《计算机应用》唯一官方网站 ›› 2023, Vol. 43 ›› Issue (1): 321-328.DOI: 10.11772/j.issn.1001-9081.2021111888

• 前沿与综合应用 • 上一篇    

联合MOD11A1和地面气象站点数据的多站点温度预测深度学习模型

张军1,2, 吴朋莉1,2, 石陆魁1,2, 史进1, 潘斌3   

  1. 1.河北工业大学 人工智能与数据科学学院,天津 300401
    2.河北省大数据计算重点实验室(河北工业大学),天津 300401
    3.南开大学 统计与数据科学学院,天津 300071
  • 收稿日期:2021-11-08 修回日期:2022-05-12 发布日期:2023-01-12
  • 通讯作者: 史进(1981—),男,河北张家口人,助理研究员,硕士,主要研究方向:人工智能、数据挖掘 Email:shijin@hebut.edu.cn
  • 作者简介:张军(1976—),男,河北张家口人,副教授,博士,CCF会员,主要研究方向:机器学习、智能计算;吴朋莉(1998—),女,河南商丘人,硕士研究生,主要研究方向:机器学习、智能计算;石陆魁(1974—),男,河北邯郸人,教授,博士,CCF会员,主要研究方向:机器学习、数据挖掘;潘斌(1990—),男,山东烟台人,副教授,博士,主要研究方向:机器学习、遥感图像处理、多目标优化;
  • 基金资助:
    国家自然科学基金资助项目(62001252);河北省自然科学基金资助项目(F2020202008);河北省教育厅科学技术研究项目(ZD2021311)。

Deep learning model for multi-station temperature prediction combined with MOD11A1 and surface meteorological station data

ZHANG Jun1,2, WU Pengli1,2, SHI Lukui1,2, SHI Jin1, PAN Bin3   

  1. 1.School of Artificial Intelligence, Hebei University of Technology, Tianjin 300401, China
    2.Hebei Province Key Laboratory of Big Data Calculation (Hebei University of Technology), Tianjin 300401, China
    3.School of Statistics and Data Science, Nankai University, Tianjin 300071, China
  • Received:2021-11-08 Revised:2022-05-12 Online:2023-01-12
  • Contact: SHI Jin, born in 1981, M. S., research assistant. His research interests include artificial intelligence, data mining.
  • About author:ZHANG Jun, born in 1976, Ph. D., associate professor. His research interests include machine learning, intelligent computing;WU Pengli, born in 1998, M. S. candidate. Her research interests include machine learning, intelligent computing;SHI Lukui, born in 1974, Ph. D., professor. His research interests include machine learning, data mining;PAN Bin, born in 1990, Ph. D., associate professor. His research interests include machine learning, remote sensing image processing, multi-objective optimization;
  • Supported by:
    This work is partially supported by National Natural Science Foundation of China (62001252), Natural Science Foundation of Hebei Province (F2020202008), Science and Technology Project of Hebei Education Department (ZD2021311).

摘要: 针对地面气象站点分布稀疏影响站点间关系以及站点间的关系强度推理难的问题,提出一种基于联合MOD11A1和地面气象站点数据的多站点温度预测深度学习模型(GDM)。GDM包括时空注意力(TSA)、双向图神经长短期记忆(DG-LSTM)网络编码和边-点转换双向门控循环网络解码(EN-GRU)模块。首先使用TSA模块提取MOD11A1图像特征并形成多个虚拟气象站点的温度时间序列,缓解地面气象站点分布稀疏对站点间关系的影响;然后用DG-LSTM编码器通过融合两组温度时间序列来计算地面气象站点间和虚拟气象站点间的关系强度;最后用EN-GRU解码器通过结合站点间的关系强度对地面气象站点的温度时间序列关系进行建模。实验结果表明,相较于二维卷积神经网络(2D-CNN)、长短期记忆全连接网络(LSTM-FC)、长短期记忆神经网络扩展网络(LSTME)和长短记忆与自适应提升集成网络(LSTM-AdaBoost),GDM在10个地面气象站点24 h内温度预测的平均绝对误差(MAE)分别减小0.383 ℃、0.184 ℃、0.178 ℃和0.164 ℃,能提高未来24 h多个气象站点温度的预测精度。

关键词: 温度预测, 注意力机制, 深度学习, 长短期记忆网络, 门控循环单元, 图神经网络, MOD11A1, 地面气象站点

Abstract: Focusing on the issues that the relationships between the stations are affected by the sparse distribution of surface meteorological stations and it is difficult to infer the strengths of relationships between the stations, a Deep learning Model for multi-station temperature prediction combined with MOD11A1 and surface meteorological station data was proposed, namely GDM, which included Spatio-Temporal Attention (TSA) , Double Graph neural Long Short-Term Memory (DG-LSTM) network encoding and Edge-Node transform Gated Recurrent Unit (EN-GRU) decoding modules. Firstly, TSA module was utilized to extract MOD11A1 image features and form the temperature time series of multiple virtual meteorological stations, so as to alleviate the impact of sparse distribution of surface meteorological stations on the relationships between the stations. Secondly, DG-LSTM encoder was used to calculate the strengths of the relationships among surface meteorological stations and virtual meteorological stations via fusing two sets of temperature time series. Finally, EN-GRU decoder was adopted to model the temperature time series relationships between surface meteorological stations through combining the inter-station relationship strengths. Experimental results show that compared with 2-Dimensional Convolutional Neural Network (2D-CNN), Long Short-Term Memory-Fully Connected network (LSTM-FC), Long Short-Term Memory neural network Extended (LSTME) and Long Short-Term Memory and AdaBoost network (LSTM-AdaBoost), GDM has the Average Absolute Error (MAE) of temperature prediction in 24 hours at 10 surface meteorological stations reduced by 0.383 ℃, 0.184 ℃, 0.178 ℃ and 0.164 ℃ respectively. It can be seen that GDM can improve the prediction accuracy of the temperature for meteorological stations in the next 24 hours.

Key words: temperature prediction, attention mechanism, deep learning, Long Short-Term Memory (LSTM) network, Gated Recurrent Unit (GRU), Graph Neural Network (GNN), MOD11A1, surface meteorological station

中图分类号: