针对当前风电功率预测过程中历史信息利用不充分及多维输入权重值固定忽略了不同时间维度的特征重要性的问题,提出一种基于特征变权的风电功率预测模型。该方法利用随机森林（RF）分析不同高度处的风速、风向、温度等气象特征对风电输出功率的影响程度,并利用累积贡献率完成气象特征的提取。对提取的特征及历史功率信息利用奇异谱分析（SSA）去噪,以去噪后的数据作为输入建立级联式FA-CNN-LSTM多变量预测模型对超短期风电功率进行预测。通过在CNN-LSTM网络中增加特征注意力机制（FA）自适应挖掘不同时刻的特征关系,动态调整不同时间维度各输入特征的权重,加强预测时刻关键特征的注意力,从而提升预测性能。基于某风电场实测数据的算例分析表明,所提方法可有效提高超短期风电功率预测精度。
Abstract
Aiming at the problems of insufficient utilization of historical information and the fixed multi-dimensional input weight ignoring the importance of features in different time dimensions in current wind power prediction process, a wind power prediction model based on feature variable weight is proposed. Random forest （RF） is used to analyze the degree of influence of wind speed, wind direction, temperature and other meteorological characteristics at different heights on the wind power and cumulative contribution rate is used to complete the extraction of meteorological features. Singular spectrum analysis （SSA） is used to denoise the extracted features and historical power information, and the denoised data is used as input to establish a cascaded FA-CNN-LSTM multivariate prediction model to predict ultra-short-term wind power. By adding feature attention mechanism （FA） to CNN-LSTM network to adaptively mine feature relationships at different time, the weights of input features at different time dimensions can be dynamically adjusted to enhance the attention of key features at prediction moment, and the prediction performance can be improved. The case study shows that the proposed method can effectively improve the accuracy of ultra-short-term wind power prediction.
关键词
风电功率预测 /
长短期记忆 /
随机森林 /
奇异谱分析 /
注意力机制
Key words
wind power prediction /
long short-term memory /
random forest /
singular spectrum analysis /
attention mechanism
{{custom_sec.title}}
{{custom_sec.title}}
{{custom_sec.content}}
参考文献
[1] YANG L, HE M, ZHANG J S, et al.Support-vector-machine-enhanced Markov model for short-term wind power forecast[J]. IEEE transactions on sustainable energy, 2015, 6(3): 791-799.
[2] 殷豪, 欧祖宏, 陈德, 等. 基于二次模式分解和级联式深度学习的超短期风电功率预测[J]. 电网技术, 2020,44(2): 445-453.
YIN H, OU Z H, CHEN D, et al.Ultra-short-term wind power prediction based on two-layer mode decomposition and cascaded deep learning[J]. Power system technology, 2020, 44(2): 445-453.
[3] 卢继平, 曾燕婷, 喻华, 等. 基于改进AWNN的风电功率超短期多步预测[J]. 太阳能学报, 2021, 24(1): 166-173.
LU J P, ZENG Y T, YU H, et al.Ultra-short-term wind power multi-step forecasting based on improve AWNN[J]. Acta energiae solaris sinica, 2021, 24(1): 166-173.
[4] 钱政, 裴岩, 曹利宵, 等. 风电功率预测方法综述[J]. 高电压术, 2016, 42(4): 1047-1060.
QIAN Z, PEI Y, CAO L X, et al.Review of wind power forecasting method[J]. High voltage engineering, 2016, 42(4): 1047-1060.
[5] 杨茂, 张罗宾. 基于数据驱动的超短期风电功率预测综述[J]. 电力系统保护与控制, 2019, 47(13): 171-186.
YANG M, ZHANG L B.Based deterministic and probabilistic wind speed forecasting approach[J]. Power system protection and control, 2016, 182: 80-93.
[6] 李俊卿, 李佳秋. 基于Kriging和长短期记忆网络的风电功率预测方法[J]. 太阳能学报, 2021, 41(11): 241-247.
LI J Q, LI J Q.Wind power prediction method based on Kriging and LSTM network[J]. Acta energiae solaris sinica, 2021, 41(11): 241-247.
[7] 薛阳, 王琳, 王舒, 等. 一种结合CNN和GRU网络的超短期风电预测模型[J]. 可再生能源, 2019, 37(3): 456-462.
XUE Y, WANG L, WANG S, et al.An ultra-short-term wind power forecasting model combined with CNN and GRU networks[J]. Renewable energy resources, 2019, 37(3): 456-462.
[8] HAGHI H V, LOTFIFARD S, QU Z H.Multivariate predictive analytics of wind power data for robust control of energy storage[J]. IEEE transactions on industrial informatics, 2016, 12(4): 1350-1360.
[9] 钱勇生, 邵洁, 季欣欣, 等. 基于LSTM-Attention网络的短期风电功率预测[J]. 电机与控制应用, 2019, 46(9): 95-100.
QIAN Y S, SHAO J, JI X X, et al.Short-term wind power prediction based on LSTM-attention network[J]. Motor and control applications, 2019, 46(9): 95-100.
[10] 杨茂, 白玉莹. 基于多位置NWP和门控循环单元的风电功率超短期预测[J]. 电力系统自动化, 2021, 45(1): 177-183.
YANG M, BAI Y Y.Ultra-short-term prediction of wind power based on multi-location numerical weather prediction and gated recurrent unit[J]. Automation of electric power systems, 2021, 45(1): 177-183.
[11] AFSHAR K, BIGDELI N.Data analysis and short term load forecasting in Iran electricity market using singular spectral analysis(SSA)[J]. Energy, 2011, 36: 2620-2627.
[12] LIN Z F, CHENG L L, HUANG G H.Electricity consumption prediction based on LSTM with attention mechanism[J]. IEEE transactions on electrical and electronic engineering, 2020, 15(4): 556-562.
[13] LIN X, JING X L, HU A J, et al.Deterministic and probabilistic multi-step forecasting for short-term wind speed based on secondary decomposition and a deep learning method[J]. Energy conversion and management, 2020, 220: 0196-8014.
[14] 梁志峰, 王峥, 冯双磊, 等. 基于波动规律挖掘的风电功率超短期预测方法[J]. 电网技术, 2020, 44(11): 4096-4104.
LIANG Z F, WANG Z, FENG S L, et al.Ultra-short-term forecasting method of wind power based on fluctuation law mining[J]. Power system technology, 2020, 44(11): 4096-4104.
基金
辽宁省中央引导地方科技发展资金计划（2021JH6/10500166）; 辽宁省“兴辽英才计划”（XLYC1802041）
PDF(1741 KB)
