Looking for breakthrough ideas for innovation challenges? Try Patsnap Eureka!

CNN multi-information fusion-based GNSS-R sea surface wind speed inversion method and system

A technology of multi-information fusion and sea surface wind speed, applied in neural learning methods, ICT adaptation, climate sustainability, etc., can solve the problems of DDM feature loss of effective information, difficult to consider GNSS-R wind speed retrieval process information, etc., to achieve High training and use efficiency, high precision, and comprehensive effects

Pending Publication Date: 2022-08-05
WUHAN UNIV
View PDF0 Cites 2 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0004] The purpose of the present invention is to solve the problem that the existing artificial extraction of DDM features will lose effective information and it is difficult to consider all the information that affects the GNSS-R wind speed retrieval process

Method used

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View more

Image

Smart Image Click on the blue labels to locate them in the text.
Viewing Examples
Smart Image
  • CNN multi-information fusion-based GNSS-R sea surface wind speed inversion method and system
  • CNN multi-information fusion-based GNSS-R sea surface wind speed inversion method and system
  • CNN multi-information fusion-based GNSS-R sea surface wind speed inversion method and system

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0041] Embodiment 1 of the present invention provides a GNSS-R sea surface wind speed inversion method based on CNN multi-information fusion, which uses convolutional neural network to extract DDM features, and then uses the latitude and longitude, RCG, signal incident angle, and effective wave height of the specular reflection point. Feature fusion, input wind speed inversion model, output inversion wind speed, the realization process includes the following steps:

[0042] Step S1: constructing the DDM and the convolution input vector corresponding to the effective scattering area and the auxiliary vector of the longitude and latitude, RCG, signal incident angle, and effective wave height of the specular reflection point;

[0043] Further, step S1 may specifically include:

[0044] Step S1.1: Match the GNSS-R data with the effective wave height data in space and time;

[0045] Step S1.2: Combine the GNSS-R L1b BRCS DDM and the effective scattering cross-sectional area of ​​t...

Embodiment 2

[0065] Embodiment 2 of the present invention provides a GNSS-R sea surface wind speed inversion method based on CNN multi-information fusion, which can also be implemented using this process, which mainly includes the following steps:

[0066] The first step is to download and save the dataset. Use python script to batch download CYGNSS L1 data (CYGNSS data is a kind of spaceborne GNSS-R data), ECMWF wind speed data and ECMWF effective wave height data of the specified date range, and save them to the specified database;

[0067] The second step is data preprocessing and data set matching. Calculate the RCG in the data sample corresponding to CYNGSS and write it to the file, and interpolate and match the CYGNSS data with the wind speed data and effective wave height according to space and time in turn;

[0068] The third step is to generate training and test sets. According to the preset data screening conditions, filter the matched CYGNSS data set, and then divide the train...

Embodiment 3

[0106] In another embodiment 3 of the present invention, a GNSS-R sea surface wind speed inversion system based on CNN multi-information fusion is also provided, such as image 3 shown, including:

[0107] The data processing module is used to calculate the RCG and perform space-time matching between the CYGNSS data and the ECMWF significant wave height data;

[0108] Model storage module, used to store the trained CNN model;

[0109] The wind speed inversion module inputs the processed CYGNSS data into the CNN model for wind speed inversion, and gives the corresponding inversion performance.

[0110] Further, the wind speed inversion module specifically includes:

[0111] Combine the BRCS DDM in the GNSS-R data to be inverted and the effective scattering cross-sectional area of ​​the corresponding size into the input vector of the convolution layer, and normalize it;

[0112] Combine the longitude and latitude, RCG, signal incident angle, and effective wave height of the s...

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

PUM

No PUM Login to View More

Abstract

The invention discloses a CNN multi-information fusion-based GNSS-R sea surface wind speed inversion method and system, and the method comprises the steps: extracting DDM features through a convolutional neural network, carrying out the feature fusion of the DDM features with the longitude and latitude, RCG, signal incident angle and significant wave height of a mirror reflection point, inputting a wind speed inversion model, and outputting an inversion wind speed. Comprising the following steps: constructing a convolution input vector of DDM and a corresponding effective scattering area, and auxiliary vectors of longitude and latitude, RCG, a signal incident angle and an effective wave height of a specular reflection point; inputting the convolution input vector and the auxiliary vector into a wind speed inversion model, and outputting a corresponding inversion wind speed; the wind speed inversion model is a CNN-based effective feature extraction and multi-information fusion GNSS-R sea surface wind speed inversion model. According to the method, the characteristics of the CNN are fully utilized, the two steps of feature extraction and model fitting are unified into the end-to-end CNN when the GMF is constructed, the GNSS-R wind speed inversion model with high inversion precision, high comprehensiveness and high robustness is obtained through training of a deep learning method, and efficient automatic inversion is achieved.

Description

technical field [0001] The invention relates to the fields of electronics, information, atmospheric science and the like, in particular to a multi-information fusion CNN sea surface wind speed inversion method and system. Background technique [0002] The sea surface wind field is an important meteorological parameter, which has a significant impact on the global climate, and the sea surface wind speed is one of the most important parameters of the sea surface wind field. an important question for research. Traditional sea surface wind detection methods include buoys, meteorological remote sensing satellites, etc., which have the defects of limited measurement range and high power consumption cost. In recent years, the theory of sea surface wind speed inversion by global navigation satellite system reflectometry (GNSS-R) technology has been continuously improved and developed, providing a sea surface wind field with large coverage, high temporal and spatial resolution, and ...

Claims

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

Application Information

Patent Timeline
no application Login to View More
Patent Type & Authority Applications(China)
IPC IPC(8): G06F30/27G06K9/62G06N3/04G06N3/08
CPCG06F30/27G06N3/084G06F2113/08G06N3/045G06F18/214Y02A90/10
Inventor 郭文飞杜皓郭迟
Owner WUHAN UNIV
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Patsnap Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Patsnap Eureka Blog
Learn More
PatSnap group products

Browse by: Latest US Patents, China's latest patents, Technical Efficacy Thesaurus, Application Domain, Technology Topic, Popular Technical Reports.

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap|About US| Contact US: help@patsnap.com