碳中和背景下多通道特征组合超短期风电功率预测

doi:10.12096/j.2096-4528.pgt.20103

发电技术 ›› 2021, Vol. 42 ›› Issue (1): 60-68.DOI: 10.12096/j.2096-4528.pgt.20103

碳中和背景下多通道特征组合超短期风电功率预测

黄树帮¹^,²(), 陈耀¹^,²(), 金宇清³^,*()

¹ 新疆金风科技股份有限公司, 北京市大兴区 100176
² 江苏金风软件技术有限公司, 江苏省无锡市 214000
³ 河海大学能源与电气学院, 江苏省南京市 211100

收稿日期:2020-10-09 出版日期:2021-02-28 发布日期:2021-03-12
通讯作者: 金宇清
作者简介:金宇清(1980)，男，博士，副教授，研究方向为新能源发电系统的建模与控制，本文通信作者，jyq16@hhu.edu.cn
黄树帮(1979)，男，硕士，高级工程师，研究方向为智能变电站及新能源发电并网技术，huangshubang@goldwind.com.cn
陈耀(1990)，男，硕士，工程师，研究方向为人工智能与功率预测，yao_chen@aliyun.com
基金资助:
国家自然科学基金项目(52077059)

A Multi-channel Feature Combination Model for Ultra-short-term Wind Power Prediction Under Carbon Neutral Background

Shubang HUANG¹^,²(), Yao CHEN¹^,²(), Yuqing JIN³^,*()

¹ Xinjiang Goldwind Science & Technology Co., Ltd., Daxing District, Beijing 100176, China
² Jiangsu Goldwind Software & Technology Co., Ltd., Wuxi 214000, Jiangsu Province, China
³ College of Energy and Electrical Engineering, Hohai University, Nanjing 211100, Jiangsu Province, China

Received:2020-10-09 Published:2021-02-28 Online:2021-03-12
Contact: Yuqing JIN
Supported by:
National Natural Science Foundation of China(52077059)

摘要/Abstract

摘要：

碳中和背景下，风电将成为我国的主导能源之一。随着人工智能技术快速发展，人工神经网络被广泛应用于风力发电功率预测。传统的人工神经网络算法采用固定形式数据集和单一网络结构，限制了整体表达能力，导致超短期风电功率预测由于各种不确定因素造成难以控制的误差。为此，提出一种基于人工神经网络的多通道特征组合模型，用于超短期风电功率预测。首先将数据集进行重新分类，分别输入到3个神经网络，建立3种特征组合形式；再将多通道特征进行拼接融合，并将融合后的特征加入到全连接神经网络中进行功率预测，可消除不同特征之间的干扰，有效学习到长期依赖的数据特征；最后对5个风电场数据进行算法验证。实验结果表明，该方法比单通道模型能够获得更好的预测精度，而且增加了网络稳定性。

关键词: 碳中和, 风力发电, 超短期功率预测, 人工神经网络, 多通道

Abstract:

Wind power will become one of the dominant power sources of China oriented to carbon neutral. With the rapid development of artificial intelligence technology, artificial neural networks are widely used in wind power generation forecasting. Traditional artificial neural network algorithms use fixed-form data sets and simple network structures, which limits the overall expression ability and results in uncontrollable errors in ultra-short-term wind power forecasting due to various uncertain factors. In this work, a multi-channel feature combination model based on artificial neural network for ultra-short-term wind power prediction was proposed. Firstly, the data were reclassified and input into three neural networks to establish three feature combinations. After that, multi-channel features splicing and fusion were performed. The fused features were added to the fully connected neural network for power prediction, which can eliminate the interference between different features and effectively learn long-term dependent data features. Finally, the algorithm was verified on the actual data of five wind farms. The experimental results show that this method has better prediction accuracy than the single-channel model, and can improve the network stability.

Key words: carbon neutral, wind power generation, ultra-short-term power prediction, artificial neural network, multi-channel

中图分类号:

TK89

黄树帮, 陈耀, 金宇清. 碳中和背景下多通道特征组合超短期风电功率预测[J]. 发电技术, 2021, 42(1): 60-68.

Shubang HUANG, Yao CHEN, Yuqing JIN. A Multi-channel Feature Combination Model for Ultra-short-term Wind Power Prediction Under Carbon Neutral Background[J]. Power Generation Technology, 2021, 42(1): 60-68.

图/表 18

图1 卷积神经网络基本模型

Fig.1 Basic model of convolutional neural network

图2 LSTM网络结构

Fig.2 LSTM network structure

图3 多通道特征组合模型

Fig.3 Multi-channel feature combination model

图4 超短期功率预测可用数据形式

Fig.4 Available data forms for ultra-short-term power prediction

图5 膨胀卷积原理图

Fig.5 Schematic diagram of dilated convolution

图6 多通道神经网络改进形式

Fig.6 Improved form of multi-channel neural network

图7 多通道神经网络算法流程图

Fig.7 Algorithmic flowchart of multi-channel neural network

图8 多通道神经网络结构图

Fig.8 Structure of multi-channel neural network

图9 实测功率与实测风速散点图

Fig.9 Scatter diagram of measured power and wind speed

图10 实测功率与实测风速折线图过滤前后对比

Fig.10 Comparison of the measured power and wind speed before and after filtering

表1 江苏地区2种算法的预测结果对比

Tab. 1 Comparison of prediction results of two algorithms in Jiangsu

风电场	误差准确率		平均准确率
风电场	单通道均方根	多通道均方根	单通道调和	多通道调和
江苏A	0.852 1	0.855 3	0.674 4	0.679 8
江苏B	0.814 9	0.820 2	0.640 2	0.643 5

图11 江苏地区2种算法预测值折线对比图

Fig.11 Comparison of the predicted power of two algorithms in Jiangsu

表2 内蒙古地区2种算法的预测结果对比

Tab. 2 Comparison of prediction results of two algorithms in Inner Mongolia

风电场	误差准确率		平均准确率
风电场	单通道均方根	多通道均方根	单通道调和	多通道调和
内蒙C	0.753 6	0.764 6	0.530 3	0.533 8

图12 内蒙古地区2种算法预测值折线对比图

Fig.12 Comparison of the predicted power of two algorithms in Inner Mongolia

表3 云南地区2种算法的预测结果对比

Tab. 3 Comparison of prediction results of two algorithms in Yunnan

风电场	误差准确率		平均准确率
风电场	单通道均方根	多通道均方根	单通道调和	多通道调和
云南D	0.856 2	0.860 3	0.774 7	0.775 9

图13 云南地区2种算法预测值折线对比图

Fig.13 Comparison of the predicted power of two algorithms in Yunnan

表4 甘肃地区2种算法的预测结果对比

Tab. 4 Comparison of prediction results of two algorithms in Gansu

风电场	误差准确率		平均准确率
风电场	单通道均方根	多通道均方根	单通道调和	多通道调和
甘肃E	0.848 2	0.852 2	0.601 5	0.608 9

图14 甘肃地区2种算法预测值折线对比图

Fig.14 Comparison of the predicted power of two algorithms in Gansu

参考文献 25

1	全球风能理事会. 全球风能报告[R]. 全球风能理事会, 2019.
	Global Wind Energy Council.Global wind energy report[R].Global Wind Energy Council, 2019.
2	钱峰, 陈艺, 刘俊磊, 等. 大规模风电接入的电力系统协调控制策略[J]. 广东电力, 2019, 32 (11): 12- 18.
	QIAN F , CHEN Y , LIU J L , et al. Coordinated control strategy for power systems with large-scale wind power integration[J]. Guangdong Electric Power, 2019, 32 (11): 12- 18.
3	施涛, 李春来, 朱慧敏. 考虑功率预测不确定性的风电消纳随机调度[J]. 电网与清洁能源, 2019, 35 (4): 55- 59.
	SHI T , LI C L , ZHU H M . Stochastic dispatching of wind power accommodation considering uncertainty of power prediction[J]. Power System and Clean Energy, 2019, 35 (4): 55- 59.
4	张坤, 张金环, 张巍巍, 等. 基于ASD-KDE的风电出力超短期区间预测[J]. 智慧电力, 2019, 47 (5): 32- 37.
	ZHANG K , ZHANG J H , ZHANG W W , et al. Ultra-short-term wind power interval prediction based on ASD-KDE[J]. Smart Power, 2019, 47 (5): 32- 37.
5	张金环, 王超群, 张彤, 等. 基于高斯混合分布模型的风电功率预测误差统计分析研究[J]. 智慧电力, 2020, 48 (7): 59- 64.
	ZHANG J H , WANG C Q , ZHANG T , et al. Statistical analysis of wind power forecasting errors based on Gaussian mixture model[J]. Smart Power, 2020, 48 (7): 59- 64.
6	赵滨滨, 王莹, 王彬, 等. 基于ARIMA时间序列的分布式光伏系统输出功率预测方法研究[J]. 可再生能源, 2019, 37 (6): 820- 823.
	ZHAO B B , WANG Y , WANG B , et al. Photovoltaic power prediction in distribution network based on ARIMA model time series[J]. Renewable Energy Resources, 2019, 37 (6): 820- 823.
7	田波, 朴在林, 王慧. 基于时间序列建模在风力发电功率短期预测中的研究[J]. 电网与清洁能源, 2016, 32 (3): 115- 119.
	TIAN B , PIAO Z L , WANG H . Short term prediction of wind power based on time series modeling[J]. Power System and Clean Energy, 2016, 32 (3): 115- 119.
8	朱红路, 李旭, 姚建曦, 等. 基于小波分析与神经网络的光伏电站功率预测方法[J]. 太阳能学报, 2015, 36 (11): 2725- 2730.
	ZHU H L , LI X , YAO J X , et al. The power prediction method for photovoltaic power station based on wavelet analysis and neural networks[J]. Acta Energiae Solaris Sinica, 2015, 36 (11): 2725- 2730.
9	潘志刚, 刘三明, 李莹, 等. 基于深度学习网络的风电场功率短期预测研究[J]. 科技与创新, 2015, (19): 4- 6.
	PAN Z G , LIU S M , LI Y , et al. Research on the short-term forecasting of wind power based on deep learning network[J]. Science and Technology & Innovation, 2015, (19): 4- 6.
10	叶興, 薛家祥. 改进型LSTM网络光伏发电功率预测研究[J]. 中国测试, 2019, 45 (11): 14- 20.
	YE Y , XUE J X . Research on photovoltaic power generation prediction based on improved LSTM network[J]. China Measurement & Testing Technology, 2019, 45 (11): 14- 20.
11	王瑞, 刘明山, 周原. 一种基于DBN和卡尔曼滤波算法的AGV定位算法:
	CN201710027637.3[P].2017-05-17. WANG R, LIU M S, ZHOU Y.An AGV location algorithm based on DBN and Kalman filter algorithm: CN201710027637.3[P].2017-05-17.
12	赵怡茗. 基于VMD-SE-LSSVM和迭代误差修正的光伏发电功率预测[D]. 西安: 西安理工大学, 2018.
	ZHAO Y M.PV power prediction based on VMD-SE- LSSVM and iterative error correction[D].Xi'an: Xi'an University of Technology, 2018.
13	CHANG C C , LIN C J . LIBSVM: a library for support vector machines[J]. ACM Transactions on Intelligent Systems & Technology, 2011, 2 (3): 1- 27.
14	杨延西, 刘丁. 基于小波变换和最小二乘支持向量机的短期电力负荷预测[J]. 电网技术, 2005, 29 (13): 60- 64.
	YANG Y X , LIU D . Short-term load forecasting based on wavelet transform and least square support vector machines[J]. Power System Technology, 2005, 29 (13): 60- 64.
15	周志华. 机器学习[J]. 中国民商, 2016, 3 (21): 93.
	ZHOU Z H . Machine learning[J]. China's civil and commercial, 2016, 3 (21): 93.
16	李相俊, 许格健. 基于长短期记忆神经网络的风力发电功率预测方法[J]. 发电技术, 2019, 40 (5): 426- 433.
	LI X J , XU G J . Wind power prediction method based on long short-term memory neural network[J]. Power Generation Technology, 2019, 40 (5): 426- 433.
17	郎伟明, 麻向津, 周博文, 等. 基于LSTM和非参数核密度估计的风电功率概率区间预测[J]. 智慧电力, 2020, 48 (2): 31- 37.
	LANG W M , MA X J , ZHOU B W , et al. Wind power probabilistic intervals prediction based on LSTM and nonparametric kernel density estimation[J]. Smart Power, 2020, 48 (2): 31- 37.
18	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.
	ZHOU F Y , JIN L P , DONG J . Review of convolutional neural network[J]. Chinese Journal of Computers, 2017, 40 (6): 1229- 1251.
19	牛哲文, 余泽远, 李波, 等. 基于深度门控循环单元神经网络的短期风功率预测模型[J]. 电力自动化设备, 2018, 38 (5): 43- 49.
	NIU Z E , YU Z Y , LI B , et al. Short-term wind power forecasting model based on deep gated recurrent unit neural network[J]. Electric Power Automation Equipment, 2018, 38 (5): 43- 49.
20	AHMED A , KHALID M . A review on the selected applications of forecasting models in renewable power systems[J]. Renewable and Sustainable Energy Reviews, 2019, 100, 9- 21. DOI
21	杨正瓴, 冯勇, 熊定方, 等. 基于季风特性改进风电功率预测的研究展望[J]. 智能电网, 2015, 3 (1): 1- 7.
	YANG Z L , FENG Y , XIONG D F . Research prospects of improvement in wind power forecasting based on characteristics of monsoons[J]. Smart Grid, 2015, 3 (1): 1- 7.
22	郭丽丽, 丁世飞. 深度学习研究进展[J]. 计算机科学, 2015, 42 (5): 28- 33.
	GUO L L , DING S F . Research progress of deep learning[J]. Computer Science, 2015, 42 (5): 28- 33.
23	王星, 李超, 陈吉. 基于膨胀卷积神经网络模型的中文分词方法[J]. 中文信息学报, 2019, 33 (9): 24- 30.
	WANG X , LI C , CHEN J . Dilated convolution neural networks for Chinese word segmentation[J]. Journal of Chinese Information Processing, 2019, 33 (9): 24- 30.
24	应稼田, 程良伦, 林锦发. 一种用于风电功率预测的双存储结构LSTM模型[J]. 无线互联科技, 2019, 16 (5): 117- 120.
	YING J T , CHENG L L , LIN J F . A dual storage structure LSTM model for wind power prediction[J]. Wuxian Hulian Keji, 2019, 16 (5): 117- 120.
25	刘彬, 戴桂平. 基于白化检验和3σ准则的小波阈值去噪算法[J]. 传感技术学报, 2005, 18 (3): 473- 476.
	LIU B , DAI G P . Adaptive wavelet thresholding denoising algorithm based on white noise detection and 3σ rule[J]. Chinese Journal of Sensors and Actuators, 2005, 18 (3): 473- 476.

[1]	周丹, 袁至, 李骥, 范玮. 考虑平抑未来时刻风电波动的混合储能系统超前模糊控制策略[J]. 发电技术, 2024, 45(3): 412-422.
[2]	刘洪波, 刘永发, 任阳, 孙黎, 刘珅诚. 高风电渗透率下考虑系统风电备用容量的储能配置[J]. 发电技术, 2024, 45(2): 260-272.
[3]	邵宜祥, 刘剑, 胡丽萍, 过亮, 方渊, 李睿. 一种改进组合神经网络的超短期风速预测方法研究[J]. 发电技术, 2024, 45(2): 323-330.
[4]	王若为, 李音璇, 葛维春, 张诗钽, 刘闯, 楚帅. 沙漠光伏电能外送技术综述[J]. 发电技术, 2024, 45(1): 32-41.
[5]	彭道刚, 税纪钧, 王丹豪, 赵慧荣. “双碳”背景下虚拟电厂研究综述[J]. 发电技术, 2023, 44(5): 602-615.
[6]	张宁, 朱昊, 杨凌霄, 胡存刚. 考虑可再生能源消纳的多能互补虚拟电厂优化调度策略[J]. 发电技术, 2023, 44(5): 625-633.
[7]	刘含笑. 双碳背景下电除尘器的节能减碳分析[J]. 发电技术, 2023, 44(5): 738-744.
[8]	曹冬惠, 杜冬梅, 何青. 氢储能安全及其检测技术综述[J]. 发电技术, 2023, 44(4): 431-442.
[9]	许洪华, 邵桂萍, 鄂春良, 郭金东. 我国未来能源系统及能源转型现实路径研究[J]. 发电技术, 2023, 44(4): 484-491.
[10]	张全斌, 周琼芳. 基于“双碳”目标的中国火力发电技术发展路径研究[J]. 发电技术, 2023, 44(2): 143-154.
[11]	黄宇箴, 陈彦奇, 吴志聪, 徐钢, 刘彤. 碳中和背景下热电联产机组抽汽分配节能优化[J]. 发电技术, 2023, 44(1): 85-93.
[12]	董瑞, 高林, 何松, 杨东泰. CCUS技术对我国电力行业低碳转型的意义与挑战[J]. 发电技术, 2022, 43(4): 523-532.
[13]	余潇, 卜广全, 王姗姗. 风电经柔直孤岛送出交流暂态过电压抑制策略研究[J]. 发电技术, 2022, 43(4): 618-625.
[14]	冯伟忠, 李励. “双碳”目标下煤电机组低碳、零碳和负碳化转型发展路径研究与实践[J]. 发电技术, 2022, 43(3): 452-461.
[15]	高虎, 刘凡, 李海. 碳中和目标下氨燃料的机遇、挑战及应用前景[J]. 发电技术, 2022, 43(3): 462-467.

碳中和背景下多通道特征组合超短期风电功率预测

A Multi-channel Feature Combination Model for Ultra-short-term Wind Power Prediction Under Carbon Neutral Background

RichHTML

PDF (PC)

可视化

摘要/Abstract

引用本文

图/表 18

参考文献 25

相关文章 15

编辑推荐

Metrics