基于机器学习的100 m高度风速短期预报订正方法

高金兵; 曹润东; 于廷照; 姚锦烽; 申彦波

doi:10.19912/j.0254-0096.tynxb.2023-1243

PDF(1702 KB)

太阳能学报 ›› 2025, Vol. 46 ›› Issue (1) : 71-77. DOI: 10.19912/j.0254-0096.tynxb.2023-1243

基于机器学习的100 m高度风速短期预报订正方法

高金兵^1~3, 曹润东^1~3, 于廷照^1,3, 姚锦烽^1,2, 申彦波^1~3

作者信息 +

SHORT-TERM 100-METERS WIND SPEED FORECAST CORRECTION METHOD BASED ON MACHINE LEARNING

Gao Jinbing^1-3, Cao Rundong^1-3, Yu Tingzhao^1,3, Yao Jinfeng^1,2, Shen Yanbo^1-3

Author information +

文章历史 +

摘要

提出一种基于机器学习的先格点订正区域误差、后站点订正局地误差的100 m高度风速短期预报订正方法。首先,基于100 m高度风速格点实况资料,采用深度学习算法对数值天气预报进行格点订正;然后,基于测风塔100 m高度风速观测资料,采用随机森林集成预报订正方法对站点风速预报进行订正;最后,选用内蒙古中部某站点进行分析验证。结果表明,订正前未来84小时风速预报均方根误差（RMSE）为4.25 m/s,格点订正后RMSE降至3.16~3.79 m/s,站点集成订正后RMSE降至2.81 m/s。格点订正后预报误差明显降低,站点集成订正后误差进一步减小。

Abstract

In this paper, a forecast correction method based on machine learning is proposed, which first corrects regional errors with gridded analysis products and then corrects local errors with station observation data. The method utilizes the gridded wind speed analysis products at 100-meter height to correct the results of numerical weather prediction through deep learning algorithms, and then corrects the station wind speed forecast results with a random forest integrated forecast approach based on the observation data of wind speed at 100-meter height from the anemometer tower. An empirical analysis of a meteorological station in central Inner Mongolia shows that after correction with the proposed method, the root mean square error （RMSE） of wind speed prediction for the next 84 hours is significantly reduced from 4.25 m/s to 3.16-3.79 m/s after gridded forecast correction and to 2.81 m/s after station integrated correction. Therefore, the proposed method can effectively reduce forecast errors.

导出引用

高金兵, 曹润东, 于廷照, 姚锦烽, 申彦波. 基于机器学习的100 m高度风速短期预报订正方法[J]. 太阳能学报. 2025, 46(1): 71-77 https://doi.org/10.19912/j.0254-0096.tynxb.2023-1243

Gao Jinbing, Cao Rundong, Yu Tingzhao, Yao Jinfeng, Shen Yanbo. SHORT-TERM 100-METERS WIND SPEED FORECAST CORRECTION METHOD BASED ON MACHINE LEARNING[J]. Acta Energiae Solaris Sinica. 2025, 46(1): 71-77 https://doi.org/10.19912/j.0254-0096.tynxb.2023-1243

中图分类号： TM614

参考文献

[1] 赵泽妮, 云斯宁, 贾凌云, 等. 基于统计模型的短期风能预测方法研究进展[J]. 太阳能学报, 2022, 43(11): 224-234.
ZHAO Z N, YUN S N, JIA L Y, et al.Recent progress in short-term forecasting of wind energy based on statistical models[J]. Acta energiae solaris sinica, 2022, 43(11): 224-234.
[2] 冉靖, 张智刚, 梁志峰, 等. 风电场风速和发电功率预测方法综述[J]. 数理统计与管理, 2020, 39(6): 1045-1059.
RAN J, ZHANG Z G, LIANG Z F, et al.Review of wind speed and wind power prediction methods[J]. Journal of applied statistics and management, 2020, 39(6): 1045-1059.
[3] HU Y L, CHEN L.A nonlinear hybrid wind speed forecasting model using LSTM network, hysteretic ELM and Differential Evolution algorithm[J]. Energy conversion and management, 2018, 173: 123-142.
[4] 姚万业, 黄璞, 姚吉行, 等. 一种基于深度学习的FRS-CLSTM风速预测模型[J]. 太阳能学报, 2020, 41(9): 324-330.
YAO W Y, HUANG P, YAO J X, et al.A FRS-CLSTM wind speed prediction model based on deep learning[J]. Acta energiae solaris sinica, 2020, 41(9): 324-330.
[5] 李大中, 李颖宇. 基于深度学习与误差修正的超短期风电功率预测[J]. 太阳能学报, 2021, 42(12): 200-205.
LI D Z, LI Y Y.Ultra-short term wind power prediction based on deep learning and error correction[J]. Acta energiae solaris sinica, 2021, 42(12): 200-205.
[6] 王华彪, 鲁冠军, 李小勇. 基于卷积和共享权系数长短时记忆网络的空时融合机制在风速预测上的应用研究[J]. 太阳能学报, 2022, 43(11): 156-165.
WANG H B, LU G J, LI X Y.Research on application of spatio-temporal fusion mechanism based on convolution and shared weight long short-term memory network in wind speed prediction[J]. Acta energiae solaris sinica, 2022, 43(11): 156-165.
[7] SHAHID F, ZAMEER A, MUNEEB M.A novel genetic LSTM model for wind power forecast[J]. Energy, 2021, 223: 120069.
[8] 杨国清, 刘世林, 王德意, 等. 基于Attention-GRU风速修正和Stacking的短期风电功率预测[J]. 太阳能学报, 2022, 43(12): 273-281.
YANG G Q, LIU S L, WANG D Y, et al.Short-term wind power forecasting based on Attention-GRU wind speed correction and stacking[J]. Acta energiae solaris sinica, 2022, 43(12): 273-281.
[9] HAN L, CHEN M X, CHEN K K, et al.A deep learning method for bias correction of ECMWF 24-240 h forecasts[J]. Advances in atmospheric sciences, 2021, 38(9): 1444-1459.
[10] CHEN L C, PAPANDREOU G, KOKKINOS I, et al. Semantic image segmentation with deep convolutional nets and fully connected CRFs[J/OL]. arXiv, 2014: arXiv: 1412.7062.
[11] CHEN L C, PAPANDREOU G, KOKKINOS I, et al.DeepLab: semantic image segmentation with deep convolutional nets, atrous convolution, and fully connected CRFs[J]. IEEE transactions on pattern analysis and machine intelligence, 2018, 40(4): 834-848.
[12] CHEN L C, PAPANDREOU G, SCHROFF F, et al. Rethinking atrous convolution for semantic image segmentation[J/OL]. arXiv, 2017: arXiv.1706.05587.
[13] CHEN L C, ZHU Y K, PAPANDREOU G, et al.Encoder-decoder with atrous separable convolution for semantic image segmentation[M]. Lecture Notes in Computer Science. Cham: Springer International Publishing, 2018: 833-851.
[14] DOSOVITSKIY A, BEYER L, KOLESNIKOV A,et al.An Image is worth 16x16 words: transformers for image recognition at scale[C]//International Conference on Learning Representations. Virtual Online Conference,2021.
[15] LIU Z, LIN Y T, CAO Y, et al.Swin transformer: hierarchical vision transformer using shifted windows[C]//2021 IEEE/CVF International Conference on Computer Vision (ICCV). Montreal, QC, Canada, 2021: 10012-10022.
[16] WAN Q, HUANG Z L, LU J C, et al.SeaFormer: squeeze-enhanced axial transformer for mobile semantic segmentation[C]//International Conference on Learning Representations. Kigali, Rwanda, 2023.
[17] 周海, 秦昊, 吉璐莹, 等. 地面气象要素多模式集成预报研究进展[J]. 大气科学学报, 2022, 45(6): 815-825.
ZHOU H, QIN H, JI L Y, et al.Research progresses of multimodel ensemble forecast of surface meteorological elements[J]. Transactions of atmospheric sciences, 2022, 45(6): 815-825.
[18] 江滢, 宋丽莉, 程兴宏. 风电场风速预报集合订正方法的尝试性研究[J]. 资源科学, 2013, 35(3): 673-680.
JIANG Y, SONG L L, CHENG X H.An integrated and revised method of forecasting wind speed for wind farms[J]. Resources science, 2013, 35(3): 673-680.
[19] 徐丽娜, 申彦波, 冯震, 等. 多种风电场风速预报订正方法的适用性研究与集成应用[J]. 干旱区地理, 2022, 45(4): 1114-1124.
XU L N, SHEN Y B, FENG Z, et al.Applicability research and integrated application of various correction methods for wind speed forecast in a wind farm[J]. Arid land geography, 2022, 45(4): 1114-1124.
[20] 许沛华, 陈正洪, 孙延维, 等. 湖北山区复杂地形条件下风电功率预报算法研究[J]. 干旱气象, 2021, 39(3): 524-532.
XU P H, CHEN Z H, SUN Y W, et al.Research on wind power prediction algorithm under complicated terrain in mountainous area of Hubei Province[J]. Journal of arid meteorology, 2021, 39(3): 524-532.
[21] 陈鹤, 周顺武, 熊安元, 等. 河北省风能详查区风速日变化特征[J]. 干旱气象, 2011, 29(3): 343-349.
CHEN H, ZHOU S W, XIONG A Y, et al.Analysis on diurnal variation of wind velocity in Hebei Province[J]. Journal of arid meteorology, 2011, 29(3): 343-349.
[22] 乌日柴胡, 王建捷, 孙靖, 等. 北京山区与平原冬季近地面风的精细观测特征[J]. 气象学报, 2019, 77(6): 1107-1123.
WU R, WANG J J, SUN J, et al.An observational investigation of fine features of near surface winds in winter over Beijing area[J]. Acta meteorologica sinica, 2019, 77(6): 1107-1123.
[23] 朱蓉, 王阳, 向洋, 等. 中国风能资源气候特征和开发潜力研究[J]. 太阳能学报, 2021, 42(6): 409-418.
ZHU R, WANG Y, XIANG Y, et al.Stydy on climate characteristics and development potential of wind energy resources in China[J]. Acta energiae solaris sinica, 2021, 42(6): 409-418.