Severe convective weather prediction method and system for improving three-dimensional generative adversarial neural network based on hybrid evolutionary algorithm

Abstract

The invention discloses a severe convective weather prediction method and system for improving a three-dimensional generative adversarial neural network based on a hybrid evolutionary algorithm, and the method comprises the steps: reading radar echo data from an original Doppler weather radar, generating a training data set, and dividing the training data set into a plurality of groups of input data; constructing a hybrid evolutionary algorithm by using a genetic algorithm and a cross entropy algorithm; establishing an improved three-dimensional confrontation generation neural network model based on a hybrid evolutionary algorithm; training the three-dimensional generative adversarial neural network model through the training data set to obtain a trained three-dimensional generative adversarial neural network model; the number N of radar echo data needing to be input is obtained according to the to-be-predicted time range, and the latest N pieces of preprocessed radar echo data are input into the trained three-dimensional confrontation generation neural network model for severe convective weather prediction. The severe convective weather prediction accuracy can be effectively improved.

Description

technical field [0001] The invention belongs to the technical field of short-term near-weather prediction, and in particular relates to a strong convective weather prediction method and system based on a hybrid evolution algorithm to improve a three-dimensional confrontation generation neural network. Background technique [0002] Severe convective weather is a very dangerous weather phenomenon in summer in my country. It may be accompanied by hail, lightning, strong wind and short-term heavy rainfall. Due to the small spatial scale and short duration of this kind of weather, it is difficult to effectively predict it with conventional observation instruments. Doppler weather radar, with its high spatial and temporal resolution, is currently the only practical instrument capable of routinely sampling the detailed three-dimensional structure of severe convective storms. Therefore, various nowcasting algorithms based on Doppler weather radar data have been developed to assist ...

AI Technical Summary

Problems solved by technology

[0009] Aiming at the problem of low prediction accuracy existing in existing nowcasting methods, the present invention proposes a strong convective weather prediction method and system based on a hybrid evolutionary algorithm to improve the three-dimensional confrontation generation neural network, through the extraction of the original Doppler radar The radar echo data trains the neural network, and at the same time predicts the radar echo data, and judges whether there is strong convective weather based on the radar echo data, which can effectively improve the accuracy of strong convective weather prediction

Images