An emergency simulation method and system based on rapid acquisition of real-time information of dam-break

Abstract

This invention discloses a simulation emergency response method and system for rapidly acquiring real-time information on landslide dam failure. The method includes: collecting image data and point cloud data of the landslide dam area using a drone; acquiring web text information related to the landslide dam through web crawling; matching all point cloud data into the same spatial coordinate system using an iterative nearest-point algorithm to construct a three-dimensional terrain model of the landslide dam area; interpreting the image data based on the constructed three-dimensional terrain model to obtain the physical parameters of the landslide dam; correcting the physical parameters through a multi-source data fusion mechanism; inputting the corrected physical parameters into a preset failure process calculation model to dynamically simulate the failure process, and updating the landslide dam physical parameter values ​​in conjunction with real-time image data during the simulation; outputting the failure time, peak flow, and downstream inundation range based on the simulation results, and providing emergency strategies through a visual interface based on the output results.

Description

Technical Field

[0001] This invention relates to a simulation emergency response method, and more particularly to a simulation emergency response method and system based on the rapid acquisition of real-time information on landslide dam failure. Background Technology

[0002] A landslide dam is a temporary dam formed when a mountain collapses due to natural disasters such as earthquakes, landslides, or debris flows, blocking a river channel. Its structure is loose and its stability is poor; once the water level rises to a high level or it is affected by aftershocks or rainfall, it is extremely prone to breaching, causing massive floods that pose a serious threat to the lives and property of people downstream.

[0003] Existing emergency response methods for landslide dams have the following drawbacks: they mainly rely on dispatching surveyors to the site on foot or by helicopter for manual measurement. This method is time-consuming and risky, especially when transportation is disrupted or weather conditions are severe, making it difficult to obtain accurate on-site data in the short term. Furthermore, manual measurement is limited by tools and the environment, and the key parameters obtained, such as dam dimensions, reservoir capacity, and breach morphology, are often estimated values ​​with low accuracy, directly affecting the accuracy of subsequent analysis.

[0004] Therefore, it is of great significance to provide a simulation emergency response method and system based on the rapid acquisition of real-time information on landslide dam failure. Summary of the Invention

[0005] Purpose of the invention: The purpose of this invention is to provide a simulation emergency response method and system based on the rapid acquisition of real-time information on landslide dam failure, which solves the problems of slow speed and low accuracy of manual measurement in the prior art, and improves the efficiency and effectiveness of emergency response to landslide dam disasters.

[0006] Technical solution: The present invention provides a simulation-based emergency response method for rapid acquisition of real-time information on landslide dam failure, comprising:

[0007] UAVs were used to collect image data and point cloud data of the landslide dam area, and web crawlers were used to obtain web text information related to the landslide dam. All point cloud data were matched to the same spatial coordinate system through the iterative nearest point algorithm. Based on the matched point cloud data, a three-dimensional terrain model of the landslide dam area was constructed through the triangulation algorithm.

[0008] The image data is interpreted based on the constructed 3D terrain model to obtain the physical parameters of the landslide dam; and the physical parameters are corrected through a multi-source data fusion mechanism.

[0009] The corrected physical parameters are input into the preset failure process calculation model to perform dynamic simulation of the failure process, and the physical parameter values ​​of the landslide dam are updated in combination with real-time image data during the simulation.

[0010] Based on the simulation results, the time required for the flood breach, the peak flow rate, and the downstream inundation range are output, and emergency strategies are provided through a visualization interface based on the output results.

[0011] Furthermore, the method also includes preprocessing the point cloud data by using a statistical filtering algorithm to remove outliers and noise points from the point cloud data. For each point in the point cloud... Calculate the average distance from its k nearest neighbors. Calculate the mean of all average distances in the point cloud. and standard deviation Set distance threshold range Eliminate all that satisfy point , where n represents the standard deviation multiplier.

[0012] Furthermore, the construction of a triangular terrain model of the landslide dam area based on the matched point cloud data and the triangulation algorithm specifically includes: projecting the three-dimensional point cloud onto a two-dimensional horizontal plane, performing triangulation calculations based on its X and Y coordinates, and the calculation result being a vertex index list defining which vertices constitute each triangular facet; reassigning the elevation values ​​corresponding to each vertex in the point cloud to each vertex in the two-dimensional triangle; restoring the two-dimensional planar triangulation to a three-dimensional triangulation model; and generating a three-dimensional terrain model.

[0013] Furthermore, the step of correcting the physical parameters through a multi-source data fusion mechanism specifically includes: parsing the acquired network text information to extract numerical information on the dam height, water level, and breach size; quantifying the extracted numerical information into a unified parameter value range; if the text description is a single numerical value, then constructing a parameter value range with both upper and lower bounds equal to that value; calculating a credibility score for this value range; and outputting a credibility score between 0 and 1 through a weighted scoring function, and assigning it to the text constraint boundary.

[0014] The physical parameters of the landslide dam interpreted from the image data are compared in real time with the text constraint boundaries to calculate the relative deviation value. The calculated relative deviation value is then compared with a preset deviation threshold. If the relative deviation is greater than the preset threshold, a data conflict is identified, triggering a parameter correction process. Otherwise, the physical parameter values ​​derived from the image data interpretation are used as the reliable result. The correction process includes:

[0015] Obtain the physical parameters and initial weights derived from the image data;

[0016] The median value of the text constraint boundary is obtained as the text representative value, and the credibility score of the text constraint boundary is used as its initial weight.

[0017] Retrieve historical topographic and geomorphological feature values ​​related to physical parameters from the historical hydrological database and assign them initial weights;

[0018] The initial weights of the physical parameters, textual representative values, and historical feature values ​​automatically interpreted from the image data are normalized so that the sum of the three weights is 1, resulting in the final weight coefficient. Using the normalized weights, the physical parameters, textual representative values, and historical feature values ​​are weighted and summed to calculate the final corrected value of the parameter. The above calculation is repeated for all physical parameters that trigger correction to generate the final parameter value.

[0019] Furthermore, the specific calculation process of the aforementioned failure process calculation model is as follows:

[0020] The calculation model of the breach process enters a time-stepping calculation loop. At the beginning of each time step, the model calculates the outflow velocity and energy of the water flowing through the breach based on the current reservoir water level and breach morphology.

[0021] Based on the calculated water flow energy, its scouring capacity on the dam material is assessed. When the scouring force of the water flow on the dam foundation exceeds the resistance strength of the dam material itself, the bottom of the breach will be scoured down. After the bottom of the breach is cut down, its sidewall will collapse under the action of gravity, thus causing the breach width to expand.

[0022] The model subtracts the current reservoir capacity from the total volume of water discharged, and updates the reservoir water level after the drop based on the correspondence between reservoir capacity and water level.

[0023] After completing the calculation for one time step, real-time UAV time-series image data is acquired, the measured width of the breach at that moment is automatically extracted, and the measured width is used to replace the calculated value of the breach width for the current step in the model.

[0024] The updated breach size, reservoir water level, and other variables are used as new inputs. The system automatically moves to the next time step and repeats the above operations. Finally, the model outputs the breach morphology change process, the discharge flow change process, and the reservoir water level drop process over the entire time series.

[0025] Furthermore, the entire time series output by the model includes the process of breach morphology change, the process of discharge flow change, and the process of reservoir water level drop. Based on the breach morphology change process data, the time required from the initial moment to the complete instability of the dam body and the breach failure is determined. Based on the discharge flow change process data, the peak flow and its occurrence time are extracted.

[0026] Furthermore, the extraction of peak flow and its occurrence time based on the data of the outflow change process includes: using the instantaneous outflow change of the breach as a function of time, inputting the outflow change process as a boundary condition into a hydrodynamic model containing downstream topographic data, using the outflow change process output by the breach process calculation model as an upstream boundary condition, establishing a hydraulic calculation domain in combination with the downstream digital elevation model, discretizing the calculation domain using an unstructured triangular mesh, setting initial hydraulic conditions and Manning roughness coefficients, using the two-dimensional Saint-Venant equations to represent the downstream flood evolution, solving using the finite volume method adapted to the unstructured triangular network, obtaining the spatiotemporal distribution of water depth and velocity of each mesh cell through iterative calculation using the finite volume method, and finally extracting the maximum inundation range, water depth distribution, and flood arrival time during the entire simulation period.

[0027] The present invention discloses a simulation emergency system for rapid acquisition of real-time information on landslide dam failure, comprising:

[0028] Data acquisition module: Use drones to collect image data and point cloud data of the landslide dam area, and use web crawlers to obtain web text information related to the landslide dam; all point cloud data are matched to the same spatial coordinate system through the iterative nearest point algorithm, and based on the matched point cloud data, a triangular terrain model of the landslide dam area is constructed through the triangulation algorithm.

[0029] Parameter correction module: Based on the constructed 3D terrain model, the image data is interpreted to obtain the physical parameters of the landslide dam; and the physical parameters are corrected through a multi-source data fusion mechanism.

[0030] Breach Simulation Module: Input the corrected physical parameters into the preset breach process calculation model to perform dynamic simulation of the breach process, and update the physical parameter values ​​of the landslide dam in combination with real-time image data during the simulation.

[0031] Simulation emergency module: Based on the simulation results, outputs the time required for the flood failure, peak flow rate, and downstream flooding range, and provides emergency strategies through a visual interface based on the output results.

[0032] The computer device of the present invention includes one or more processors, a memory, and one or more programs, wherein the one or more programs are stored in the memory and configured to be executed by the one or more processors, and the programs, when executed by the processors, implement the steps of the simulation emergency method for rapid acquisition of real-time information on landslide dam failure.

[0033] The present invention discloses a computer-readable storage medium having a computer program stored thereon, wherein when the computer program is executed by a processor, it implements the steps of a simulation emergency method for rapid acquisition of real-time information on landslide dam failure.

[0034] Beneficial effects: Compared with existing technologies, this invention has the following significant advantages: By using drones and web crawlers, this invention shortens the information acquisition time from hours or even days to minutes; parameter extraction and model calculation are highly automated, significantly shortening the decision-making cycle; the use of unmanned and remote operation methods avoids personnel entering dangerous areas, ensuring the safety of surveyors; high-resolution imagery and automated interpretation technology make the acquired parameters more accurate than manual estimation, providing a solid foundation for the accuracy of subsequent simulation and prediction, providing quantitative and visualized scientific evidence for emergency command of landslide dams, and improving the efficiency and effectiveness of disaster emergency response. Attached Figure Description

[0035] Figure 1 This is a flowchart of the method described in this invention. Detailed Implementation

[0036] The technical solution of the present invention will be further described below with reference to the accompanying drawings.

[0037] Example 1

[0038] This embodiment provides a simulation-based emergency response method for rapid acquisition of real-time information on landslide dam failure, including:

[0039] Step S1: Use drones to collect image data and point cloud data of the landslide dam area, and use web crawlers to obtain web text information related to the landslide dam.

[0040] Specifically, the drone equipment scans the target area via a preset flight path, collecting high-resolution orthophotos, oblique photographs, and point cloud data. The drone's onboard GPS / IMU system records the precise location and flight altitude metadata of the photos. Simultaneously, a web crawler, based on internet protocols, automatically captures online text information related to the landslide dam event. The raw point cloud data collected by the drone is then processed for denoising, registration, and 3D modeling to generate a high-precision 3D terrain model of the landslide dam area. Statistical filtering algorithms are used to remove outliers and noise points from the point cloud data. For each point in the point cloud... Calculate the average distance from it to its k nearest neighbors. Calculate the mean of all average distances in the entire point cloud. and standard deviation Set a distance threshold Eliminate all that satisfy point Where n represents the standard deviation multiplier, which is a threshold scaling factor. Then, a registration operation is performed to unify the point cloud datasets acquired from different flight strips and viewpoints into the same spatial coordinate system using an iterative nearest-point algorithm. The specific calculation formula is as follows:

[0041]

[0042] Where R represents the rotation matrix and t represents the translation vector. This represents a point in the original point cloud P. Indicating the target point cloud Q and The corresponding nearest point, N represents the number of matching point pairs;

[0043] Based on the registered point cloud data, a triangular terrain model of the landslide dam area is constructed using a triangulation algorithm. The three-dimensional point cloud is projected onto a two-dimensional horizontal plane, and triangulation calculations are performed based on the X and Y coordinates. The algorithm outputs a list of vertex indices that define which vertices constitute each triangular facet. After the mesh is constructed, the elevation values ​​corresponding to each vertex in the original point cloud are reassigned to each vertex in the two-dimensional triangle, thereby restoring the two-dimensional planar triangulation mesh to a three-dimensional triangulation mesh model to generate a three-dimensional terrain model.

[0044] The web crawler system, based on web crawler programs, monitors social media, news portals, and meteorological and hydrological websites in real time, automatically capturing text, images, and videos related to landslide dams, such as on-site photos posted by the public, official announcements, real-time rainfall, and upstream water flow.

[0045] Step S2: The image data is interpreted to obtain the physical parameters of the landslide dam; and the physical parameters are corrected through a multi-source data fusion mechanism.

[0046] Specifically, the ground resolution of the image is calculated based on the metadata of the UAV orthophoto image. The specific calculation formula is as follows:

[0047]

[0048] in, The resolution represents the actual size of the ground corresponding to each pixel; f represents the focal length of the UAV camera; H represents the UAV's altitude above the ground; and p represents the pixel size of the sensor. On the generated 3D terrain model, key physical parameters are measured through image recognition, including the dam height, dam crest length, initial breach width, and breach location. The water body is delineated by identifying the shoreline in the image. Contour recognition technology automatically identifies the shoreline from the image to determine the water body's horizontal extent. Elevation information from point cloud data is then fused, and the current reservoir capacity and real-time water level are obtained through integral calculation. The specific calculation formula is as follows:

[0049]

[0050] Where V represents the total storage capacity, This represents the area of ​​the i-th water layer. The elevation interval is represented by the surface velocity of the water body. By analyzing the displacement trajectory of floating objects on the water surface in continuous time-series images, and combining the ground resolution and time interval, the surface velocity of the water body is estimated. The specific calculation formula is as follows:

[0051]

[0052] in, Indicates the surface velocity of water. Indicates that the floating object is in The actual displacement over time. This represents the pixel displacement of a floating object in a continuous image. Indicates the time interval between consecutive image captures;

[0053] A multi-source data fusion verification mechanism is initiated to perform structured parsing on the text information obtained by web crawlers, extracting key numerical descriptions of dam height, water level, and breach size. The extracted numerical information is quantified into a unified parameter value range. If the text description is a single numerical value, a parameter value range with both upper and lower bounds equal to that value is constructed. For example, if the water level is between 50 and 55 meters, the range [50, 55] is constructed. If the dam height is 50 meters, the range [50, 50] is constructed. A credibility score is calculated for this value range. Through a weighted scoring function, a credibility score between 0 and 1 is output and assigned to the text constraint boundary.

[0054] The key physical parameters automatically interpreted from the image are compared with the text constraint boundaries in real time to calculate the relative deviation value. If the key physical parameters automatically interpreted from the image fall within the text range, the relative deviation is zero. If it is below the lower limit of the range, the relative deviation is ((lower limit of the range - key physical parameters automatically interpreted from the image) divided by the image interpretation value). If it is above the lower limit of the range, the relative deviation is equal to ((key physical parameters automatically interpreted from the image - upper limit of the range) divided by the image interpretation value). The calculated relative deviation value is compared with a preset deviation threshold. If the relative deviation is greater than the preset threshold, it is determined that there is a significant conflict in the data, and the parameter correction process is automatically triggered. Otherwise, the image interpretation value is adopted as the reliable result.

[0055] When the correction process is triggered, the following actions will be performed automatically:

[0056] Acquire the key physical parameters and their initial weights derived from the automatic image interpretation;

[0057] The median value of the text constraint boundary is obtained as the text representative value, and the credibility score of the text constraint boundary is used as its initial weight.

[0058] Retrieve historical topographic and geomorphological feature values ​​related to this parameter from the historical hydrological database and assign it an initial weight;

[0059] The initial weights of the key physical parameters, textual representative values, and historical feature values ​​automatically interpreted from the image are normalized so that the sum of their weights is 1, resulting in the final weight coefficient. Using the normalized weights, the image interpretation value, textual representative value, and historical feature value are weighted and summed respectively to calculate the final corrected value of the parameter. The above fusion calculation is repeated for all key parameters that trigger correction to generate the final parameter value.

[0060] The preset threshold is determined by statistical analysis of historical landslide dam case data and risk tolerance in conjunction with emergency decision-making scenarios. The dam height is obtained by identifying the dam crest ridgeline and the downstream dam toe and calculating the vertical distance between them. The dam crest length is determined by extracting the actual length of the three-dimensional spatial curve of the dam crest along the river channel. The initial breach width is obtained by locating the narrowest cross-section on the identified flow channel and measuring the horizontal distance between the two intact banks. The breach location is described in three-dimensional coordinates by determining the plane coordinates of the center point of the narrowest cross-section or the scour initiation point and reading its corresponding elevation from the three-dimensional model.

[0061] Rapid acquisition of multi-source heterogeneous information also includes using web crawlers to capture real-time weather and hydrological data related to landslide dams, and using these as input parameters for dynamic simulation and prediction of the breach process.

[0062] Step S3: Input the corrected physical parameters into the preset failure process calculation model to perform dynamic simulation of the failure process, and update the physical parameter values ​​of the landslide dam in combination with real-time image data during the simulation.

[0063] Specifically, the steps of the breakdown process computation model are as follows:

[0064] Step S31: The model enters the time-stepping calculation loop. At the beginning of each time step, the model calculates the outflow velocity and energy of the water flow through the breach based on the current reservoir water level and breach morphology. The greater the height difference between the water level and the bottom of the breach, the stronger the water flow energy and the larger the outflow rate.

[0065] Step S32: Based on the calculated water flow energy, assess its scouring capacity on the dam material. When the scouring force of the water flow on the dam foundation exceeds the resistance strength of the dam material itself, the bottom of the breach will be scoured and cut down. After the bottom of the breach is cut down, its sidewall will collapse under the action of gravity, thus causing the breach width to expand. The rate of cutting down and widening depends on the strength of the water flow energy and the scouring characteristics of the dam material.

[0066] Step S33: As the water flow continues to decrease, the total water storage capacity of the reservoir is relatively reduced. The model subtracts the current reservoir capacity from the total volume of the outflow and updates the reservoir water level after the decrease based on the correspondence between the reservoir capacity and the water level.

[0067] Step S34: After completing the calculation of a time step, start the dynamic boundary condition update, obtain the latest round of UAV time series images, automatically extract the measured width of the breach at that moment, and replace the calculated value of the breach width in the model for the current step with this measured value.

[0068] Step S35: Using the updated breach size, reservoir water level, and other variables as new initial conditions, the system automatically enters the next time step and repeats steps one through four. Through this cyclical process, the system dynamically simulates the continuous evolution of the breach gradually expanding, the reservoir water continuously flowing out, and the water level continuously decreasing. Finally, the model outputs the breach morphology change process, the outflow change process, and the reservoir water level decrease process over the entire time series.

[0069] Step S4: Output the time required for the flood breach, peak flow rate, and downstream flooding range based on the simulation results, and provide emergency strategies through a visualization interface based on the output results.

[0070] Specifically, after the simulation calculation is completed, the complete time series data output by the failure process calculation model is obtained, including the failure morphology change process, the discharge flow change process, and the reservoir water level drop process. Based on the failure morphology change process data, the time required from the initial moment to the complete instability of the dam is determined. The failure time is defined as the time from the initial moment... At the critical moment of dam instability The interval is calculated using the following formula:

[0071]

[0072] in, The time required for the collapse is represented by the peak flow rate and its occurrence time extracted by analyzing the data of the downstream flow rate changes. The specific calculation formula is as follows:

[0073]

[0074] in, Indicates peak traffic. Indicates the time when peak traffic occurs. This represents the instantaneous discharge flow rate at the breach as a function of time during the dam break. The discharge flow rate variation is used as a boundary condition input to a hydrodynamic model incorporating downstream topographic data. The discharge flow rate variation output from the dam break model is used as the upstream boundary condition. Simultaneously, a downstream digital elevation model is integrated to establish the hydraulic computational domain. Subsequently, an unstructured triangular mesh is used to discretize the computational domain. Initial hydraulic conditions and Manning roughness coefficients are set, and the downstream flood evolution is described using the two-dimensional Saint-Venant equations. The finite volume method adapted to the unstructured triangular mesh is used for solution. The specific calculation formulas are as follows:

[0075]

[0076] in, , The x and y components of the unit normal vector of the control volume boundary are represented by ds, which represents the boundary arc length element. The spatiotemporal distribution of water depth and flow velocity of each grid cell is obtained by iterative calculation using the finite volume method. Finally, the maximum inundation range, water depth distribution and flood arrival time during the entire simulation period are extracted to form quantitative inundation prediction data that can be used for emergency decision-making. Based on these quantitative results, the emergency command center is provided with decision-making basis for early warning duration, disaster intensity and impact range through a visualization interface.

[0077] The specific steps for iterative calculation using the finite volume method are as follows:

[0078] The downstream area is divided into a large number of closely connected triangular grids, each grid being an independent computational unit. At the initial time (t=0), an initial water depth and initial flow velocity are set for each grid, and a cyclical calculation proceeds second by second. Starting from the first second, the simulation is performed progressively forward with a time step t. For each second, the following steps are executed:

[0079] Step S41: For each side of each triangle, check the water depth and flow rate of the cells on both sides of the side. Based on this information, calculate how much water flows from one side to the other in that brief second.

[0080] Step S42: For each triangular cell, sum up all the inflow and outflow of water on its three sides. Based on the net change in water volume, update the new water depth and new flow rate of this cell in the next second. New water volume = old water volume + total inflow - total outflow.

[0081] Step S43: After updating the water volume, deduct the energy lost by the water flow based on the roughness of the riverbed, and adjust the direction and speed of the water flow according to the topographic slope.

[0082] Step S44: Update the time to the next second. Execute step S41, using the new state of all cells as the starting point, continue to calculate the changes in the next second, and repeat the above steps until the total simulation time ends. Output the water depth and flow velocity of each cell at each moment, thereby generating a complete flood range map, water depth distribution map, and flood evolution animation of the entire downstream area.

[0083] After the simulation calculation is completed, the model outputs the prediction results and presents them to the emergency command center through a visualization interface. The output results include: key time nodes for dam failure: predicting the total time required from the current moment to the complete failure of the dam; flood hydrograph: predicting the peak flow at the breach and its occurrence time; downstream impact analysis: combining downstream geographic information data to predict the flood evolution process and generate flood arrival time, inundation range and inundation depth distribution maps at different downstream locations.

[0084] Among them, determining the image scale in the intelligent interpretation and calculation of key physical parameters of the collapse is achieved by calculating the ground resolution using UAV flight altitude metadata or by identifying reference objects of known size in the image. The interpretation capacity is achieved by identifying the shoreline in the image and calculating the volume in combination with the digital elevation model (DEM) data of the area.

[0085] Example 2

[0086] This embodiment also provides a simulation emergency system for rapid acquisition of real-time information on landslide dam failure, including:

[0087] Data acquisition module: Use drones to collect image data and point cloud data of the landslide dam area, and use web crawlers to obtain web text information related to the landslide dam; all point cloud data are matched to the same spatial coordinate system through the iterative nearest point algorithm, and based on the matched point cloud data, a triangular terrain model of the landslide dam area is constructed through the triangulation algorithm.

[0088] Parameter correction module: Based on the constructed 3D terrain model, the image data is interpreted to obtain the physical parameters of the landslide dam; and the physical parameters are corrected through a multi-source data fusion mechanism.

[0089] Breach Simulation Module: Input the corrected physical parameters into the preset breach process calculation model to perform dynamic simulation of the breach process, and update the physical parameter values ​​of the landslide dam in combination with real-time image data during the simulation.

[0090] Simulation emergency module: Based on the simulation results, outputs the time required for the flood failure, peak flow rate, and downstream flooding range, and provides emergency strategies through a visual interface based on the output results.

[0091] This embodiment is based on the same inventive concept as Embodiment 1, and will not be repeated here.

[0092] Example 3

[0093] This embodiment also provides a computer device, including one or more processors, a memory, and one or more programs, wherein the one or more programs are stored in the memory and configured to be executed by the one or more processors, and the programs, when executed by the processors, implement the steps of the simulation emergency method for rapid acquisition of real-time information on landslide dam failure.

[0094] Example 4

[0095] This embodiment also provides a computer-readable storage medium storing a computer program thereon, which, when executed by a processor, implements the steps of the simulation emergency method based on rapid acquisition of real-time information on landslide dam failure.

Claims

1. A simulation-based emergency response method for rapid acquisition of real-time information on landslide dam failure, characterized in that, include: UAVs were used to collect image data and point cloud data of the landslide dam area, and web crawlers were used to obtain web text information related to the landslide dam. All point cloud data were matched to the same spatial coordinate system through the iterative nearest point algorithm. Based on the matched point cloud data, a three-dimensional terrain model of the landslide dam area was constructed through the triangulation algorithm. The image data is interpreted based on the constructed 3D terrain model to obtain the physical parameters of the landslide dam; and the physical parameters are corrected through a multi-source data fusion mechanism. The corrected physical parameters are input into the preset failure process calculation model to perform dynamic simulation of the failure process, and the physical parameter values ​​of the landslide dam are updated in combination with real-time image data during the simulation. Based on the simulation results, the time required for the flood breach, the peak flow rate, and the downstream inundation range are output, and emergency strategies are provided through a visualization interface based on the output results.

2. The simulation-based emergency response method for rapid acquisition of real-time information on landslide dam failure as described in claim 1, characterized in that, The method further includes preprocessing the point cloud data, removing outliers and noise points from the point cloud data using a statistical filtering algorithm, for each point in the point cloud. Calculate the average distance from its k nearest neighbors. Calculate the mean of all average distances in the point cloud. and standard deviation Set distance threshold range Eliminate all that satisfy point , where n represents the standard deviation multiplier.

3. The simulation-based emergency response method for rapid acquisition of real-time information on landslide dam failure as described in claim 1, characterized in that, The process of constructing a triangular terrain model of the landslide dam area based on the matched point cloud data and using a triangulation algorithm includes: projecting the three-dimensional point cloud onto a two-dimensional horizontal plane, performing triangulation calculations based on its X and Y coordinates, and defining a vertex index list that defines which vertices constitute each triangular facet; reassigning the elevation values ​​corresponding to each vertex in the point cloud to each vertex in the two-dimensional triangle; restoring the two-dimensional planar triangulation to a three-dimensional triangulation model; and generating a three-dimensional terrain model.

4. The simulation-based emergency response method for rapid acquisition of real-time information on landslide dam failure as described in claim 1, characterized in that, The step of correcting the physical parameters through a multi-source data fusion mechanism specifically includes: parsing the acquired network text information, extracting numerical information on the dam height, water level, and breach size of the landslide dam, quantifying the extracted numerical information into a unified parameter value range, and constructing a parameter value range with both upper and lower bounds equal to the value if the text description is a single numerical value; calculating a credibility score for the value range, outputting a credibility score between 0 and 1 through a weighted scoring function, and assigning it to the text constraint boundary; The physical parameters of the landslide dam interpreted from the image data are compared in real time with the text constraint boundaries to calculate the relative deviation value. The calculated relative deviation value is then compared with a preset deviation threshold. If the relative deviation is greater than the preset threshold, a data conflict is identified, triggering a parameter correction process. Otherwise, the physical parameter values ​​derived from the image data interpretation are used as the reliable result. The correction process includes: Obtain the physical parameters and initial weights derived from the image data; The median value of the text constraint boundary is obtained as the text representative value, and the credibility score of the text constraint boundary is used as its initial weight. Retrieve historical topographic and geomorphological feature values ​​related to physical parameters from the historical hydrological database and assign them initial weights; The initial weights of the physical parameters, textual representative values, and historical feature values ​​automatically interpreted from the image data are normalized so that the sum of the three weights is 1, resulting in the final weight coefficient. Using the normalized weights, the physical parameters, textual representative values, and historical feature values ​​are weighted and summed to calculate the final corrected value of the parameter. The above calculation is repeated for all physical parameters that trigger correction to generate the final parameter value.

5. The simulation-based emergency response method for rapid acquisition of real-time information on landslide dam failure as described in claim 1, characterized in that, The specific calculation process of the breakdown process calculation model is as follows: The calculation model of the breach process enters a time-stepping calculation loop. At the beginning of each time step, the model calculates the outflow velocity and energy of the water flowing through the breach based on the current reservoir water level and breach morphology. Based on the calculated water flow energy, its scouring capacity on the dam material is assessed. When the scouring force of the water flow on the dam foundation exceeds the resistance strength of the dam material itself, the bottom of the breach will be scoured down. After the bottom of the breach is cut down, its sidewall will collapse under the action of gravity, thus causing the breach width to expand. The model subtracts the current reservoir capacity from the total volume of water discharged, and updates the reservoir water level after the drop based on the correspondence between reservoir capacity and water level. After completing the calculation for one time step, real-time UAV time-series image data is acquired, the measured width of the breach at that moment is automatically extracted, and the measured width is used to replace the calculated value of the breach width for the current step in the model. The updated breach size, reservoir water level, and other variables are used as new inputs. The system automatically moves to the next time step and repeats the above operations. Finally, the model outputs the breach morphology change process, the discharge flow change process, and the reservoir water level drop process over the entire time series.

6. A simulation-based emergency response method for rapid acquisition of real-time information on landslide dam failure, as described in claim 5, is characterized in that... The model outputs the entire time series, including the process of breach morphology change, the process of discharge flow change, and the process of reservoir water level drop. Based on the breach morphology change data, the time required from the initial moment to the complete instability of the dam body and the collapse is determined. Based on the discharge flow change data, the peak flow and its occurrence time are extracted.

7. A simulation-based emergency response method for rapid acquisition of real-time information on landslide dam failure as described in claim 6, characterized in that, The extraction of peak flow and its occurrence time based on the data of the downstream flow change process includes: using the instantaneous downstream flow change of the breach as a function of time, inputting the downstream flow change process as a boundary condition into a hydrodynamic model containing downstream topographic data, using the downstream flow change process output by the breach process calculation model as an upstream boundary condition, establishing a hydraulic calculation domain in combination with the downstream digital elevation model, discretizing the calculation domain using an unstructured triangular mesh, setting initial hydraulic conditions and Manning roughness coefficients, using the two-dimensional Saint-Venant equations to represent the downstream flood evolution, solving the problem using the finite volume method adapted to the unstructured triangular mesh, obtaining the spatiotemporal distribution of water depth and velocity of each mesh cell through iterative calculation using the finite volume method, and finally extracting the maximum inundation range, water depth distribution, and flood arrival time during the entire simulation period.

8. A simulation-based emergency response system for rapid acquisition of real-time information on landslide dam failure, characterized in that, include: Data acquisition module: Use drones to collect image data and point cloud data of the landslide dam area, and use web crawlers to obtain web text information related to the landslide dam; all point cloud data are matched to the same spatial coordinate system through the iterative nearest point algorithm, and based on the matched point cloud data, a triangular terrain model of the landslide dam area is constructed through the triangulation algorithm. Parameter correction module: Based on the constructed 3D terrain model, the image data is interpreted to obtain the physical parameters of the landslide dam; The physical parameters are then corrected using a multi-source data fusion mechanism. Breach Simulation Module: Input the corrected physical parameters into the preset breach process calculation model to perform dynamic simulation of the breach process, and update the physical parameter values ​​of the landslide dam in combination with real-time image data during the simulation. Simulation emergency module: Based on the simulation results, outputs the time required for the flood failure, peak flow rate, and downstream flooding range, and provides emergency strategies through a visual interface based on the output results.

9. A computer device, characterized in that, It includes one or more processors, a memory, and one or more programs, wherein the one or more programs are stored in the memory and configured to be executed by the one or more processors, and the programs, when executed by the processors, implement the steps of a simulation emergency method for rapid acquisition of real-time information on landslide dam failure as described in any one of claims 1-7.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the steps of a simulation emergency method for rapid acquisition of real-time information on landslide dam failure as described in any one of claims 1-7.

Images