一种新型天气分型方法及其在光伏功率预测中的应用

doi:10.16183/j.cnki.jsjtu.2021.264

摘要/Abstract

摘要：

为提高光伏功率预测准确率提出了一种新的天气分型方法,该方法首先按总云量大小区分晴天和云天,然后根据太阳被遮蔽的程度将云天进一步细分为三类.该方法能有效识别影响光伏出力的关键气象环境因子特征,并对其加权求和得到新型分类指标Sky Condition Factor(SCF).该方法物理意义明确,区分度较好且易于量化.对天气类型合理细分后,可消除众多气象环境因子之间的耦合关系,降低输入变量维度,易于统计建模.最后分别基于原理和统计方法进行建模验证,结果显示该方法可以有效提高光伏功率预测的准确率.

关键词: 光伏功率预测, 天气分型, 气象环境因子, 原理预测法, 统计预测法

Abstract:

To improve the accuracy of photovoltaic (PV) power prediction, this paper proposes a novel weather classification method. First, it distinguishs the clear days and cloudy days according to the total cloud cover. Then, it further classifies the cloudy days into three subtypes to investigate whether the sun is obscured by clouds. This method can effectively identify the characteristics of key meteorological environmental factors that affect PV output and form a new classification index sky condition factor (SCF) by weighted summation. This method has clear physical meanings, good discrimination, and easy quantification. The reasonable classification of weather types can eliminate the coupling relationship between many meteorological environmental factors and reduce the dimension of input variables, which makes it easy for statistical modeling. Based on the theoretical and the statistical approachs respectively, the modeling and verification are conducted and the results show that the method can effectively improve the accuracy of PV power prediction.

Key words: photovoltaic (PV) power prediction, weather type classification, meteorological environmental factors, physical approach, statistic approach

中图分类号:

TM615
P49

李芬, 周尔畅, 孙改平, 白永清, 童力, 刘邦银, 赵晋斌. 一种新型天气分型方法及其在光伏功率预测中的应用[J]. 上海交通大学学报, 2021, 55(12): 1510-1519.

LI Fen, ZHOU Erchang, SUN Gaiping, BAI Yongqing, TONG Li, LIU Bangyin, ZHAO Jinbin. A Novel Weather Classification Method and Its Application in Photovoltaic Power Prediction[J]. Journal of Shanghai Jiao Tong University, 2021, 55(12): 1510-1519.

图/表 16

图1

表1

表2

图2

图3

图4

表3

非线性回归模型

模型名称	表达式	RSS	R²
Metcherlich	$y = b 1 + b 2 exp (- b 3 P)$	3.140	0.983
Johnson-Schumacher	$y = b 1 exp (- b 2 / (P + b 3))$	1.376	0.992
Morgen-Mercer-Florin	$y = (b 1 b 2 + b 3 P b 4) / (b 2 + P b 4)$	1.372	0.992
Peal-Reed	$y = b 1 / {1 + b 2 exp [- (b 3 P + b 4 P 2 + b 5 P 3)]}$	1.807	0.990
Richards	$y=b_{1}/\{[1+b_{3}exp(-b_{2}P)]^{1/b_{4}}\}$	1.470	0.992

表3

表4

图5

表5

表6

图6

表7

图7

表8

表9

参考文献 26

[1]	中华人民共和国国家统计局. 中华人民共和国2020年国民经济和社会发展统计公报[EB/OL].(2021-02-28)[2021-06-01]. http://www.gov.cn/xinwen/2021-02/28/content_5589283.htm.
	National Bureau of Statistics of the People’s Republic of China. 2020 statistical bulletin on national economic and social development of the People’s Republic of China[EB/OL].(2021-02-28)[2021-06-01]. http://www.gov.cn/xinwen/2021-02/28/content_5589283.htm.
[2]	CALCABRINI A, ZIAR H, ISABELLA O, et al. A simplified skyline-based method for estimating the annual solar energy potential in urban environments[J]. Nature Energy, 2019, 4(3):206-215. doi: 10.1038/s41560-018-0318-6 URL
[3]	SHENG H M, XIAO J, CHENG Y H, et al. Short-term solar power forecasting based on weighted Gaussian process regression[J]. IEEE Transactions on Industrial Electronics, 2018, 65(1):300-308. doi: 10.1109/TIE.2017.2714127 URL
[4]	AMEUR A, BERRADA A, LOUDIYI K, et al. Forecast modeling and performance assessment of solar PV systems[J]. Journal of Cleaner Production, 2020, 267:122167.
[5]	ABDEL-NASSER M, MAHMOUD K. Accurate photovoltaic power forecasting models using deep LSTM-RNN[J]. Neural Computing and Applications, 2019, 31(7):2727-2740. doi: 10.1007/s00521-017-3225-z URL
[6]	YU D, CHOI W, KIM M, et al. Forecasting day-ahead hourly photovoltaic power generation using convolutional self-attention based long short-term memory[J]. Energies, 2020, 13(15):4017. doi: 10.3390/en13154017 URL
[7]	杨晶显, 张帅, 刘继春, 等. 基于VMD和双重注意力机制LSTM的短期光伏功率预测[J]. 电力系统自动化, 2021, 45(3):174-182.
	YANG Jingxian, ZHANG Shuai, LIU Jichun, et al. Short-term photovoltaic power prediction based on variational mode decomposition and long short term memory with dual-stage attention mechanism[J]. Automation of Electric Power Systems, 2021, 45(3):174-182.
[8]	单英浩, 付青, 耿炫, 等. 基于改进BP-SVM-ELM与粒子化SOM-LSF的微电网光伏发电组合预测方法[J]. 中国电机工程学报, 2016, 36(12):3334-3343.
	SHAN Yinghao, FU Qing, GENG Xuan, et al. Combined forecasting of photovoltaic power generation in microgrid based on the improved BP-SVM-ELM and SOM-LSF with particlization[J]. Proceedings of the CSEE, 2016, 36(12):3334-3343.
[9]	袁晓玲, 施俊华, 徐杰彦. 计及天气类型指数的光伏发电短期出力预测[J]. 中国电机工程学报, 2013, 33(34):57-64.
	YUAN Xiaoling, SHI Junhua, XU Jieyan. Short-term power forecasting for photovoltaic generation considering weather type index[J]. Proceedings of the CSEE, 2013, 33(34):57-64.
[10]	廖卫强, 张认成, 俞万能, 等. 基于相似样本及PCA的光伏输出功率预测[J]. 太阳能学报, 2016, 37(9):2377-2385.
	LIAO Weiqiang, ZHANG Rencheng, YU Wanneng, et al. Prediction of output power of photovoltaic based on similar samples and principal component analysis[J]. Acta Energiae Solaris Sinica, 2016, 37(9):2377-2385.
[11]	王飞, 米增强, 甄钊, 等. 基于天气状态模式识别的光伏电站发电功率分类预测方法[J]. 中国电机工程学报, 2013, 33(34):75-82.
	WANG Fei, MI Zengqiang, ZHEN Zhao, et al. A classified forecasting approach of power generation for photovoltaic plants based on weather condition pattern recognition[J]. Proceedings of the CSEE, 2013, 33(34):75-82.
[12]	CHEN C S, DUAN S X, CAI T, et al. Online 24-h solar power forecasting based on weather type classification using artificial neural network[J]. Solar Energy, 2011, 85(11):2856-2870. doi: 10.1016/j.solener.2011.08.027 URL
[13]	ARMSTRONG S, HURLEY W G. A new methodology to optimise solar energy extraction under cloudy conditions[J]. Renewable Energy, 2010, 35(4):780-787. doi: 10.1016/j.renene.2009.10.018 URL
[14]	DEMAIN C, JOURNÉE M, BERTRAND C. Evaluation of different models to estimate the global solar radiation on inclined surfaces[J]. Renewable Energy, 2013, 50:710-721. doi: 10.1016/j.renene.2012.07.031 URL
[15]	LI F, LIN Y L, GUO J P, et al. Novel models to estimate hourly diffuse radiation fraction for global radiation based on weather type classification[J]. Renewable Energy, 2020, 157:1222-1232. doi: 10.1016/j.renene.2020.05.080 URL
[16]	中华人民共和国国家质量监督检验检疫总局, 中国国家标准化管理委员会. 中华人民共和国国家标准: 地面气象观测规范总则. GB/T 35221—2017[S]. 北京: 中国质量标准出版传媒有限公司, 2017.
	General Administration of Quality Supervision, Inspection and Quarantine of the People’s Republic of China, Standardization Administration of the People’s Republic of China. National Standard of the People’s Republic of China: Specifications for surface meteorological observation—General. GB/T 35221—2017[S]. Beijing: China Quality and Standards Publishing & Media Co., Ltd., 2017.
[17]	汪冬华. 多元统计分析与SPSS应用[M]. 上海: 华东理工大学出版社, 2017.
	WANG Donghua. Multivariate statistical analysis and application of SPSS[M]. Shanghai: East China University of Science and Technology Press, 2017.
[18]	盛裴轩, 毛节泰, 李建国, 等. 大气物理学[M]. 第2版. 北京: 北京大学出版社, 2013.
	SHENG Peixuan, MAO Jietai, LI Jianguo, et al. Atmospheric physics[M]. 2nd ed. Beijing: Peking University Press, 2013.
[19]	LIUB Y H, JORDAN R C. The interrelationship and characteristic distribution of direct, diffuse and total solar radiation[J]. Solar Energy, 1960, 4(3):1-19.
[20]	HAY J E. Calculation of monthly mean solar radiation for horizontal and inclined surfaces[J]. Solar Energy, 1979, 23(4):301-307. doi: 10.1016/0038-092X(79)90123-3 URL
[21]	KLUCHER T M. Evaluation of models to predict insolation on tilted surfaces[J]. Solar Energy, 1979, 23(2):111-114. doi: 10.1016/0038-092X(79)90110-5 URL
[22]	PEREZ R, STEWART R, ARBOGAST C, et al. An anisotropic hourly diffuse radiation model for sloping surfaces: Description, performance validation, site dependency evaluation[J]. Solar Energy, 1986, 36(6):481-497. doi: 10.1016/0038-092X(86)90013-7 URL
[23]	REINDL D T, BECKMAN W A, DUFFIE J A. Evaluation of hourly tilted surface radiation models[J]. Solar Energy, 1990, 45(1):9-17. doi: 10.1016/0038-092X(90)90061-G URL
[24]	LI F, LI C Y, SHI J, et al. Evaluation index system for photovoltaic systems statistical characteristics under hazy weather conditions in central China[J]. IET Renewable Power Generation, 2017, 11(14):1794-1803. doi: 10.1049/rpg2.v11.14 URL
[25]	李芬, 陈正洪, 马年骏, 等. 光伏资源精细化评估与预报技术研究[M]. 北京: 气象出版社, 2016.
	LI Fen, CHEN Zhenghong, MA Nianjun, et al. Photovoltaic resources of fine evaluation and forecast technology research[M]. Beijing: China Meteorological Press, 2016.
[26]	张犁, 胡海兵, 冯兰兰, 等. 模块化光伏并网系统欧洲效率优化控制方法[J]. 中国电机工程学报, 2012, 32(9):7-13.
	ZHANG Li, HU Haibing, FENG Lanlan, et al. European efficiency improvement control for grid-connected modular photovoltaic generation systems[J]. Proceedings of the CSEE, 2012, 32(9):7-13.

变量	R
I	0.868
S_p	0.672
k'_T	0.657
B_d	0.612
RH	-0.442
V	0.378
T	0.262
C	-0.254
W	0.150
RA	-0.123
p	-0.007

天气类型	个案数目	μ_SCF取值范围
2	1714	ξ₁<μ_SCF≤ξ₂
3	1477	ξ₂<μ_SCF≤ξ₃
4	2468	ξ₃<μ_SCF≤ξ₄

模型	天气类型1		天气类型2		天气类型3		天气类型4
模型	MAPE/%	NRMSE/%	MAPE/%	NRMSE/%	MAPE/%	NRMSE/%	MAPE/%	NRMSE/%
Perez	24.87	36.32	23.32	29.95	38.28	52.57	52.68	88.86
Hay	27.73	39.95	26.23	33.62	38.20	52.67	52.51	88.71
Reindl	27.64	39.88	26.03	33.49	38.16	52.55	52.38	88.59
Klucher	27.55	39.24	25.90	33.06	43.16	55.78	59.49	89.50
Liu & Jordan	25.05	36.48	23.54	30.07	37.14	51.08	20.72	32.91

变量名	与P_ac的相关系数
变量名	天气类型1	天气类型2	天气类型3	天气类型4
I	0.884	0.736	0.748	0.690
I_b	0.820	0.668	0.523	0.194
I_d	0.604	0.425	0.664	0.690
B_d	0.598	0.311	-0.045	-0.092
S_p	0.581	0.368	0.258	0.112
k'_T	0.536	0.502	0.331	0.007
V	0.385	0.152	0.233	0.200
W	0.341	0.166	0.080	-0.031
RH	-0.247	-0.283	-0.307	-0.348
T	0.119	0.087	0.249	0.308
C	-0.052	0.046	-0.008	-0.072
R	-0.041	0.007	-0.126	-0.119

模型	天气类型1		天气类型2		天气类型3		天气类型4
模型	MAPE/%	NRMSE/%	MAPE/%	NRMSE/%	MAPE/%	NRMSE/%	MAPE/%	NRMSE/%
LR	20.06	25.11	18.87	22.30	33.43	58.09	47.50	70.39
GPR	13.18	19.73	15.78	18.47	32.94	58.93	34.73	61.58
SVR	14.52	20.40	14.30	16.94	33.87	58.33	31.56	54.50
Adaboost	16.61	24.15	19.38	23.42	31.02	52.56	42.94	66.29