CN-122024456-A - Disaster chain real-time monitoring and early warning method, device, equipment and storage medium

CN122024456ACN 122024456 ACN122024456 ACN 122024456ACN-122024456-A

Abstract

The application discloses a disaster chain real-time monitoring and early warning method, device, equipment and storage medium, relating to the technical field of disaster monitoring, comprising the steps of acquiring multi-source remote sensing data, a disaster chain knowledge graph and a current disaster chain risk threshold matrix based on grid units of a target area; the method comprises the steps of updating a disaster chain knowledge graph based on multi-source remote sensing data, inputting the updated disaster chain knowledge graph into a disaster-crossing type coupling model to obtain a disaster chain risk probability matrix in a set period in the future, wherein the disaster-crossing type coupling model comprises a landslide instability model, a barrier lake overtop-piping collapse model and a flood evolution model, and comparing the disaster chain risk probability matrix with a current disaster chain risk threshold matrix to trigger hierarchical early warning. The method solves the technical problems that single disaster model is independent, and the complete physical process of triggering the follow-up disaster by the front disaster in the disaster chain cannot be accurately simulated, and improves the accuracy of the risk early warning of the cross disaster.

Inventors

XU SHIRUI
HUANG MIN
ZHANG JIZHOU
NAN DAN
ZHAO ZIHUAN
LI YUHANG

Assignees

北京数慧时空信息技术有限公司

Dates

Publication Date: 20260512
Application Date: 20260413

Claims (10)

1. The disaster chain real-time monitoring and early warning method is characterized by comprising the following steps of: acquiring multisource remote sensing data, a disaster chain knowledge graph and a current disaster chain risk threshold matrix based on grid units of a target area; updating the disaster chain knowledge graph based on the multi-source remote sensing data; Inputting the updated disaster chain knowledge graph into a disaster-crossing type coupling model to obtain a disaster chain risk probability matrix in a set period in the future, wherein the disaster-crossing type coupling model comprises a landslide instability model, a barrier lake overtop-piping burst model and a flood evolution model; And comparing the disaster chain risk probability matrix with the current disaster chain risk threshold matrix to trigger hierarchical early warning.
2. The method of claim 1, wherein acquiring multi-source telemetry data based on grid cells of a target area comprises: acquiring an optical remote sensing image, a synthetic aperture radar SAR image, a thermal infrared image, a LiDAR point cloud and a deformation time sequence; sub-pixel level registration is carried out on the optical remote sensing image and the synthetic aperture radar SAR image, and registration image data are obtained; interpolation is carried out on the missing values of the deformation time sequence to obtain complete deformation time sequence data; Correcting based on the registration image data, the complete deformation time series data, the thermal infrared image and the LiDAR point cloud to obtain correction data; and aligning the correction data with grid cells of the target area to obtain multi-source remote sensing data.
3. The method of claim 1, wherein obtaining a disaster chain knowledge graph comprises: Acquiring historical disaster seed data; Based on the historical disaster species data, obtaining physical triggering relation, disaster-enriched environmental factors, remote sensing invertible parameters and disaster-bearing body exposure degree among different disaster species; determining disaster-bearing nodes based on the disaster-tolerant environmental factors, the remote sensing invertible parameters and the disaster-bearing body exposure; And establishing a disaster chain knowledge graph according to the disaster node and the physical triggering relationship.
4. The method of claim 1, wherein obtaining a current disaster chain risk threshold matrix based on grid cells of the target area comprises: Acquiring a historical disaster chain risk threshold matrix, weather forecast, night light data, population density grid data and road network density data; Obtaining a social vulnerability index based on the night light data, the population density grid data and the road network density data; obtaining a weather condition enhancement factor based on the weather forecast; And aligning the grid units of the target area with the historical disaster chain risk threshold matrix, the meteorological condition enhancement factors and the social weakness index to obtain a current disaster chain risk threshold matrix.
5. The method of claim 1, wherein the inputting the updated disaster chain knowledge graph into a cross-disaster type coupling model to obtain a disaster chain risk probability matrix in a future set period, the cross-disaster type coupling model including a landslide instability model, a barrier lake overtop-piping collapse model and a flood evolution model includes: inputting the updated disaster chain knowledge graph into a disaster-crossing type coupling model, wherein the disaster-crossing type coupling model comprises a landslide instability model, a barrier lake overtop-piping collapse model and a flood evolution model; acquiring a landslide safety coefficient output by the landslide instability model, and marking a grid with the landslide safety coefficient smaller than a set value as a potential landslide source; Inputting the potential landslide source into the barrier lake overtopping-piping collapse model to obtain the collapse flood peak flow; inputting the burst flood peak flow into the flood evolution model to obtain downstream flood risk probability in a set period in the future; and determining a disaster chain evolution probability matrix based on the downstream flood risk probability in the future set period.
6. The method of claim 1, wherein the comparing the disaster chain risk probability matrix to the current disaster chain risk threshold matrix triggers a hierarchical pre-warning comprising: comparing the disaster chain risk probability matrix with the current disaster chain risk threshold matrix to determine a grid early warning area and a grid early warning level, wherein the grid early warning area and the grid early warning level are in one-to-one correspondence; performing color coding according to the grid early warning area and the grid early warning level to obtain a grid level early warning result; determining early warning information based on the grid level early warning result; and triggering hierarchical early warning based on the early warning information.
7. The method of claim 6, wherein the color coding is performed according to the grid early warning area and the grid early warning level, and further comprising, after obtaining the grid level early warning result: When the cellular network is interrupted, returning the grid level early warning result to the cloud end through a Beidou No. three short message or a LoRa ad hoc network; when the communication is normal, the original remote sensing slices and the feature images are uploaded to the cloud for model retraining in batches; And uploading the grid level early warning result in a preset range when the electric quantity is lower than a preset electric quantity threshold value.
8. A disaster chain real-time monitoring and early warning device, characterized in that the device comprises: The acquisition module is used for acquiring multi-source remote sensing data, a disaster chain knowledge graph and a current disaster chain risk threshold matrix based on grid units of the target area; The updating module is used for updating the disaster chain knowledge graph based on the multi-source remote sensing data; The computing module is used for inputting the updated disaster chain knowledge graph into a disaster-crossing type coupling model to obtain a disaster chain risk probability matrix in a set period in the future, wherein the disaster-crossing type coupling model comprises a landslide instability model, a barrier lake overtop-piping burst model and a flood evolution model; And the early warning module is used for comparing the disaster chain risk probability matrix with the current disaster chain risk threshold matrix and triggering hierarchical early warning.
9. A disaster chain real-time monitoring and early warning device, characterized in that the device comprises a memory, a processor and a computer program stored on the memory and executable on the processor, the computer program being configured to implement the steps of the disaster chain real-time monitoring and early warning method according to any one of claims 1 to 7.
10. A storage medium, characterized in that the storage medium is a computer-readable storage medium, on which a computer program is stored, which computer program, when being executed by a processor, implements the steps of the disaster chain real-time monitoring and early warning method according to any one of claims 1 to 7.

Description

Disaster chain real-time monitoring and early warning method, device, equipment and storage medium Technical Field The application relates to the technical field of disaster monitoring, in particular to a disaster chain real-time monitoring and early warning method, device, equipment and storage medium. Background When remote sensing equipment such as satellites and unmanned aerial vehicles are utilized to monitor disasters, a key technical requirement exists that dynamic changes of on-site computing resources and real-time data acquisition conditions can be achieved according to different stages of disaster chain evolution. The current state of the art is that the individual components of the disaster monitoring system are often independent of each other. Links such as data acquisition, model calculation, threshold analysis, early warning release and the like are generally responsible for different subsystems or modules, data transmission is carried out among the modules by adopting preset and static interfaces, the modules are operated separately, the data transmission is not smooth, the analysis cannot be carried out from the whole angle of a disaster chain, and the prediction of the whole disaster chain is not accurate enough. Therefore, the problem to be solved is how to improve the accuracy of the disaster risk early warning. The foregoing is provided merely for the purpose of facilitating understanding of the technical solutions of the present application and is not intended to represent an admission that the foregoing is prior art. Disclosure of Invention The application mainly aims to provide a disaster chain real-time monitoring and early warning method, device, equipment and storage medium, and aims to solve the technical problem of how to improve the accuracy of cross-disaster risk early warning. In order to achieve the above purpose, the application provides a disaster chain real-time monitoring and early warning method, which comprises the following steps: acquiring multisource remote sensing data, a disaster chain knowledge graph and a current disaster chain risk threshold matrix based on grid units of a target area; updating the disaster chain knowledge graph based on the multi-source remote sensing data; Inputting the updated disaster chain knowledge graph into a disaster-crossing type coupling model to obtain a disaster chain risk probability matrix in a set period in the future, wherein the disaster-crossing type coupling model comprises a landslide instability model, a barrier lake overtop-piping burst model and a flood evolution model; And comparing the disaster chain risk probability matrix with the current disaster chain risk threshold matrix to trigger hierarchical early warning. In one embodiment, acquiring multi-source remote sensing data based on a grid cell of a target area includes: acquiring an optical remote sensing image, a synthetic aperture radar SAR image, a thermal infrared image, a LiDAR point cloud and a deformation time sequence; sub-pixel level registration is carried out on the optical remote sensing image and the synthetic aperture radar SAR image, and registration image data are obtained; interpolation is carried out on the missing values of the deformation time sequence to obtain complete deformation time sequence data; Correcting based on the registration image data, the complete deformation time series data, the thermal infrared image and the LiDAR point cloud to obtain correction data; and aligning the correction data with grid cells of the target area to obtain multi-source remote sensing data. In an embodiment, acquiring a disaster chain knowledge graph based on a grid unit of a target area includes: Acquiring historical disaster seed data; Based on the historical disaster species data, obtaining physical triggering relation, disaster-enriched environmental factors, remote sensing invertible parameters and disaster-bearing body exposure degree among different disaster species; determining disaster-bearing nodes based on the disaster-tolerant environmental factors, the remote sensing invertible parameters and the disaster-bearing body exposure; Correlating the grid units of the target area with the disaster nodes to obtain grid disaster nodes; And establishing a disaster chain knowledge graph according to the grid disaster species nodes and the physical triggering relationship. In an embodiment, the acquiring the current disaster chain risk threshold matrix based on the grid unit of the target area includes: Acquiring a historical disaster chain risk threshold matrix, weather forecast, night light data, population density grid data and road network density data; Obtaining a social vulnerability index based on the night light data, the population density grid data and the road network density data; obtaining a weather condition enhancement factor based on the weather forecast; And aligning the grid units of the target area with the historical disaster chain ris