology, 2024, 45 (4): 684- 695. DOI
13	JIAO X, LI X S, LIN D Y, et al. A graph neural network based deep learning predictor for spatio-temporal group solar irradiance forecasting[J]. IEEE Transactions on Industrial Informatics, 2022, 18 (9): 6142- 6149. DOI
14	张赟宁, 魏广军. 考虑特征选择的短期光伏功率组合预测模型[J]. 电力系统及其自动化学报, 2024, 36 (8): 122- 132.
	ZHANG Yunning, WEI Guangjun. Combined prediction model for short-term photovoltaic power considering feature selection[J]. Proceedings of the CSU-EPSA, 2024, 36 (8): 122- 132.
15	毕贵红, 张梓睿, 赵四洪, 等. 基于多模式分解和多分支输入的光伏功率超短期预测[J]. 高电压技术, 2024, 50 (9): 3837- 3849.
	BI Guihong, ZHANG Zirui, ZHAO Sihong, et al. Ultra-short-term prediction of photovoltaic power based on multi-mode decomposition and multi-branch input[J]. High Voltage Engineering, 2024, 50 (9): 3837- 3849.
16	欧阳静, 秦龙, 王坚锋, 等. 基于PCA-ShapeDTW-QWGRU的分布式光伏集群短期功率预测[J]. 太阳能学报, 2024, 45 (5): 458- 467.
	OUYANG Jing, QIN Long, WANG Jianfeng, et al. Short-term power prediction for distributed pv clusters based on pca-shapedtw-qwgru[J]. Acta Energiae Solaris Sinica, 2024, 45 (5): 458- 467.
17	刘源延, 孔小兵, 马乐乐, 等. 基于小波包变换与深度学习的超短期光伏功率预测[J]. 太阳能学报, 2024, 45 (5): 537- 546.
	LIU Yuanyan, KONG Xiaobing, MA Lele, et al. Ultra-short-term photovoltaic power forecasting based on wavelet packet transform and deep learning[J]. Acta Energiae Solaris Sinica, 2024, 45 (5): 537- 546.
18	王于波, 郝玲, 徐飞, 等. 分布式光伏集群发电功率波动模式识别与超短期概率预测[J]. 上海交通大学学报, 2024, 58 (9): 1334- 1343.
	WANG Yubo, HAO Ling, XU Fei, et al. Pattern recognition and ultra-short-term probabilistic forecasting of power fluctuating in aggregated distributed photovoltaics clusters[J]. Journal of Shanghai Jiao Tong University, 2024, 58 (9): 1334- 1343.
19	QU J Q, QIAN Z, PEI Y. Day-ahead hourly photovoltaic power forecasting using attention-based CNN-LSTM neural network embedded with multiple relevant and target variables prediction pattern[J]. Energy, 2021, 232, 120996. DOI
20	毛明轩, 冯心营, 陈思宇, 等. 基于贝叶斯优化卷积神经网络的路面光伏阵列最大功率点电压预测方法[J]. 中国电机工程学报, 2024, 44 (2): 620- 631.
	MAO Mingxuan, FENG Xinying, CHEN Siyu, et al. A novel maximum power point voltage forecasting method for pavement photovoltaic array based on Bayesian optimization convolutional neural network[J]. Proceedings of the CSEE, 2024, 44 (2): 620- 631.
21	HUANG C J, KUO P H. Multiple-input deep convolutional neural network model for short-term photovoltaic power forecasting[J]. IEEE Access, 2019, 7, 74822- 74834. DOI
22	李彦伦, 窦晓波, 卜强生, 等. 基于改进FCM和最小互信息算法的户变关系辨识方法[J]. 电力系统保护与控制, 2024, 52 (11): 102- 111.
	LI Yanlun, DOU Xiaobo, BU Qiangsheng, et al. Identification method of transformer-customer relationship based on an improved FCM algorithm and minimum mutual information[J]. Power System Protection and Control, 2024, 52 (11): 102- 111.
23	王煜尘, 窦银科, 孟润泉. 基于模糊C均值聚类-变分模态分解和群智能优化的多核神经网络短期负荷预测模型[J]. 高电压技术, 2022, 48 (4): 1308- 1319.
	WANG Yuchen, DOU Yinke, MENG Runquan. Forecasting model for multicore neural network short-term load based on fuzzy C-mean clustering-variational modal decomposition and chaotic swarm intelligence optimization[J]. High Voltage Engineering, 2022, 48 (4): 1308- 1319.
24	HU J, SHEN L, SUN G. Squeeze-and-excitation networks[C]//2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition. Salt Lake City, UT, USA. IEEE, 2018: 7132–7141.
25	邢晨, 张照贝. 基于改进时间卷积网络的短期光伏出力概率预测方法[J]. 太阳能学报, 2023, 44 (2): 373- 380.
	XING Chen, ZHANG Zhaobei. Short-term probabilistic forecasting method of photovoltaic output power based on improved temporal convolutional network[J]. Acta Energiae Solaris Sinica, 2023, 44 (2): 373- 380.
26	朱永清, 林佳宁, 李庆生, 等. 冰灾下考虑多重不确定性的负荷聚合商市场力评估方法[J]. 浙江电力, 2024, 43 (1): 64- 71.
	ZHU Yongqing, LIN Jianing, LI Qingsheng, et al. A market power assessment method for load aggregators considering multiple uncertainties under ice disasters[J]. Zhejiang Electric Power, 2024, 43 (1): 64- 71.
27	王光达, 杨彪, 张琛. 从电网规划视角看电力供需[J]. 中国电力企业管理, 2023, (4): 33- 35.
28	于松源, 张峻松, 元志伟, 等. 计及热惯性的热电联产虚拟电厂韧性提升策略[J]. 发电技术, 2023, 44 (6): 758- 768.
	YU Songyuan, ZHANG Junsong, YUAN Zhiwei, et al. Resilience enhancement strategy of combined heat and power-virtual power plant considering thermal inertia[J]. Power Generation Technology, 2023, 44 (6): 758- 768.
29	黄四亮, 孟湛博, 邢晓敏, 等. 考虑灾前预防和灾时响应状态下的弹性配电网提升策略研究[J]. 东北电力大学学报, 2023, 43 (4): 82- 89.
	HUANG Siliang, MENG Zhanbo, XING Xiaomin, et al. Consider the research on the improvement strategy of resilient distribution network under pre-disaster prevention and disaster response[J]. Journal of Northeast Electric Power University, 2023, 43 (4): 82- 89.

[1]	史子轶, 夏向阳, 刘佳斌, 谷阳洋, 王玉龙, 洪佳瑶. 基于AKNN异常检验与ADPC聚类的低压台区拓扑识别方法[J]. 中国电力, 2024, 57(5): 168-177.
[2]	战文华, 车建峰, 王勃, 丁禹. 基于网格化数值天气预报的区域光伏发电多输出功率预测方法[J]. 中国电力, 2024, 57(3): 144-151.
[3]	胡志强, 叶远誉, 俞林刚, 伍栋文, 胡涛. 主动负荷参与的低压台区三相负荷不平衡自动均衡方法[J]. 中国电力, 2024, 57(2): 49-54.
[4]	李翰章, 冯江涛, 王鹏程, 荣澔洁, 柴宇唤. 基于TDE-SO-AWM-GRU的光伏发电功率预测模型[J]. 中国电力, 2024, 57(12): 41-49.
[5]	柳守诚, 王淳, 邹智辉, 陈佳慧, 周晗, 刘伟, 张旭. 基于t-SNE降维和放射传播聚类算法的低压配电网相位识别[J]. 中国电力, 2023, 56(5): 108-117.
[6]	张娜, 任强, 刘广忱, 郭力萍, 李静宇. 基于VMD-GWO-ELMAN的光伏功率短期预测方法[J]. 中国电力, 2022, 55(5): 57-65.
[7]	徐遵义, 刘文慧, 张旭冉, 张海燕. 基于数据驱动的浆液循环泵运行优化研究[J]. 中国电力, 2021, 54(3): 197-204.
[8]	梁适春, 张晓冬, 林培峰, 牛萌. 一种混合储能光伏发电系统的功率预测算法[J]. 中国电力, 2014, 47(3): 24-27.