ZHEN Fudong,ZHOU Shuyue,CHEN Lijing,et al.Research on Short-term Prediction Method of High-altitude Wind Field based on ADS-B[J].Journal of Chengdu University of Information Technology,2022,37(06):642-650.[doi:10.16836/j.cnki.jcuit.2022.06.005]
基于广播式自动相关监视系统的高空风场短时预测方法研究
- Title:
- Research on Short-term Prediction Method of High-altitude Wind Field based on ADS-B
- 文章编号:
- 2096-1618(2022)06-0642-09
- 分类号:
- TP311.13
- 文献标志码:
- A
- 摘要:
- 使用飞机作为天气传感器是空中交通管理和气象研究的最新进展,其数据有利于提高气象信息监测和预报的准确性,这将是中国航空气象领域的重大突破。结合ADS-B和Mode S数据的风场反演结果,更加丰富的数据集使短时风场的预测效果优于传统方法。利用一种新的、多功能、高精度的优化风场数据进行短时高空风场的预测研究。为提高中尺度高空风场短时预报的精度,基于高精度、高覆盖的风场反演数据,研究了高空风场的短时预测算法。通过分析高空风场时间序列的分布特征和变化趋势,建立SARIMA和GPR两种时间序列预测模型,分别讨论两种模型的优缺点及其应用方向,并对模型的预测准确性进行定量分析。结果表明,基于GPR的方法优于基于SARIMA的方法。在可接受的预报时间范围内,基于GPR的预测模型能较好地捕捉高空风场的变化趋势,具有较好的泛化能力。在可用数据较少的情况下,GPR模型的复杂度不太高,时间相对较少,即用较少的时间换取更高的准确性,非常适合高空风场的短时预测。
- Abstract:
- The use of aircraft as weather sensors is the latest development in air traffic management and meteorological research. Aircraft data will help to improve the accuracy of meteorological information monitoring and forecasting, which will be a major breakthrough in the field of aviation meteorology in China. Combined with the wind field inversion results of ADS-B and ModeS data, the richer data set makes the prediction effect of short-term wind field better than the traditional methods. In this paper, a new, multi-functional and high-precision optimized wind field data is used to predict the short-term high-altitude wind field. In order to improve the accuracy of short-term prediction of mesoscale high-altitude wind field, this paper studies the short-term prediction algorithm of high-altitude wind field based on high-precision and high-coverage wind field inversion data. By analyzing the distribution characteristics and changing trend of time series of high-altitude wind field, two time series prediction models of SARIMA and GPR are established. The advantages and disadvantages of the two models and their application directions are discussed respectively, and the prediction accuracy of the model is quantitatively analyzed. The results show that the method based on GPR is better than the method based on SARIMA. In the acceptable forecast time range, the prediction model based on GPR can better capture the changing trend of high-altitude wind field and has better generalization ability. In the case of less available data, the complexity of the GPR model will not be too high, and the time will be relatively less, that is, less time in exchange for higher accuracy, which is very suitable for the short-term prediction of high-altitude wind fields.
参考文献/References:
[1] 崔世锋,程正飞,吴月友.安徽省安庆地区高空风特征分析[J].安徽农业科学,2010,38(22):11803-11805.
[2] 黄成涛,王立新,钟伯文.高空风对大型客机航线性能的量化影响[J].北京航空航天大学学报,2017,43(7):1348-1354.
[3] Zhu J,Wang H,Li J,et al.Research and Optimization of Meteo-Particle Model for Wind Retrieval[J].Atmosphere,2021,12(9):1114.
[4] 陈天舒.基于ARIMA与GPR组合模型的人民币汇率预测[D].济南:山东大学,2015.
[5] Swastanto B A.Gaussian process regression for long-term time series forecasting[D].Delft:Technische Universiteit Delft,2016.
[6] 徐俊,徐文辉,曾鑫,等.基于大数据挖掘电量预测方法的创新及应用[J].电力大数据,2018,21(10):57-61.
[7] 庄景璞. 基于神经网络的SARIMA模型人才需求与培养匹配预测研究[D]. 南京:南京邮电大学, 2020.
[8] 王晓.基于时间序列的铁路客流量预测研究[D].西安:长安大学,2020.
[9] 张亚婕.基于ARIMA模型对股票和指数预测结果的简单比较分析[J].市场研究,2019(11):23-26.
[10] Roberts S,Osborne M,Ebden M,et al.Gaussian processes for time-series modelling[J].Philosophical Transactions of the Royal Society A:Mathematical,Physical and Engineering Sciences,2013,371(1984):20110550.
[11] Rasmussen C E,Williams C K I.Model selection and adaptation of hyperparameters[M].Gaussian processes for machine learning.MIT Press,2005:105-128.
[12] Cai H,Jia X,Feng J,et al.Gaussian Process Regression for numerical wind speed prediction enhancement[J].Renewable energy,2020,146:2112-2123.
[13] Roman I,Santana R,Mendiburu A,et al.Evolving Gaussian process kernels from elementary mathematical expressions for time series extrapolation[J].Neurocomputing,2021,462:426-439.
相似文献/References:
[1]刘 涛,廖 伟,卿烈华,等.基于ADS-B数据的风矢量反演方法[J].成都信息工程大学学报,2020,35(04):412.[doi:10.16836/j.cnki.jcuit.2020.04.008]
LIU Tao,LIAO Wei,QIN Liehua,et al.Wind Vector Inversion Method of based on ADS-B Data[J].Journal of Chengdu University of Information Technology,2020,35(06):412.[doi:10.16836/j.cnki.jcuit.2020.04.008]
[2]朱嘉慧,王海江,李 静,等.Meteo-particle模型在ADS-B风场反演中的性能研究[J].成都信息工程大学学报,2021,36(05):479.[doi:10.16836/j.cnki.jcuit.2021.05.001]
ZHU Jiahui,WANG Haijiang,LI Jing,et al.Study on the Performance of the Meteo-particle Model in Wind Field Retrieval with ADS-B[J].Journal of Chengdu University of Information Technology,2021,36(06):479.[doi:10.16836/j.cnki.jcuit.2021.05.001]
备注/Memo
收稿日期:2022-06-25
基金项目:国家自然科学基金资助项目(U1733103)