Deep learning-based water regime trend prediction method for dense river network basin and application thereof

Abstract

The invention discloses a deep learning-based river network dense drainage basin water regime trend prediction method and application thereof, and the method comprises the following steps: collecting daily scale rainfall and water level time sequence point data of all rainfall stations and river channel water level stations in a whole drainage basin range, and longitude and latitude information of each observation station, and recompling the above; converting the reorganized rainfall and water level time sequence point data into spatio-temporal data through an inverse distance weight method and a contour surface diagram rendering method; constructing and training a watershed high-resolution water level forecasting model, and respectively inputting rainfall and water level spatio-temporal data tensors; wherein the watershed high-resolution water level forecasting model comprises a preliminary spatial feature extraction module, a spatial-temporal feature extraction module, a stacking module, a splicing module and a feature reduction module; and the model outputs a water level contour surface diagram of the whole watershed in the prediction period to obtain a watershed water regimen change trend. The runoff production and confluence mechanism of the drainage basin can be fully mined from historical hydrological data, and the space-time water regimen change of the whole drainage basin can be accurately forecasted.

Description

Technical field [0001] The present invention relates to the trend of the whole basin regime prediction techniques, and in particular relates to a method of forecasting trends intensive regime Application of a river basin and its depth based learning. Background technique [0002] Most previous studies of hydrological forecast focused on making quantitative or qualitative prediction of a future hydrology hydrological stations, specific river section or a single body of water, unable to make a high-resolution grid point forecasts of a region or the entire basin. When the river basin densely distributed, each river flow reciprocating motion, difficult to use river water level of a process or a region of a site to measure the water potential of the entire basin, when multiple sites simultaneously modeling hydrological forecasting, possible input the problem of data complexity and dimension disasters, leading to the prediction accuracy is too low, a separate forecast hydrological cond...

AI Technical Summary

Problems solved by technology

The physical hydrodynamic process model is mainly based on various hydrodynamic equations to perform real simulation calculations on the drainage system in the watershed. The modeling process is cumbersome and the calculation speed is extremely slow.

The neural network model has no clear physical modeling process, so it is also called a black box model. Its biggest defect is determined by its modeling characteristics. If the underlying surface of the watershed changes, the model needs to be retrained, but the watershed It is difficult for the underlying surface to undergo major changes in a short period of time

Images