Flood flow prediction method

Abstract

The invention discloses a flood flow prediction method, and belongs to the field of deep learning. The method comprises the following steps: 1, inputting a collected past flood session data set of the Changhua region, and extracting data of the input data set; 2, transmitting the data into a feature enhancement module, enlarging the dimension difference of input and output data by rewriting a multi-level convolutional neural network, and highlighting and obtaining key information in flood factors; 3, a feature extraction module and a time information coding module are introduced into the recurrent neural network, modeling is performed from different perspectives, and enhanced data are transmitted into the constructed neural network for training; 4, balancing the double-view-angle weight, adjusting the proportion of the double-view-angle weight, and carrying out perception information fusion through a merging unit to complete training; and 5, predicting the test set data to obtain a final prediction result. The prediction method is high in accuracy and high in prediction efficiency, and flood prediction can be rapidly completed.

Description

technical field [0001] The invention relates to a method for predicting flood flow, which belongs to the field of deep learning. Background technique [0002] Flood is a natural phenomenon in which the water volume of rivers and lakes increases rapidly and the water level rises rapidly due to factors such as torrential rain and wind tide. Because the water body rises beyond a certain water level, it threatens the safety of relevant areas and even causes disasters. Therefore, flood prediction is extremely important. Once we can accurately predict floods in advance, the lives and properties of thousands of people can be protected. This makes flood prediction urgent for researchers in the computer and hydrology fields. And important task. [0003] Today, many researchers are working on designing accurate and reliable flood forecasting models, and these methods are generally divided into two categories: hydrological models and data-driven models. The hydrological model summari...

AI Technical Summary

Problems solved by technology

[0005] However, due to the low frequency of floods and the complex mechanism, various problems will arise when using deep learning methods to explore the relationship between river flow and characteristic factors, such as insufficient data, resulting in over-fitting or low-problem modeling ability; at the same time, how to effectively extract time information and feature information plays a key role in understanding sequence information, which also brings about the problem of how to conduct accurate modeling

Images