Flood season climate trend prediction method based on TCN

Abstract

The invention discloses a flood season climate trend prediction method based on TCN. The method comprises the following steps: analyzing a plurality of collected flood season meteorological observation index data, and screening out flood season meteorological observation indexes related to flood season weather trend factors; training a twin time domain convolution network corresponding to the weather trend factor, and mapping flood season meteorological observation index data related to the weather trend factor to a high-dimensional feature vector; inputting the high-dimensional feature vectoroutput by the trained twin time domain convolution network corresponding to each flood season weather trend factor into a corresponding weather trend prediction time domain convolution network, a temperature trend prediction time domain convolution network and a precipitation prediction time domain convolution network to obtain predicted weather, temperature and precipitation; and drawing a weather trend, a temperature trend and a precipitation trend chart according to the prediction result, and analyzing the climate trend. According to the invention, flood season climate prediction is realized, and a better prediction effect can be achieved.

Description

technical field [0001] The invention relates to the technical fields of climate prediction, deep learning, and artificial intelligence, in particular to a TCN-based method for predicting climate trends in flood seasons. Background technique [0002] Flood season is a period of concentrated and frequent occurrence of disastrous weather in a year. The forecast of climate trend in flood season is the forecast of the average climate state in flood season. The forecast of climate trend in flood season is very important for disaster prevention work. At present, in the field of climate trend prediction, some methods analyze the data of factors affecting precipitation through temporal convolutional networks to predict precipitation values. This method requires a large amount of sample data to train the network, and the training time is long, so it can only predict precipitation values. Moreover, the input data representation ability of temporal convolutional network is poor, which i...

AI Technical Summary

Problems solved by technology

At present, in the field of climate trend prediction, some methods analyze the data of factors affecting precipitation through temporal convolutional networks to predict precipitation values. This method requires a large amount of sample data to train the network, and the training time is long, so it can only predict precipitation values. Moreover, the input data representation ability of the temporal convolutional network is poor, which is not conducive to network convergence

Images