摘要
准确的风速预测对风能资源的充分利用和风电场的经济效益提升具有显著的意义。为提高集合数值预报的风速预报能力,弥补现有深度学习集合预报订正模型对格点预报数据时间特征提取的不足,引入ConvLSTM深度学习模型,对CMA-REPS(中国气象局区域集合预报模式)预测的华北地区近地面10 m风速格点数据进行偏差订正实验,以均方根误差(RMSE)作为评分标准将订正结果与CMA-REPS原始预报数据和Unet深度学习模型方法得到的订正结果进行对比。结果表明,ConvLSTM模型的订正效果相比Unet模型有进一步的提升,经ConvLSTM模型订正后的近地面10 m风速预报数据整体上更趋近于实况数据。
Accurate wind speed prediction is of great significance for the full utilization of wind energy resources and the improvement of the economic benefits of wind farms.To improve the wind speed forecasting capability of ensemble numerical forecasting,this paper introduces the ConvLSTM deep learning model to perform a bias correction test on the grid point data of near-surface 10 m wind speed predicted by CMA-REPS(China Meteorological Administration-Regional Ensemble Prediction System),and the root mean square error(RMSE)was used as the scoring criterion to compare the correction results with the original prediction data of CMA-REPS and the correction results obtained by the Unet deep learning model method.The results show that the correction effect of the ConvLSTM model can be further improved than that of the Unet model,and the prediction data of 10 m wind speed near the surface after the modification of the ConvLSTM model is closer to the real data.
作者
毛波
杨昊
周世杰
杨康权
陈敏
MAO Bo;YANG Hao;ZHOU Shijie;YANG Kangquan;CHEN Min(Department of Computer Science,Chengdu University of Information Technology,Chengdu 610225,China;School of Information and Software Engineering,University of Electronic Science and Technology of China,Chengdu 610054,China;Sichuan Meteorological Observatory,Chengdu 610072,China;Key Laboratory of Sichuan Meteorological Bureau,Chengdu 610072,China)
出处
《成都信息工程大学学报》
2023年第3期264-270,共7页
Journal of Chengdu University of Information Technology
基金
国家重点研发计划资助项目(2021YFC3000902)
四川省科技计划重点研发专项资助项目(2022YFS0542)。
