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

doi:10.19595/j.cnki.1000-6753.tces.191715

摘要
图/表
参考文献
相关文章 (15)

全文: PDF (2705 KB) HTML
输出: BibTeX | EndNote (RIS)

摘要在传统风电场风速预测中,往往将风电场视作一个整体,进而给出较大空间尺度下的单一整体风速预测结果。但在实际风电场中,多台风机分布在不同的地理位置上,其各自的本地风速也存在明显差异。对此,该文提出一种通过学习历史风速和历史风向中空间相关性和时间相关性,从而实现不同地理位置上多风机各自本地风速的预测方法。该方法首先利用卷积神经网络（CNN）提取空间相关性信息,然后将蕴含空间信息的特征序列交由简单循环单元（SRU）处理,进而学习时间相关性信息。风速和三角函数化的风向构成类似于RGB图像结构的三维矩阵。CNN非常适合处理RGB图像类型的数据,故使用CNN提取多风机的风速和风向在同一个时刻下的空间信息。与其他循环神经网络相比,由于SRU可大幅减小计算代价,故使用SRU提取多风机的风速和风向随时间变化的动态信息。在实际风电场数据上的仿真测试表明：在多个不同的预测时域下,该方法与现有利用时空相关性的预测方法相比,不仅显著减小了计算代价,而且提高了预测的准确性。

	服务

	把本文推荐给朋友
	加入我的书架
	加入引用管理器
	E-mail Alert
	RSS
	作者相关文章
	王晨
	寇鹏

关键词 ：多风机风速预测, 卷积神经网络, 简单循环单元, 改进循环神经网络, 时空相关性

Abstract：Conventional wind speed forecasts focused on the overall wind speed of a wind farm. However, in an actual wind farm, multiple wind turbines are widely distributed in different geographical locations, and their local wind speeds are significantly different. To address this, this paper proposes a method for the turbine-specific wind speed forecasts in a wind farm. This method learns the historical spatial and temporal correlation between wind speed and wind direction. Firstly, the convolutional neural network (CNN) is used to extract the spatial correlation information among multiple wind turbines. Subsequently, the extracted spatial information is processed by the simple recurrent unit (SRU), which learns the temporal correlation information. During the data preprocess procedure, wind speed and trigonometric wind direction form a three-dimensional matrix, which is similar to the RGB image series. CNN is very suitable for processing data of RGB image type, so the spatial information of wind speed and wind direction of multiple wind turbines at the same moment is extracted by CNN. Compared with other recurrent neural network (RNN), SRU has much lower computational cost. So SRU is utilized to extract the dynamic information of wind speed and wind direction of multiple wind turbines over time. The simulation results on actual wind farm data validate the effectiveness of the proposed method.

Key words： Wind speed forecast of multiple wind turbines convolutional neural network simple recurrent unit improved recurrent neural network spatio-temporal correlation

收稿日期: 2019-12-08

PACS:

TM614

基金资助:国家自然科学基金（51777162）和中央高校基本科研业务费专项资金（xzy012019022）资助项目

通讯作者: 寇鹏男,1983年生,副教授,博士生导师,研究方向为新能源发电与智能微电网、电机先进控制、预测控制与智能控制等。E-mail：koupeng@mail.xjtu.edu.cn

作者简介: 王晨男,1995年生,硕士研究生,研究方向人工智能技术在电力系统中的应用。E-mail：wangchenedu@stu.xjtu.edu.cn

引用本文:

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. Wang Chen, Kou Peng. Wind Speed Forecasts of Multiple Wind Turbines in a Wind Farm Based on Integration Model Built by Convolutional Neural Network and Simple Recurrent Unit. Transactions of China Electrotechnical Society, 2020, 35(13): 2723-2735.

链接本文:

https://dgjsxb.ces-transaction.com/CN/10.19595/j.cnki.1000-6753.tces.191715 https://dgjsxb.ces-transaction.com/CN/Y2020/V35/I13/2723

[1] World Wind Energy Association. Wind power capacity worldwide reaches 597 GW, 50.1 GW added in 2018 [EB/OL]. [2019-07-04]. http://wwindea.org/ blog/2019/02/25/wind-power-capacity-worldwide-reaches-600-gw-539-gw-added-in-2018.
[2] 林旻威, 温步瀛. 大规模风电接入对电力系统暂态稳定性影响研究综述[J]. 电气技术, 2017, 20(4): 1-8, 38.
Lin Menwei, Wen Buying.The overview of influence of large scale wind generation on transient stability of power system[J]. Electrical Engineering, 2017, 20(4): 1-8, 38.
[3] 唐程辉, 张凡, 张宁, 等. 基于风电场总功率条件分布的电力系统经济调度二次规划方法[J]. 电工技术学报, 2019, 34(10): 2069-2078.
Tang Chenghui, Zhang Fan, Zhang Ning, et al.Quadratic programming for power system economic dispatch based on the conditional probability distribution of wind farms sum power[J]. Transactions of China Electrotechnical Society, 2019, 34(10): 2069-2078.
[4] 穆钢, 蔡婷婷, 严干贵, 等. 双馈风电机组参与持续调频的双向功率约束及其影响[J]. 电工技术学报, 2019, 34(8): 1750-1759.
Mu Gang, Cai Tingting, Yan Gangui, et al.Bidirectional power constraints and influence of doubly fed induction generator participating in continuous frequency regulation[J]. Transactions of China Electrotechnical Society, 2019, 34(8): 1750-175.
[5] 关中杰, 鲁效平, 李钢强, 等. 基于风速模型的风电机组动态转矩前馈控制技术[J]. 电工技术学报, 2018, 33(22): 5338-5345.
Guan Zhongjie, Lu Xiaoping, Li Gangqiang, et al.Dynamic torque feed forward control technology of wind turbine based on wind speed model[J]. Transactions of China Electrotechnical Society, 2018, 33(22): 5338-5345.
[6] 李军, 杜雪. 稀疏高斯过程在短期风电功率概率预测中的应用[J]. 电机与控制学报, 2019, 23(8): 67-77.
Li Jun, Du Xue.Application of sparse gaussian process in short-term wind power probability prediction[J]. Electric Machines and Control, 2019, 23(8): 67-77.
[7] Kou Peng, Liang Deliang, Yu Linbo, et al.Nonlinear model predictive control of wind farm for system frequency support[J]. IEEE Transactions on Power Systems, 2019, 34(5): 3547-3561.
[8] 沈小军, 付雪姣, 周冲成, 等. 风电机组风速-功率异常运行数据特征及清洗方法[J]. 电工技术学报, 2018, 33(14): 3353-3361.
Shen Xiaojun, Fu Xuejiao, Zhou Chongcheng, et al.Characteristics of outliers in wind speed-power operation data of wind turbines and its cleaning method[J]. Transactions of China Electrotechnical Society, 2018, 33(14): 3353-3361.
[9] 杨正瓴, 刘阳, 张泽, 等. 采用最近历史观测值和PLSR进行空间相关性超短期风速预测[J]. 电网技术, 2017, 41(6): 1816-1823.
Yang Zhengling, Liu Yang, Zhang Ze, et al.Ultra-short-term wind speed prediction with spatial correlation using recent historical observations and PLSR[J]. Power System Technology, 2017, 41(6): 1816-1823.
[10] 殷豪, 董朕, 陈云龙. 基于CEEMD 和膜计算优化支持向量机的风速预测[J]. 电力系统保护与控制, 2017, 45(21): 27-34.
Yin Hao, Dong Zhen, Chen Yunlong.Wind speed forecasting based on complementary ensemble empirical mode decomposition and support vector regression optimized by membrane computing optimization[J]. Power System Protection and Control, 2017, 45(21): 27-34.
[11] 叶瑞丽, 郭志忠, 刘瑞叶, 等. 基于小波包分解和改进Elman神经网络的风电场风速和风电功率预测[J]. 电工技术学报, 2017, 32(21): 103-111.
Ye Ruili, Guo Zhizhong, Liu Ruiye, et al.Wind speed and wind power forecasting method based on wavelet packet decomposition and improved Elman neural network[J]. Transactions of China Electrotechnical Society, 2017, 32(21): 103-111.
[12] 林洁, 吴布托, 陈伟. 基于深层玻尔兹曼机的风电场异常风速值自适应检测预处理方法[J]. 电工技术学报, 2018, 33(增刊1): 205-212.
Lin Jie, Wu Butuo, Chen Wei.Adaptive detection and preprocessing method for abnormal wind speed of wind farm based on deep Boltzmann machine[J]. Transactions of China Electrotechnical Society, 2018, 33(S1): 205-212.
[13] Wang Huaizhi, Li Gangqiang, Wang Guibing, et al.Deep learning based ensemble approach for probabilistic wind power forecasting[J]. Applied Energy, 2017, 188(188): 56-70.
[14] 鲁迪, 王星华, 贺小平. 基于混合粒子群算法和多分位鲁棒极限学习机的短期风速预测方法[J]. 电力系统保护与控制, 2019, 47(5): 115-122.
Lu Di, Wang Xinghua, He Xiaoping.Hybrid population particle algorithm and multi-quantile robust extreme learning machine based short-term wind speed forecasting[J]. Power System Protection and Control, 2019, 47(5): 115-122.
[15] 朱霄珣, 徐搏超, 焦宏超, 等. 遗传算法对SVR风速预测模型的多参数优化[J]. 电机与控制学报, 2017, 21(2): 70-75.
Zhu Xiaoxun, Xu Bochao, Jiao Hongchao, et al.Wind speed prediction method based on SVR and multi-parameter optimization of GA[J]. Electric Machines and Control, 2017, 21(2): 70-75.
[16] 杨正瓴, 吴炳卫, 赵强, 等. 我国冬季风路径上风速与其影响因子的优化延迟相关性分析[J]. 电力系统保护与控制, 2018, 46(19): 57-64.
Yang Zhengling, Wu Bingwei, Zhao Qiang, et al.Analysis of the optimal lag time of wind speeds and their influence factors on the main path of China winter monsoon[J]. Power System Protection and Control, 2018, 46(19): 57-64.
[17] Wang Huaizhi, Wang Guibing, Li Gangqiang, et al.Deep belief network based deterministic and probabilistic wind speed forecasting approach[J]. Applied Energy, 2016, 182(8): 80-93.
[18] Khodayar M, Kaynak O, Khodayar M E.Rough deep neural architecture for short-term wind speed forecasting[J]. IEEE Transactions on Industrial Informatics, 2017, 13(6): 2770-2779.
[19] 杨茂, 周宜. 计及风电场状态的风电功率超短期预测[J]. 中国电机工程学报, 2019, 39(5): 1259-1268.
Yang Mao, Zhou Yi.Ultra-short-term prediction of wind power considering wind farm status[J]. Proceedings of the CSEE, 2019, 39(5): 1259-1268.
[20] 薛禹胜, 陈宁, 王树民, 等. 关于利用空间相关性预测风速的评述[J]. 电力系统自动化, 2017, 41(10): 161-169.
Xue Yusheng, Chen Ning, Wang Shumin, et al.Review on wind speed prediction based on spatial correlation[J]. Automation of Electric Power Systems, 2017, 41(10): 161-169.
[21] 陈宁, 薛禹胜, 丁杰, 等. 利用空间相关性的超短期风速预测[J]. 电力系统自动化, 2017, 41(12): 124-130.
Chen Ning, Xue Yusheng, Ding Jie, et al.Ultra-short term wind speed prediction using spatial correlation[J]. Automation of Electric Power System, 2017, 41(12): 124-130.
[22] Chen Niya, Qian Zheng, Nabney I T, et al.Wind power forecasts using gaussian processes and numerical weather prediction[J]. IEEE Transactions on Power Systems, 2014, 29(2): 656-665.
[23] Wu Wenzu, Chen Kunjin, Qiao Ying, et al.Probabilistic short-term wind power forecasting based on deep neural networks[C]//2016 International Conference on Probabilistic Methods Applied to Power Systems, Beijing, China, 2016: 1-8.
[24] 张文, 冯洋, 刘群. 基于简单循环单元的深层神经网络机器翻译模型[J]. 中文信息学报, 2018, 32(10): 36-44.
Zhang Wen, Feng Yang, Liu Qun.Deep neural machine translation model based on simple recurrent units[J]. Journal of Chinese Information Processing, 2018, 32(10): 36-44.
[25] Tao Lei, Zhang Yu.Training RNNs as fast as CNNs[J]. arXiv preprint, 2017.
[26] Lecun Y, Bottou L, Bengio Y, et al.Gradient-based learning applied to document recognition[J]. Proceedings of the IEEE, 1998, 86(11): 2278-2324.
[27] Liu Tongtong, Wang Rubing, Chen Jikang, et al.Fine-grained classification of product images based on convolutional neural networks[J]. Advances in Molecular Imaging, 2018, 8: 69-87.
[28] He Kaiming, Zhang Xiangyu, Ren Shaoqing, et al.Spatial pyramid pooling in deep convolutional networks for visual recognition[J]. IEEE Transactions on Pattern Analysis & Machine Intelligence, 2014, 37(9): 1904-1916.
[29] Saratha S, Tajuddin W A W A. Logic learning in hopfield networks[J]. Modern Applied Science, 2008, 2(3): 57-63.
[30] Jin Hui, Yang Libin, Gan Jiatian, et al.Risk assessment of grid frequency deviation related to correlation considering wind speed and wind direction in a wind farm[C]//2018 2nd IEEE Conference on Energy Internet and Energy System Integration, Beijing, China, 2018: 1-5.