Large-watershed runoff simulation method based on computer vision and LSTM neural network

A computer vision and neural network technology, applied in the field of large watershed runoff simulation, can solve the problems of difficult to effectively screen and make full use of data space information, unable to make full use of input data space characteristics, low model operation efficiency, etc. The effect of shortening operation time and improving operation efficiency

Inactive Publication Date: 2021-12-03
YUNNAN UNIV
View PDF2 Cites 1 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0006] The present invention aims at the problem that the existing data-driven model is difficult to effectively screen and make full use of data space information when it is applied in a large watershed, so that the simulation accuracy is low and the model operation efficiency is low. It provides a large-scale model based on computer vision and LSTM neural network. Watershed runoff simulation method, that is, a runoff simulation method suitable for large watersheds based on computer vision, MIV average influence value, LSTM neural network and other methods
It can realize the screening of massive grid data, the extraction and representation of spatial feature information, and is suitable for the runoff simulation of large watersheds, improving the simulation accuracy and model operation efficiency; it solves the defect that the existing technology cannot make full use of the spatial features of the input data, thereby improving Performance and Simulation Accuracy of Data-Driven Hydrological Models in Large Watersheds

Method used

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View more

Image

Smart Image Click on the blue labels to locate them in the text.
Viewing Examples
Smart Image
  • Large-watershed runoff simulation method based on computer vision and LSTM neural network
  • Large-watershed runoff simulation method based on computer vision and LSTM neural network
  • Large-watershed runoff simulation method based on computer vision and LSTM neural network

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0072] Embodiment 1: a kind of large watershed runoff simulation method based on computer vision and LSTM neural network, concrete steps are as follows:

[0073] (1) Collect the meteorological and hydrological data sets of the watershed, in which the meteorological and hydrological data sets include the flow time series of the target section and the precipitation data of the whole watershed;

[0074] Meteorological and hydrological data sets can also include temperature, wind speed, humidity, solar radiation, snow cover, snow water equivalent, leaf area index (LAI), normalized difference vegetation index (NDVI), soil moisture, water storage (gravity One or more of satellite data derivation);

[0075] (2) standardize and normalize the meteorological and hydrological data set in step (1) to 0-1 to obtain the model input data set;

[0076] The data is standardized to a unified data resolution, and the normalization process is to normalize the original meteorological and hydrolog...

Embodiment 2

[0130] Embodiment 2: In this embodiment, runoff simulation is performed in the Ayeyarwady River Basin;

[0131] Overview of the Ayeyarwady River Basin: The Ayeyarwady River flows through China, India, and Myanmar. The basin covers about 60% of Myanmar's territory, and more than 90% of Myanmar's population lives there. It is known as the mother river of Myanmar. The Ayeyarwady River originates in Zayu County, Tibet, China. Its main tributaries include Daying River, Ruili River, Qindun River, Mu River, Yao River and Meng River. 430,000km 2 ;

[0132] A large watershed runoff simulation method based on computer vision and LSTM neural network, the specific steps are as follows:

[0133] (1) Collect the meteorological and hydrological data sets of the basin. According to the data availability of the Ayeyarwady River basin, collect the daily flow time series of the PYAY station of the control section of the basin from 1996 to 2010, and the model input data set of the whole basin f...

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to view more

PUM

No PUM Login to view more

Abstract

The invention relates to a large-watershed runoff simulation method based on computer vision and an LSTM neural network, and belongs to the technical field of hydrological simulation. The method comprises the following steps: collecting a meteorological and hydrological data set of a watershed, and standardizing and normalizing the meteorological and hydrological data set to obtain an input data set; identifying and extracting spatial texture features of the input data set except the traffic time sequence by adopting a local binary pattern algorithm of a computer vision technology; representing the spatial texture features and the intensity features of the input data set by using spatial pyramid mapping SPM; adopting an MIV average influence value method to calculate the average influence value of the intensity characteristics of the data on the flow, and screening the input data; dividing the screened data into a training data set and a verification data set, and training the LSTM neural network based on the training data set; and carrying out runoff simulation based on the verification data set and the trained LSTM neural network, and verifying a simulation result by adopting the actually measured flow.

Description

technical field [0001] The invention relates to a large watershed runoff simulation method based on computer vision and LSTM neural network, belonging to the technical field of hydrological simulation. Background technique [0002] Hydrological models can be divided into two categories: physical process-based hydrological models and data-driven hydrological models. The simulation method based on physical process refers to the hydrological principle, based on the mathematical model to describe the runoff process and river evolution process. This method has a large number of model parameters that need to be calibrated and its optimization is difficult. Compared with the hydrological model based on physical process, the data-driven hydrological model method weakens the description of the physical process of the water cycle. It is a black box method with the goal of establishing the optimal mathematical relationship between input and output data. This method has a simple structu...

Claims

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to view more

Application Information

Patent Timeline
no application Login to view more
Patent Type & Authority Applications(China)
IPC IPC(8): G06F30/27G06N3/04G06F113/08
CPCG06F30/27G06F2113/08G06N3/044
Inventor 袁旭陆颖何大明李亚张珂瑶王加红王海龙郭子璞赖红
Owner YUNNAN UNIV
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Try Eureka
PatSnap group products

Browse by: Latest US Patents, China's latest patents, Technical Efficacy Thesaurus, Application Domain, Technology Topic.

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap