基于SAM-WGAN-GP的短期风电功率预测

黄玲; 李林霞; 程瑜; 许子健

doi:10.19912/j.0254-0096.tynxb.2022-0265

PDF(1732 KB)

太阳能学报 ›› 2023, Vol. 44 ›› Issue (4) : 180-188. DOI: 10.19912/j.0254-0096.tynxb.2022-0265

基于SAM-WGAN-GP的短期风电功率预测

黄玲^1~3, 李林霞¹, 程瑜¹, 许子健¹

作者信息 +

SHORT-TERM WIND POWER PREDICTION BASED ON SAM-WGAN-GP

Huang Ling^1-3, Li Linxia¹, Cheng Yu¹, Xu Zijian¹

Author information +

文章历史 +

摘要

针对风电功率预测精度低且模型不稳定的问题,提出基于双阶段注意力机制生成对抗网络（SAM-WGAN-GP）的短期风电功率预测模型。首先,在生成对抗网络的生成模型中引入自注意力机制和时间注意力机制,通过自注意力机制自适应的选择输入特征,并通过时间注意力机制捕获风电数据时间序列的长时间依赖性;判别模型采用卷积神经网络,提高模型的预测精度。其次,将SAM-WGAN-GP网络的生成器损失函数和均方根误差结合作为目标函数,以提高模型的稳定性,同时为解决判别器缓慢学习的问题,引入双时间尺度更新规则（TTUR）以平衡网络的训练过程。最后,以甘肃省酒泉市某风电场的实际运行数据为例,验证SAM-WGAN-GP模型不仅能自适应选择输入特征,而且可捕捉风电数据的长时间依赖性,并提高预测精度。

Abstract

Aiming at the problem of low prediction accuracy and unstable model of wind power prediction, a short-term wind power prediction model based on dual-stage self-attention mechanism wasserstein generative adversarial networks with gradient penalty （SAM-WGAN-GP） was proposed in this paper. Firstly, self-attention mechanism and time attention mechanism are introduced in the generative model of generative adversarial network to construct a dual-stage attention-based wasserstein generative adversarial networks with Gradient Penalty, the self-attention mechanism is brought into the generator to adaptively select the input features, a time attention mechanism is brought into the generator to capture the long-term dependence of wind power data time series; convolutional neural network is selected as the discriminator to improve the prediction accuracy of the model. Secondly, the generator loss function and root mean square error of SAM-WGAN-GP network are combined as the objective function to improve the stability of the model. Meanwhile, in order to solve the problem of slow learning of discriminators, TTUR is introduced to balance the training process of the network. Finally, the actual operation data of a wind farm in Jiuquan, Gansu Province are used as an example to verify that the SAM-WGAN-GP model. The results indicate that the model can not only adaptively select input features, but also capture the long time dependence of wind power date and improve the prediction accuracy.

导出引用

黄玲, 李林霞, 程瑜, 许子健. 基于SAM-WGAN-GP的短期风电功率预测[J]. 太阳能学报. 2023, 44(4): 180-188 https://doi.org/10.19912/j.0254-0096.tynxb.2022-0265

Huang Ling, Li Linxia, Cheng Yu, Xu Zijian. SHORT-TERM WIND POWER PREDICTION BASED ON SAM-WGAN-GP[J]. Acta Energiae Solaris Sinica. 2023, 44(4): 180-188 https://doi.org/10.19912/j.0254-0096.tynxb.2022-0265

中图分类号： TM614

参考文献

[1] 李伟, 王冰, 曹智杰, 等. 基于混沌理论的鸡群改进算法及其在风电功率区间预测中的应用[J]. 太阳能学报, 2021,42(7): 350-358.
LI W, WANG B, CAO Z J, et al.Improved chicken swarm algorithm based on Chaos Theory and its application in wind power interval prediction[J]. Acta energiae solaris sinica, 2021, 42(7): 350-358.
[2] 杨锡运, 张艳峰, 叶天泽, 等. 基于朴素贝叶斯的风电功率组合概率区间预测[J]. 高电压技术, 2020, 46(3): 1099-1108.
YANG X Y, ZHANG Y F, YE T Z, et al.Combination probability interval prediction of wind power based on Naive Bayes[J]. High voltage engineering, 2020, 46(3): 1099-1108.
[3] 梁志峰, 王铮, 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 prediction method of wind power based on fluctuation law mining[J]. Power system technology, 2020, 44(11): 4096-4104.
[4] 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 quadratic mode decomposition and cascaded deep learning[J]. Power system technology, 2020, 44(2): 445-453.
[5] 刘云凯, 彭显刚, 袁浩亮, 等. 基于VMD与改进QRGRU的超短期风电功率概率预测[J]. 电力工程技术, 2021, 40(3): 72-77.
LIU Y K, PENG X G, YUAN H L, et al.Ultra-short-term wind power probability prediction based on VMD and improved QRGRU[J]. Electric power engineering technology, 2021, 40(3): 72-77.
[6] 姚越, 刘达. 基于注意力机制的卷积神经网络-长短期记忆网络的短期风电功率预测[J]. 现代电力, 2021, 38(6): 1-7.
YAO Y, LIU D.Convolutional neural network-long short-term memory network based on attention mechanism short-term wind power forecasting[J]. Modern electric power, 2021, 38(6): 1-7.
[7] WANG C Y, XU C, YAO X, et al.Evolutionary generative adversarial networks[J]. IEEE transactions on evolutionary computation, 2019, 23(6): 921-934.
[8] 卢锦玲, 张祥国, 张伟, 等. 基于改进辅助分类生成对抗网络的风机主轴承故障诊断[J]. 电力系统自动化, 2021, 45(7): 148-154.
LU J L, ZHANG X G, ZHANG W, et al.Fault diagnosis of wind turbine main bearing based on improved auxiliary classification generative confrontation network[J]. Automation of electric power systems, 2021, 45(7): 148-154.
[9] 肖白, 黄钰茹, 姜卓, 等. 数据匮乏场景下采用生成对抗网络的空间负荷预测方法[J]. 中国电机工程学报, 2020, 40(24): 7990-8001, 8236.
XIAO B, HUANG Y R, JIANG Z, et al.A spatial load prediction method using generative adversarial networks for data scarcity scenarios[J]. Proceedings of the CSEE, 2020, 40(24): 7990-8001, 8236.
[10] 纪旭. 基于组合分析方法的短期风电功率预测研究[D].北京:华北电力大学, 2019.
JI X.Research on short-term wind power forecast based on combined analysis method[D]. Beijing: North China Electric Power University, 2019.
[11] 陈刚, 王印, 单锦宁, 等. 生成对抗网络在风电功率预测中的应用[J]. 辽宁工程技术大学学报(自然科学版), 2021, 40(3): 258-264.
CHEN G, WANG Y, SHAN J N, et al.Application of generative countermeasure network in wind power forecasting[J]. Journal of Liaoning Technical University (natural science edition), 2021, 40(3): 258-264.
[12] WANG K F, GOU C, DUAN Y J, et al.Generative adversarial networks: introduction and outlook[J]. IEEE/CAA journal of automatica sinica, 2017, 4(4): 588-598.
[13] 谭本东, 杨军, 赖秋频, 等. 基于改进CGAN的电力系统暂态稳定评估样本增强方法[J]. 电力系统自动化, 2019, 43(1): 149-157.
TAN B D, YANG J, LAI Q P, et al.Power system transient stability assessment sample enhancement method based on improved CGAN[J]. Automation of electric power systems, 2019, 43(1): 149-157.
[14] KARRAS T, AITTALA M, HELLSTEN J, et al.Training generative adversarial networks with limited data[J]. Advances in neural information processing systems, 2020, 33: 12104-12114.
[15] WANG F, ZHANG Z Y, LIU C, et al.Generative adversarial networks and convolutional neural networks based weather classification model for day ahead short-term photovoltaic power forecasting[J]. Energy conversion and management, 2019, 181: 443-462.
[16] 王越, 王德兴, 袁红春, 等. 基于金字塔注意力机制和生成对抗网络的水下图像增强[J]. 激光与光电子学进展, 2021, 58(16): 209-218.
WANG Y, WANG D X, YUAN H C, et al.Underwater image enhancement based on pyramid attention mechanism and generative confrontation network[J]. Progress in laser and optoelectronics, 2021, 58(16): 209-218.
[17] 石磊, 王毅, 成颖, 等. 自然语言处理中的注意力机制研究综述[J]. 数据分析与知识发现, 2020, 4(5): 1-14.
SHI L, WANG Y, CHENG Y, et al.A survey of attention mechanisms in natural languagep Processing[J]. Data analysis and knowledge discovery, 2020, 4(5): 1-14.
[18] 蒋明峰, 支明豪, 李杨, 等. 基于自注意力机制生成对抗网络的超分辨率磁共振图像重建[J]. 中国科学:信息科学, 2021, 51(6): 959-970.
JIANG M F, ZHI M H, LI Y, et al.Super-resolution MRI reconstruction based on self-attention mechanism to generate confrontation network[J]. Science in China: information science, 2021, 51(6): 959-970.
[19] 孙敏, 李旸, 庄正飞, 等. 基于BGRU和自注意力机制的情感分析[J]. 江汉大学学报(自然科学版), 2020, 48(4): 80-89.
SUN M, LI Y, ZHUANG Z F, et al.Sentiment analysis based on BGRU and self-attention mechanism[J]. Journal of Jianghan University(natural science edition), 2020, 48(4): 80-89.
[20] 阳雨妍, 宋爱国, 沈书馨, 等. 基于CNN-GRU的遥操作机器人操作者识别与自适应速度控制方法[J]. 仪器仪表学报, 2021, 42(3): 123-131.
YANG Y Y, SONG A G, SHEN S X, et al.Operator recognition and adaptive speed control method of teleoperation robot based on CNN-GRU[J]. Chinese journal of scientific instrument, 2021, 42(3): 123-131.
[21] WU Q Y, GUAN F, LYU C, et al.Ultra-short-term multi-step wind power forecasting based on CNN-LSTM[J]. IET renewable power generation, 2021, 15(5): 1019-1029.
[22] ZHEN H, NIU D X, YU M, et al.A hybrid deep learning model and comparison for wind power forecasting considering temporal-spatial feature extraction[J]. Sustainability, 2020, 12(22): 9490.
[23] 罗佳. 基于生成式对抗网络的机械故障诊断方法研究[D]. 太原: 中北大学, 2021.
LUO J.Research on mechanical fault diagnosis method based on generative adversarial networks[D]. Taiyuan: North University of China, 2021.
[24] 王静, 邹慧敏, 曲东东, 等. 基于经验模态分解生成对抗网络的金融时间序列预测[J]. 计算机应用与软件, 2020, 37(5): 293-297.
WANG J, ZOU H M, QU D D, et al.Financial time series prediction based on empirical mode decomposition to generate adversarial networks[J]. Computer applications and software, 2020, 37(5): 293-297.
[25] 赵传, 戴朝华, 付洋, 等. 考虑风电预测误差与系统安全域的风电装机规划多目标优化方法[J]. 太阳能学报, 2020, 41(2): 110-117.
ZHAO C, DAI C H, FU Y, et al.A multi-objective optimization method for wind power installation planning considering wind power forecast errors and system security domains[J]. Acta energiae solaris sinica, 2020, 41(2): 110-117.
[26] 杨子民, 彭小圣, 郎建勋, 等. 基于集群动态划分与BLSTM深度学习的风电集群短期功率预测[J]. 高电压技术, 2021, 47(4): 1195-1203.
YANG Z M, PENG X S, LANG J X, et al.Short-term power prediction of wind power clusters based on cluster dynamic division and BLSTM deep learning[J]. High voltage engineering, 2021, 47(4): 1195-1203.