基于深度学习的气象要素预测方法

马景奕; 刘维成; 闫文君

doi:10.16032/j.issn.1004-4965.2021.018

基于深度学习的气象要素预测方法

doi: 10.16032/j.issn.1004-4965.2021.018

1.
中国气象局气象干部培训学院甘肃分院，甘肃兰州 730020
2.
兰州市中心气象台，甘肃兰州 730020

基金项目:

国家自然科学基金项目 41505036

详细信息

通讯作者:
马景奕，甘肃省人，高级工程师，主要的研究方向：信息技术、人工智能、现代教育技术。Email: elose@126.com

中图分类号: P456
计量
- 文章访问数: 458
- HTML全文浏览量: 45
- PDF下载量: 67
- 被引次数: 0
出版历程
- 收稿日期: 2020-04-10
- 修回日期: 2020-11-18
- 刊出日期: 2021-04-01

METEOROLOGICAL ELEMENTS FORECASTING METHOD BASED ON DEEP LEARNING

1.
China Meteorological Administration Meteorological Cadre Training Institute Gansu Branch, Lanzhou 730020, China
2.
Lanzhou Central Meteorological Station Lanzhou 730020, China

摘要

摘要: 针对气象预测内容繁多且影响因素多样的问题，提出了一种基于长短时记忆（LSTM）的气象预测方法。方法能够对繁杂的气象数据进行自动预处理，提取相应的特征信息。通过神经网络的前向训练、长短时记忆反馈学习，经过多隐藏层地自主训练，对能见度、温度、露点、风速、风向以及压力气象信息实现准确预测。通过实验以及与经典机器学习预测方法的比较，验证了本文方法在气象预测中的有效性，进一步提升了气象预测的准确性，各项预测值的均方检验误差平均值为0.35。
- 气象预测 /
- 深度学习 /
- 神经网络 /
- 长短时记忆
Abstract: Aiming at the problems of various weather forecasting content and various influencing factors, a weather forecasting method based on long short-term memory (LSTM) is proposed. The method can automatically preprocess the multifari-ous meteorological data and extract the corresponding characteristic information. Through the neural network forward training, long and short time memory feedback learning, through the multi-layer neural network autonomous training, to achieve the visibility, temperature, dew point, wind speed, wind direction and pressure meteorological information accu-rate prediction. Through simulation experiments and comparison with classical machine learning prediction methods, we verify the effectiveness of the method proposed in this paper in weather prediction, and further improve the accuracy of weather prediction. The mean square test error of the predicted values is 0.35.
- meteorological forecast /
- deep learning /
- neural network /
- long short term memory

HTML全文

图 1 多层神经网络结构

下载: 全尺寸图片幻灯片

图 2 基于LSTM的深度学习网络

a.长短期存储器单元结构；b.LSTM单元结构。

下载: 全尺寸图片幻灯片

图 3 气象数据地域来源示意图

下载: 全尺寸图片幻灯片

图 4 在小数据集上的温度预测

下载: 全尺寸图片幻灯片

图 5 训练误差随训练次数的变化情况

下载: 全尺寸图片幻灯片

图 6 气象预测值误差随训练次数变化情况

下载: 全尺寸图片幻灯片

图 7 温度预测值与真实情况分布

下载: 全尺寸图片幻灯片

图 8 露点预测值与真实情况分布

下载: 全尺寸图片幻灯片

图 9 不同方法对风速流量的预测分布

下载: 全尺寸图片幻灯片

图 10 两种方法预测值均方检验误差分布图

下载: 全尺寸图片幻灯片

参考文献(32)

[1]	REICHSTEIN M, CAMPS V G, STEVENS B. Deep learning and process understanding for data-driven Earth system science[J]. Nature, 2019, 566: 195-204.
[2]	HAIDAR A, VERMA B. Monthly rainfall forecasting using one-dimensional deep convolutional neural network[J]. IEEE Access 2018, 6: 69 053-69 063.
[3]	程文聪, 史小康, 张文军, 等. 基于深度学习的数值模式降水产品降尺度方法[J]. 热带气象学报, 2020, 36(3): 307-316.
[4]	周满国, 黄艳国, 杨训根. 基于GRU神经网络与灰色模型集成的气温预报[J]. 热带气象学报, 2020, 36(6): 855-864.
[5]	谢超, 马学款, 张恒德. 华南低能见度天气特征及客观预报研究[J]. 气象科学, 2019, 39(4): 556-561.
[6]	MAQSOOD I, KHAN M, ABRAHAM A. An ensemble of neural networks for weather forecasting[J]. Neural Comput Appl, 2004, 13: 112-122.
[7]	SHI X, CHEN Z, WANG H, et al. Convolutional LSTM network: A machine learning approach for precipitation nowcasting[C]//Proceedings of the 28th International Conference on Neural Information Processing Systems, Montreal, QC, Canada, 7-12 December 2015.
[8]	SCHER S. Toward data-driven weather and climate forecasting: approximating a simple general circulation model with deep learning[J]. Geophys Res Lett, 2018, 45: 12 616-12 622.
[9]	CANDY B, SAUNDERS R W, GHENT D, et al. The impact of satellite-derived land surface temperatures on numerical weather prediction analyses and forecasts[J]. J Geophys Res, 2017, 122: 9 783-9 802.
[10]	PATHAK J, WIKNER A, FUSSELL R. Hybrid forecasting of chaotic processes: Using machine learning in conjunction with a knowledge- based model[C]. Chaos, 2018, 28: 041101.
[11]	ABRAHART R J, ANCTIL F, COULIBALY P, et al. Two decades of anarchy? Emerging themes and outstanding challenges for neural network river forecasting. Prog[J]. Phys Geogr Earth Environ, 2012, 36: 480-513.
[12]	吴进, 董国豪, 李乔深. 基于区域卷积神经网络和光流法的目标跟踪[J]. 电讯技术, 2018, 58(1): 6-12.
[13]	CHEN L, YE L, SINGH V, et al. Determination of input for artificial neural networks for flood forecasting using the copula entropy method [J]. J Hydrol Eng, 2013, 19: 04014021.
[14]	KASIVISWANATHAN K S, SUDHEER K P. Comparison of methods used for quantifying prediction interval in artificial neural network hydrologic models[J]. Model Earth Syst Environ, 2016, 2: 22.
[15]	KASIVISWANATHAN K S, SUDHEER K P. Methods used for quantifying the prediction uncertainty of artificial neural network based hydrologic models. Stoch[J]. Environ Res Risk Assess, 2017, 31: 1 659-1 670.
[16]	袁雷, 程岳, 牛文生, 等. 基于深度学习的跑道前视红外图像轮廓线提取[J]. 电讯技术, 2019, 59(2): 179-184.
[17]	吴进, 闵育, 马思敏, 等. 一种基于CNN与LSTM结合的微表情识别算法[J]. 电讯技术, 2020, 60(1): 1-7.
[18]	HSU K L, GUPTA H V, SOROOSHIAN S. Artificial neural network modeling of rainfall-runoff process[J]. Water Resources Res, 1995, 31: 2 517-2 530.
[19]	LIU J, N K, HU Y, HE Y, et al. Deep neural network modeling for big data weather forecasting, information granularity, big data, and computational intelligence[M] Springer International Publishing: Cham, Switzerland, 2015: 389-408.
[20]	FUENTE A, MERUANE V, MERUANE C. Hydrological early warning system based on a deep learning runoff model coupled with a meteorological forecast[J]. Water, 2019, 11(9): 1 808.
[21]	LU J, HU W, ZHANG X K. Precipitation data assimilation system based on a neural network and case-based reasoning system[J]. Information, 2018, 9(5): 106.
[22]	HERNÁNDEZ E, ANGUIX V S, JULIAN V, et al. Rainfall prediction: A deep learning approach[C]. In International Conference on Hybrid Artifcial Intelligence Systems, Springer, 2016: 151-162.
[23]	SHI X J, SHI Z R, WAN H G, et al. Convolutional LSTM network: A machine learning approach for precipitation nowcasting[J]. In Neural Information Processing Systems, 2015: 802-810.
[24]	HA J H, YONG H L, KIM Y H. Forecasting the precipitation of the next day using deep learning[J]. Korean Inst. Intell. Syst, 2016, 26: 93-98.
[25]	KOLANOWSKI K, ŚWIETLICKA A, KAPELA R, et al. Multisensor data fusion using Elman neural networks[J]. Applied Mathematics and Computation, 2018, 319: 236-244.
[26]	HASSOUN M H. Fundamentals of artificial neural networks[M]. MIT Press: Cambridge, MA, USA, 1995.
[27]	GREFF K, SRIVASTAVA R K, KOUTNÍK J, et al. LSTM: A search space odyssey[J]. IEEE Trans. Neural Netw. Learn. Syst, 2016, 28: 2 222-2 232.
[28]	方春, 孙福振, 李彩虹. 基于深度学习和字符嵌入的细胞穿透肽预测[J]. 计算机仿真, 2019, 36(10): 353-358.
[29]	SCHMIDHUBER J. Deep learning in neural networks: An overview[J]. Neural Netw. 2015, 61: 85-117.
[30]	HOCHREITER S, SCHMIDHUBER J. Long short-term memory[J]. Neural Comput, 1997, 9: 1 735-1 780.
[31]	刘昕玥, 姚建国, 万定生. 基于组合核函数SVM的中小流域流量预测研究[J]. 计算机仿真, 2019, 36(11): 454-457.
[32]	于广宇, 董学平, 王祥民, 等. 弹性网下基于LSTM的分解炉出口温度预测[J/OL]. 系统仿真学报: 1-8[2020-04-04]. http://kns.cnki.net/kcms/detail/11.3092.V.20200318.1918.005.html.