针对光伏发电功率存在的较大随机性和不确定性问题,提出一种基于改进长短期记忆神经网络的光伏发电功率预测方法,以此提高光伏发电功率预测的准确性。首先,分析与光伏发电出力相关性较强的气象特征,并利用t分布近邻嵌入降维技术将被选取的特征数据降至二维,以减小数据复杂度。然后,通过密度峰值聚类将降维后的数据自动聚成3类,帮助训练长短期记忆神经网络预测模型。与传统循环神经网络和长短期记忆神经网络模型相比,所提模型在光伏发电功率预测方面表现出较高的预测精度,MSE减少49.00%和31.77%,RMSE减少28.59%和17.41%,MAE减少62.35%和53.52%。研究结果表明,该模型在光伏发电功率预测方面具有较好的适用性。
Abstract
In response to the significant randomness and uncertainty issues in photovoltaic power, this paper proposes an improved LSTM based photovoltaic power prediction method to improve the accuracy of photovoltaic power prediction. Firstly, meteorological features with strong correlation with photovoltaic output were analyzed, and the t-distribution nearest neighbor embedding dimensionality reduction technique was used to reduce the selected feature data to 2D to reduce data complexity. Then, the reduced dimensionality data is automatically clustered into three categories through density peak clustering to help train the LSTM prediction model. Compared with the traditional Recurrent neural network and Long short-term memory neural network models, the model proposed in this paper shows higher prediction accuracy in photovoltaic power prediction. MSE decreases by 49.00% and 31.77%, RMSE decreases by 28.59% and 17.41%, and MAE decreases by 62.35% and 53.52%. The research results indicate that the model has good applicability in photovoltaic power prediction, providing valuable reference for the optimization and scheduling of integrated energy systems.
关键词
光伏出力 /
预测 /
神经网络 /
聚类分析 /
t分布近邻嵌入
Key words
photovoltaic output /
prediction /
neural network /
cluster analysis /
t-distribution nearest neighbor embedding
{{custom_sec.title}}
{{custom_sec.title}}
{{custom_sec.content}}
参考文献
[1] 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.
JIA L Y, YUN S N, ZHAO Z N, et al.Recent progress of short-term forecasting of photovoltaic generation based on artificial neural networks[J]. Acta energiae solaris sinica, 2022, 43(12): 88-97.
[2] DUDA J, KUSA R, PIETRUSZKO S, et al.Development of roadmap for photovoltaic solar technologies and market in Poland[J]. Energies, 2021, 15(1): 174.
[3] 宋文峰, 王倩, 李亚楠, 等. 基于双馈风电场低电压穿越的高温超导故障限流器建模及其参数优化[J]. 电力系统保护与控制, 2019, 47(18): 99-107.
SONG W F, WANG Q, LI Y N, et al.Modeling and parameter optimization of HTS-FCL based on low voltage ride-through of doubly-fed wind farm[J]. Power system protection and control, 2019, 47(18): 99-107.
[4] ZHANG F, WANG X H, WU M Y, et al.Optimization design of uncertain parameters for improving the stability of photovoltaic system[J]. Journal of power sources, 2022, 521: 230959.
[5] SENS L, NEULING U, KALTSCHMITT M.Capital expenditure and levelized cost of electricity of photovoltaic plants and wind turbines: development by 2050[J]. Renewable energy, 2022, 185: 525-537.
[6] 李燚航, 翟卫欣, 颜寒祺, 等. 基于U-net神经网络模型的PM2.5逐小时浓度值预测模型[J]. 北京大学学报(自然科学版), 2020, 56(5): 796-804.
LI Y H, ZHAI W X, YAN H Q, et al.Prediction of PM2.5 hour concentration based on U-net neural network[J]. Acta scientiarum naturalium universitatis pekinensis, 2020, 56(5): 796-804.
[7] 宋飞扬, 铁治欣, 黄泽华, 等. 基于KNN-LSTM的PM2.5浓度预测模型[J]. 计算机系统应用, 2020, 29(7): 193-198.
SONG F Y, TIE Z X, HUANG Z H, et al.PM2.5 concentration prediction model based on KNN-LSTM[J]. Computer systems & applications, 2020, 29(7): 193-198.
[8] 肖云鹏, 王锡凡, 王秀丽, 等. 面向高比例可再生能源的电力市场研究综述[J]. 中国电机工程学报, 2018, 38(3): 663-674.
XIAO Y P, WANG X F, WANG X L, et al.Review on electricity market towards high proportion of renewable energy[J]. Proceedings of the CSEE, 2018, 38(3): 663-674.
[9] WANG K J, QI X X, LIU H D.Photovoltaic power forecasting based LSTM-convolutional network[J]. Energy, 2019, 189: 116225.
[10] 盛四清, 金航, 刘长荣. 基于VMD-WSGRU的风电场发电功率中短期及短期预测[J]. 电网技术, 2022, 46(3): 897-904.
SHENG S Q, JIN H, LIU C R.Short-term and mid-short-term wind power forecasting based on VMD-WSGRU[J]. Power system technology, 2022, 46(3): 897-904.
[11] 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 C, ZHANG Z B.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.
[12] 连子宽, 姚力, 刘晟源, 等. 基于t-SNE降维和BIRCH聚类的单相用户相位及表箱辨识[J]. 电力系统自动化, 2020, 44(8): 176-184.
LIAN Z K, YAO L, LIU S Y, et al.Phase and meter box identification for single-phase users based on t-SNE dimension reduction and BIRCH clustering[J]. Automation of electric power systems, 2020, 44(8): 176-184.
[13] 王东风, 李嘉宇, 黄宇, 等. 风电场风速预测的密度峰值聚类方法研究[J]. 太阳能学报, 2021, 42(12): 110-118.
WANG D F, LI J Y, HUANG Y, et al.Research on density peaks clustering method for wind speed prediction of wind farms[J]. Acta energiae solaris sinica, 2021, 42(12): 110-118.
[14] 姚黄金, 雷霞, 付鑫权, 等. 基于改进自适应密度峰值算法的日负荷曲线聚类分析[J]. 电力系统保护与控制, 2022, 50(3): 121-130.
YAO H J, LEI X, FU X Q, et al.Cluster analysis of daily load curves based on an improved self-adaptive density peak clustering algorithm[J]. Power system protection and control, 2022, 50(3): 121-130.
[15] 姚宏民, 杜欣慧, 秦文萍. 基于密度峰值聚类及GRNN神经网络的光伏发电功率预测方法[J]. 太阳能学报, 2020, 41(9): 184-190.
YAO H M, DU X H, QIN W P.PV power forecasting approach based on density peaks clustering and general regression neural network[J]. Acta energiae solaris sinica, 2020, 41(9): 184-190.
[16] 方鹏, 高亚栋, 潘国兵, 等. 基于LSTM神经网络的中长期光伏电站发电量预测方法研究[J]. 可再生能源, 2022, 40(1): 48-54.
FANG P, GAO Y D, PAN G B, et al.Research on forecasting method of mid-and long-term photovoltaic power generation based on LSTM neural network[J]. Renewable energy resources, 2022, 40(1): 48-54.
[17] 李瑞青. 计及气象因素的超短期光伏电站出力预测研究[D]. 吉林: 东北电力大学, 2019.
LI R Q.Study on output prediction of ultra-short-term photovoltaic power station considering meteorological factors[D]. Jilin: Northeast Dianli University, 2019.
[18] GANDOMAN F H, ABDEL ALEEM S H E, OMAR N, et al. Short-term solar power forecasting considering cloud coverage and ambient temperature variation effects[J]. Renewable energy, 2018, 123: 793-805.
PDF(3032 KB)
