基于K-means++和混合深度学习的光伏功率预测

Abstract
Figure/Table
References
Related Citation (5)

Download: PDF (1242 KB) HTML (1 KB)
Export: BibTeX | EndNote (RIS)

Abstract The output of photovoltaic power generation shows strong volatility, which affects the dispatching management of power system. In this paper, a photovoltaic power prediction model using the hybrid of K-means++ and convolutional neural network and long short-term memory network (K- CNN-LSTM) is proposed. Firstly, the historical data set is classified by K-means++, then the appropriate data set is selected as the training set. Secondly, the LSTM prediction model with historical power as input is built to obtain the predicted power value to be modified. Then, the nonlinear relationship between meteorological parameters and photovoltaic power is mined by CNN, which is applied to obtain the correction coefficient for predicted power to improve the prediction accuracy. Finally, the random fluctuation of the input parameters is predicted for multiple times based on the point prediction model and the prediction error set is obtained to achieve interval prediction. Through the data set of photo- voltaic power station of Australian Desert Knowledge Solar Energy Center (DKASC), LSTM, CNN- LSTM and K-LSTM algorithms are selected for comparison. The results demonstrate that this method has high prediction accuracy and stability, and also achieves accurate output power interval prediction.

Key words： photovoltaic power prediction K-means clustering convolutional neural network (CNN) long short-term memory (LSTM) network

Received: 19 January 2021

	Service

	E-mail this article
	Add to my bookshelf
	Add to citation manager
	E-mail Alert
	RSS
	Articles by authors
	CHEN Zhenxiang
	LIN Peijie
	CHENG Shuying
	CHEN Zhicong
	WU Lijun

Cite this article:

CHEN Zhenxiang,LIN Peijie,CHENG Shuying等. Photovoltaic power prediction based on K-means++ and hybrid deep learning[J]. Electrical Engineering, 2021, 22(9): 7-13.

URL:

http://dqjs.cesmedia.cn/EN/Y2021/V22/I9/7

[1] CHOUDHARY P, SRIVASTAVA R K.Sustainability perspectives-a review for solar photovoltaic trends and growth opportunities[J]. Journal of Cleaner Production, 2019, 227: 589-612.
[2] 彭周宁, 林培杰, 赖云锋, 等. 基于混合灰色关联分析-广义回归神经网络的光伏电站短期功率预测[J]. 电气技术, 2019, 20(10): 11-18.
[3] 姜雲腾, 李萍. 基于改进粒子群神经网络短期负荷预测[J]. 电气技术, 2018, 19(2): 87-91.
[4] 龚莺飞, 鲁宗相, 乔颖, 等. 光伏功率预测技术[J]. 电力系统自动化, 2016, 40(4):140-151.
[5] 李安寿, 陈琦, 王子才, 等. 光伏发电系统功率预测方法综述[J]. 电气传动, 2016, 46(6): 93-96.
[6] 田剑刚, 张沛, 彭春华, 等. 基于分时长短期记忆神经网络的光伏发电超短期功率预测[J]. 现代电力, 2020, 37(6): 629-638.
[7] 黄亚峰, 何威, 吴光琴, 等. 基于K-means++和LSTM网络的光伏功率预测研究[J]. 电气自动化, 2020, 42(5): 25-27.
[8] 王仕俊, 平常, 薛国斌, 等. 影响光伏功率输出因素的研究与分析[J]. 电气技术, 2018, 19(8): 68-71.
[9] CHEN Biaowei, LIN Peijie, LIN Yaohai, et al.Hour- ahead photovoltaic power forecast using a hybrid GRA-LSTM model based on multivariate meteoro- logical factors and historical power datasets[C]//IOP Conference Series: Earth and Environmental Science, 2020, 431: 012059.
[10] LIN Peijie, PENG Zhouning, LAI Yunfeng, et al.Short-term power prediction for photovoltaic power plants using a hybrid improved Kmeans-GRA-Elman model based on multivariate meteorological factors and historical power datasets[J]. Energy Conversion and Management, 2018, 177: 704-717.
[11] 王晨, 寇鹏. 基于卷积神经网络和简单循环单元集成模型的风电场内多风机风速预测[J]. 电工技术学报, 2020, 35(13): 2723-2735.
[12] AJIT A, ACHARYA K, SAMANTA A.A review of convolutional neural networks[C]//International Con- ference on Emerging Trends in Information Techno- logy and Engineering, 2020, 10: 1-5.
[13] ARTHUR D, VASSILVITSKII S.K-means++: the advantages of careful seeding[C]//Eighteenth Acm- Siam Symposium on Discrete Algorithms, New Orleans, Louisiana, 2007: 1027-1035.
[14] BAE K Y, JANG H S, SUNG D K.Hourly solar irradiance prediction based on support vector machine and its error analysis[J]. IEEE Transactions on Power Systems, 2017, 32(2): 935-945.
[15] 赖昌伟, 黎静华, 陈博, 等. 光伏发电出力预测技术研究综述[J]. 电工技术学报, 2019, 34(6): 1201-1217.
[16] XIONG Rui, LI Linlin, TIAN Jinpeng.Towards a smarter battery management system: a critical review on battery state of health monitoring methods[J]. Journal of Power Sources, 2018, 405: 18-29.
[17] HOCHREITER S, SCHMIDHUBER J.Long short- term memory[J]. Neural Computation, 1997, 9(8): 1735-1780.
[18] 李彦冬, 郝宗波, 雷航. 卷积神经网络研究综述[J]. 计算机应用, 2016, 36(9): 2508-2515.
[19] 薛阳, 张宁, 俞志程, 等. 基于BiLSTM和Bootstrap方法的风电功率区间预测[J]. 可再生能源, 2020, 38(8): 1059-1064.