Please use this identifier to cite or link to this item: http://hdl.handle.net/10397/117256
PIRA download icon_1.1View/Download Full Text
Title: Typhoon dynamic data prediction method and device, electronic equipment and storage medium
Other Title: 台风动态数据预测方法和装置、电子设备及存储介质
Authors: Chen, ZW 
Zeng, YJ 
Lu, J 
Issue Date: Nov-2025
Source: 中国专利 ZL 202511094421.X
Abstract: The embodiment of the invention provides a typhoon dynamic data prediction method and device, electronic equipment and a storage medium, and belongs to the technical field of typhoon prediction. The method comprises the steps that typhoon meteorological time series data are acquired, and the typhoon meteorological time series data comprise multiple variable dimensions and multiple time nodes; based on the time node, carrying out time sequence feature coding on the typhoon meteorological time sequence data to obtain a typhoon meteorological time feature; performing variable feature coding on the typhoon meteorological time features based on the variable dimensions to obtain typhoon meteorological fusion features; performing typhoon path prediction based on the typhoon meteorological fusion features to obtain typhoon path prediction data, and performing typhoon intensity prediction based on the typhoon meteorological fusion features to obtain typhoon intensity prediction data; and carrying out data merging on the typhoon path prediction data and the typhoon intensity prediction data to obtain typhoon dynamic data. According to the embodiment of the invention, the prediction accuracy of the typhoon dynamic data can be improved.
本申请实施例提供了一种台风动态数据预测方法和装置、电子设备及存储介质,属于台风预测技术领域。该方法包括:获取台风气象时序数据,其中,所述台风气象时序数据包含多种变量维度和多个时间节点;基于所述时间节点,对所述台风气象时序数据进行时序特征编码,得到台风气象时间特征;基于所述变量维度,对所述台风气象时间特征进行变量特征编码,得到台风气象融合特征;基于所述台风气象融合特征,进行台风路径预测,得到台风路径预测数据,并基于所述台风气象融合特征,进行台风强度预测,得到台风强度预测数据;对所述台风路径预测数据及所述台风强度预测数据进行数据合并,得到台风动态数据。本申请实施例能够提高台风动态数据的预测精准度。
Publisher: 中华人民共和国国家知识产权局
Rights: Assignee: 香港理工大学深圳研究院
Appears in Collections:Patent

Files in This Item:
File Description SizeFormat 
ZL202511094421.X.PDF2.26 MBAdobe PDFView/Open
Open Access Information
Status open access
File Version Version of Record
Show full item record

Google ScholarTM

Check


Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.