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

Multi-scale convolutional neural network-based heavy rainfall and thunderstorm prediction method and system

A convolutional neural network and forecasting system technology, applied in the field of weather forecasting, can solve the problems of coarse granularity and low accuracy in time and space

Active Publication Date: 2018-09-07
南京云思创智信息科技有限公司
View PDF3 Cites 27 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0004] Aiming at the above-mentioned deficiencies existing in the prior art, the technical problem to be solved by the present invention is to provide a method and system for forecasting heavy rainfall and thunderstorms based on multi-scale convolutional neural networks, to solve the problems of traditional heavy rainfall and thunderstorm forecasting systems. Coarse granularity and low accuracy in time and space

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
  • Multi-scale convolutional neural network-based heavy rainfall and thunderstorm prediction method and system
  • Multi-scale convolutional neural network-based heavy rainfall and thunderstorm prediction method and system
  • Multi-scale convolutional neural network-based heavy rainfall and thunderstorm prediction method and system

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0039] The present invention will be further described in detail below in conjunction with the accompanying drawings and embodiments.

[0040] figure 1 It is a radar echo map of Jiangsu area at 8:06 on June 23, 2016, in which the reflectivity is marked with color, and the darker areas, such as red and purple, belong to strong reflectivity. Usually there is heavy rainfall or thunderstorms in this area, and the size of one pixel in the image represents the size of one kilometer on the ground surface. The generation cycle of the radar echo image is usually 5 minutes to obtain an echo image. Compared with the traditional meteorological index data, the data obtained by using the radar echo chart has the characteristics of better real-time performance, higher accuracy and lower cost in short-term weather forecasting.

[0041] Although the radar echo chart reflects the current weather conditions, we can form a space and time scale interrelated data through multiple correlated radar ...

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 relates to a multi-scale convolutional neural network-based heavy rainfall and thunderstorm prediction method and system. The method of the invention includes the following steps that: original radar echo data are acquired; the original radar echo data are preprocessed, so that radar echo time series images can be obtained; any three frames of radar echo time series images which aregenerated consecutively are selected, a radar echo convolutional neural network is constructed, and the next frame of radar echo image can be inferred; and radar echo change within one minute is inferred by means of an optical flow field method and a linear interpolation method. With the method adopted, thunderstorm weather in one hour can be predicted, and the space range of thunderstorm weatherprediction can be narrowed to be a range within 1 km.

Description

technical field [0001] The invention relates to the technical field of weather forecasting, in particular to a method and system for forecasting heavy rainfall and thunderstorms based on a multi-scale convolutional neural network. Background technique [0002] At present, the models that use machine learning models to forecast heavy rainfall or thunderstorms mainly include neural networks and support vector machines. These two models need to use some meteorological indicators when making forecasts, such as surface temperature, air pressure, high-altitude temperature, wind field, divergence, vertical velocity, etc., as well as some atmospheric instability factors calculated from meteorological indicators, such as K index, CT index, VT index, etc. The predicted results are relatively rough in time and space, and the accuracy is not high. In terms of time, it usually predicts whether there will be thunderstorms within 24 hours, and in terms of space, it usually forecasts the ...

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
IPC IPC(8): G01W1/10G01S13/95G06N3/04
CPCG01S13/95G01W1/10G06N3/045Y02A90/10
Inventor 汪力
Owner 南京云思创智信息科技有限公司
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