Unlock instant, AI-driven research and patent intelligence for your innovation.

Construction and application of landslide displacement prediction model driven by semantic information

A technology of semantic information and prediction model, applied in semantic analysis, neural learning method, biological neural network model, etc., can solve the problem that the landslide disaster cannot distinguish between the flat stage and the acceleration stage, affects the prediction effect of landslide disaster, and the prediction of daily displacement is not accurate enough. and other problems to achieve the effect of improving the effect of landslide warning.

Active Publication Date: 2022-04-26
CHONGQING JIAOTONG UNIVERSITY
View PDF10 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0004] Aiming at the deficiencies in the prior art, the present invention proposes a construction and application method of a landslide displacement prediction model driven by semantic information, so as to solve the problems existing in the prior art that the landslide disaster prediction cannot be distinguished from the gentle stage and the acceleration stage, and the logarithmic value fluctuates. The prediction of the daily displacement in the acceleration stage of the landslide is not accurate enough, which affects the technical problem of the prediction effect of the landslide disaster

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
  • Construction and application of landslide displacement prediction model driven by semantic information
  • Construction and application of landslide displacement prediction model driven by semantic information
  • Construction and application of landslide displacement prediction model driven by semantic information

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0047] This embodiment provides a landslide displacement prediction model driven by semantic information (hereinafter referred to as the landslide displacement prediction model), such as figure 1 As shown, the construction method of the prediction model is as follows:

[0048] S1. Through feature construction and feature combination, the landslide monitoring data is converted into semantic information, and individual combined samples and continuous combined samples are obtained

[0049] In a specific implementation, the landslide monitoring data is converted into semantic information through feature construction and feature combination, which is divided into the following two steps:

[0050] S1-1. Transform landslide monitoring data into semantic information features represented by text in a discretized manner

[0051] In a specific embodiment, the landslide monitoring data is exemplified by rainfall. Because landslide rainfall monitoring and the Meteorological Bureau use th...

Embodiment 2

[0115] This embodiment provides a landslide displacement prediction model driven by semantic information, which is constructed using the construction method in Embodiment 1, including:

[0116] The monitoring data semantic information conversion layer is used to convert landslide monitoring data into semantic information to obtain individual combined samples and continuous combined samples;

[0117] The phase division layer is used to divide the landslide cumulative displacement curve into gentle phase and acceleration phase;

[0118] The deep semantic feature extraction layer is used to extract the deep semantic features of continuous combination samples and individual combination samples according to the flat stage and the acceleration stage;

[0119] The stage identification and daily displacement prediction layer, including semantic information-driven gentle stage recognizer, used to identify the gentle stage and acceleration stage of the landslide; also includes semantic ...

Embodiment 3

[0121] This embodiment provides a landslide displacement prediction method driven by semantic information. According to landslide monitoring data, such as rainfall, using the landslide displacement prediction model driven by semantic information provided in Embodiment 2, the gentle stage and acceleration stage of the landslide Identify and predict the daily displacement during the acceleration phase of the landslide.

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 present invention provides a landslide displacement prediction model driven by semantic information, and its construction method is as follows: through feature construction and feature combination, the landslide monitoring data is converted into semantic information, and individual combined samples and continuous combined samples are obtained; sliding window variance algorithm is adopted Divide the landslide cumulative displacement curve into gentle stage and acceleration stage; use continuous combination samples for deep semantic feature extraction in the gentle stage, and use individual combination samples for deep semantic feature extraction in the acceleration stage; use continuous combination samples and separate combination samples for deep semantic feature extraction The long-short-term memory neural network is trained to obtain a semantic information-driven flat stage recognizer and a semantic information-driven accelerated stage predictor; the output of the recognizer is used as an interactive switch to control the recognizer and predictor to work together. The invention can solve the problems that the prediction of the landslide disaster cannot be distinguished from the gentle stage and the acceleration stage, and the prediction of the daily displacement in the acceleration stage is not accurate enough.

Description

technical field [0001] The invention relates to the technical field of landslide disaster prediction, in particular to a method for constructing and using a landslide displacement prediction model driven by semantic information. Background technique [0002] Landslide disasters pose a great threat to people's life safety. In order to reduce the losses caused by landslide disasters, it is necessary to predict landslide disasters. In recent years, due to the influx of a large amount of monitoring data and the development of nonlinear theory, machine learning has been gradually used to construct a prediction model of landslide displacement. In the prior art, the invention patent with the publication number CN112651314A discloses a method for automatic identification of landslide disaster-affected bodies based on semantic gate and bitemporal LSTM, which includes the following steps: cutting the entire remote sensing image and making samples , build a bitemporal long-term and s...

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 Patents(China)
IPC IPC(8): G06F40/30G06K9/62G06N3/04G06N3/08
CPCG06F40/30G06N3/08G06N3/044G06N3/045G06F18/2415
Inventor 唐菲菲唐天俊李润杰马英
Owner CHONGQING JIAOTONG UNIVERSITY