A typhoon prediction method based on a deep learning hybrid CNN-LSTM model

Abstract

The invention discloses a typhoon prediction method based on a deep learning hybrid CNN-LSTM model, and aims to solve the problems of unstable prediction result and low accuracy of an existing typhoonprediction method. According to the technical scheme, firstly, a historical optimal tropical cyclone path data set and global atmospheric marine variable data are obtained to form a data set of a training model, and then a typhoon prediction model of hybrid CNN-LSTM (Long Short Term Memory) is constructed, and training fitting is carried out on the typhoon prediction model of the hybrid CNN-LSTMby use of the data set of the training model, and finally, the trained hybrid CNN-LSTM typhoon prediction model is used for predicting whether the typhoon is formed or not and predicting the formed path and intensity. According to the typhoon prediction method and system, existing public data sets and a deep learning framework can be conveniently used for constructing and predicting the model, andafter typhoon data and atmospheric marine variable data in recent dozens of years are used for training the constructed model, the trained model can be directly used for effectively improving the typhoon prediction accuracy.

Description

technical field [0001] The invention relates to a typhoon forecasting method based on a deep learning hybrid convolutional neural network (CNN) and a long-short-term memory neural network (LSTM) model, which belongs to the intersecting fields of computers, meteorology and oceans. Background technique [0002] As an extreme weather event, a typhoon will not only affect marine activities, but also cause significant losses to people's lives and urban economies in coastal areas. Therefore, typhoon research and prediction has always been the focus of various coastal countries. Typhoon prediction generally predicts the intensity and path of typhoons. In recent years, path prediction has made great progress, but typhoon intensity prediction has not improved much. [0003] In order to provide early warning of typhoon disasters, accurate typhoon formation, intensity, and path prediction methods have always been the focus of attention in the field of meteorology. Due to the complexi...

AI Technical Summary

Problems solved by technology

[0005] The technical problem to be solved in the present invention is to propose a typhoon prediction method based on the deep learning hybrid CNN-LSTM m

Images