Loess plateau chain disaster multi-scene dynamic prediction method, system and equipment

Abstract

The invention belongs to the technical field of loess disaster simulation and stability prediction, and relates to a loess plateau chain disaster multi-scene dynamic prediction method, system and device. The method comprises the following steps: acquiring real-time data and historical data of a to-be-measured region, and preprocessing the historical data to obtain a database; constructing a numerical model and a visual terrain model according to the real-time data, and determining a parameter threshold value of disaster occurrence according to historical data and the numerical model; judging whether disaster dynamic evolution is carried out or not by taking the parameter threshold value as a judgment basis of disaster dynamic evolution; when disaster dynamic evolution is carried out, the real-time landslide stability coefficient, the sliding surface, the debris flow movement range and the accumulation height of a to-be-detected area are predicted according to real-time data and are displayed on the visual terrain model, simulation and prediction of geological disasters are achieved, disaster development and the threat range are dynamically demonstrated, disaster risks can be rapidly known, and the disaster development and threat range can be dynamically demonstrated. Effective disaster prevention and reduction measures are timely made, and disaster loss is reduced.

Description

Technical Field

[0001] The present invention belongs to the technical field of loess disaster simulation and stability prediction, and specifically relates to a multi-scenario dynamic prediction method, system and equipment for chain disasters on the Loess Plateau. Background Art

[0002] The cascading effects of chain disasters are particularly pronounced in infrastructure-intensive areas, such as urban agglomerations, transportation hubs, and energy bases. For example, on the Loess Plateau, collapsing disasters can trigger slope instability, pipeline ruptures, and road collapses, which in turn trigger secondary disasters such as landslides and debris flows, creating a chain reaction of "collapse-collapse-landslide-debris flow-infrastructure damage."

[0003] Existing technologies have three major flaws: 1) Traditional numerical models mostly use independent simulation of single disasters and do not consider multi-disaster simulation; 2) Data fusion dimensions are single, and loess parameter data have a large amount of uncertainty and lack integration with real-time monitoring data; 3) The visualization system has not established a spatiotemporal mapping relationship between infrastructure warning and disaster evolution, resulting in a lack of scientific basis for emergency evacuation route planning and emergency time.

[0004] Therefore, there is an urgent need for a comprehensive system that can realize multi-source data fusion, multi-disaster type collaborative deduction and visualization display, so as to improve the prediction accuracy and emergency response capabilities of major chain disasters, identify the development and transformation process of disaster chains, and thus serve disaster prevention and control. Summary of the Invention

[0005] The purpose of the present invention is to provide a multi-scenario dynamic prediction method, system and equipment for chain disasters on the Loess Plateau, so as to solve the technical problems in the existing technology that cannot realize multi-source data fusion, multi-disaster type collaborative deduction and visualization display.

[0006] In order to achieve the above object, the present invention adopts the following technical solutions:

[0007] In a first aspect, the present invention provides a multi-scenario dynamic prediction method for chain disasters on the Loess Plateau, comprising:

[0008] Obtain real-time data and historical data of the area to be measured, and pre-process the historical data to obtain a database;

[0009] Based on real-time data, numerical models and visual terrain models are constructed respectively. The parameter thresholds for disaster occurrence are determined based on historical data and numerical models. The parameter thresholds are used as the basis for judging the dynamic evolution of disasters and determining whether to conduct dynamic evolution of disasters.

[0010] When the disaster is dynamically evolving, the real-time landslide stability coefficient, sliding surface, debris flow movement range and accumulation height of the test area are predicted based on real-time data, and the real-time landslide stability coefficient, sliding surface, debris flow movement range and accumulation height of the test area are displayed on the visual terrain model.

[0011] Preferably, the real-time data of the area to be measured includes: remote sensing satellite images, drone photography images, DEM data maps, real-time loess shear strength parameters, real-time rainfall and real-time landslide dynamic movement parameters; the historical disaster data includes historical loess shear strength parameters, historical rainfall and historical landslide dynamic movement parameters;

[0012] The means and variances of historical loess shear strength parameters, historical rainfall, and historical landslide dynamic movement parameters are calculated to form a database.

[0013] Preferably, constructing a visual terrain model based on real-time data specifically includes:

[0014] Generate terrain contour lines based on DEM data and construct a three-dimensional digital elevation model; obtain three-dimensional terrain data based on drone photography images;

[0015] A visual terrain model is constructed based on three-dimensional digital elevation model, three-dimensional terrain data and remote sensing satellite images.

[0016] Preferably, the parameter thresholds for disaster occurrence are determined based on historical data, real-time data, and numerical models, specifically:

[0017] A first optimization function is constructed based on the mean and variance of the historical loess shear strength parameters, and the threshold value of the loess shear strength parameters in the test area is inferred through the first optimization function;

[0018] A second optimization function is constructed according to historical landslide dynamic motion parameters, and the threshold value of the landslide dynamic motion parameters in the test area is inferred through the second optimization function.

[0019] Preferably, the method uses parameter thresholds as a basis for judging the dynamic evolution of disasters, predicts the real-time landslide stability coefficient, sliding surface, debris flow movement range and accumulation height of the area to be measured, and displays the real-time landslide stability coefficient, sliding surface, debris flow movement range and accumulation height of the area to be measured on a visual terrain model; including:

[0020] S301: using the Bayesian Monte Carlo Markov chain method for the real-time loess shear strength parameter, the real-time rainfall, and the real-time landslide dynamic motion parameter to obtain the posterior distribution of each parameter; calculating the mean of the real-time loess shear strength parameter, the mean of the real-time rainfall, the mean of the real-time landslide dynamic motion parameter, the variance of the real-time loess shear strength parameter, the variance of the real-time rainfall, and the variance of the real-time landslide dynamic motion parameter based on the posterior distribution of each parameter; and calculating the system failure probability of the landslide when the real-time parameter mean of the real-time loess shear strength parameter is less than the loess shear strength parameter threshold;

[0021] S302: Determine the stability of each landslide based on the system failure probability, and judge whether each landslide has experienced unstable movement based on the stability;

[0022] S303: In the landslide where the instability movement occurs, when the real-time landslide dynamic motion parameter mean is less than the landslide dynamic motion parameter threshold, calculating the sliding surface instability probability of the sliding surface in the landslide;

[0023] S304: extracting the sliding surface with the highest probability of instability in the landslide where instability occurs;

[0024] S305: Obtaining real-time landslide dynamic motion parameters of the extracted sliding surface, and calculating the debris flow movement range and accumulation height when unstable motion occurs;

[0025] S306: Display the sliding surface, the systematic failure probability of the landslide, the movement range and accumulation height of the debris flow on the visual terrain model.

[0026] Preferably, the calculation of the system failure probability of landslide is specifically as follows:

[0027]

[0028] f′(θ)=FS(θ)-1

[0029] Where, P f represents the systematic failure probability of the landslide; Q represents the number of samples with f′(θ) < 0, and I represents the total number of sampling times; θ represents the loess shear strength parameter, and f′(θ) represents the landslide stability function; FS is the landslide stability coefficient calculated using the strength reduction method for the loess shear strength parameter.

[0030] Preferably, the extraction of the sliding surface with the highest probability of sliding surface instability in the landslide where unstable movement occurs adopts the following formula:

[0031] P=Φ(-β)=1-Φ(β)

[0032] Where P represents the probability of a landslide failing along the sliding surface; Φ represents the standard normal cumulative distribution function; and β represents the distance from the verification point to the origin in standard normal space. The point with the highest probability of failure is found within the failure points. Once this point is found, the sliding surface corresponding to it has the highest probability of failure.

[0033] A second aspect of the present invention provides a multi-scenario dynamic prediction system for chain disasters on the Loess Plateau, comprising:

[0034] A data preprocessing unit is used to obtain real-time data and historical data of the area to be measured, and preprocess the historical data to obtain a database;

[0035] The judgment unit is used to sequentially construct a numerical model and a visual terrain model based on real-time data, and determine the parameter threshold for disaster occurrence based on historical data and the numerical model; the parameter threshold is used as the basis for judging the dynamic evolution of the disaster, and whether to conduct the dynamic evolution of the disaster;

[0036] The prediction unit is used to predict the real-time landslide stability coefficient, sliding surface, debris flow movement range and accumulation height of the test area based on real-time data when the disaster is dynamically evolving, and to display the real-time landslide stability coefficient, sliding surface, debris flow movement range and accumulation height of the test area on a visual terrain model.

[0037] A third aspect of the present invention provides an electronic device, characterized in that it includes a processor and a memory, and the processor is used to execute a computer program stored in the memory to implement any one of the above-mentioned methods for dynamic prediction of multiple scenarios of chain disasters on the Loess Plateau.

[0038] A fourth aspect of the present invention provides a computer-readable storage medium, characterized in that the computer-readable storage medium stores at least one instruction, and when the at least one instruction is executed by a processor, it implements any one of the above-mentioned methods for dynamic prediction of multiple scenarios of chain disasters on the Loess Plateau.

[0039] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0040] The method disclosed in this application is based on multi-source data collection, fully utilizing various data resources to provide more comprehensive and accurate input parameters for the model. It constructs a numerical model and a visual terrain model using real-time data, determines the parameter thresholds for disaster occurrence based on historical data and the numerical model, and uses the parameter thresholds as the basis for judging the dynamic evolution of disasters to determine whether to conduct dynamic evolution of disasters. It also simulates and predicts geological disasters, dynamically demonstrating the development and threat scope of disasters through a visual terrain model, and presenting the output results of the deduction and prediction modules in the form of graphics, charts, or animations, making it easier for decision makers to intuitively understand disaster risks. The complex disaster evolution process is presented in an intuitive manner, making it easier for decision makers to quickly understand disaster risks, formulate effective disaster prevention and mitigation measures in a timely manner, and reduce disaster losses.

[0041] Furthermore, the data were preprocessed using the Bayesian MCMC algorithm to reduce parameter uncertainty, solve the single dimension of data fusion, reduce the uncertainty factors of loess parameters, and improve the accuracy and reliability of disaster prediction. BRIEF DESCRIPTION OF THE DRAWINGS

[0042] The drawings in the specification, which constitute a part of this application, are used to provide a further understanding of the present invention. The exemplary embodiments of the present invention and their descriptions are used to explain the present invention and do not constitute improper limitations on the present invention.

[0043] In the attached figure:

[0044] Figure 1 is a flow chart of a method according to an embodiment of the present invention;

[0045] Figure 2 Schematic diagram of drone imagery and DEM imagery collected by the data acquisition module proposed in the embodiment of the present invention

[0046] Figure 3 A rainfall map of the study area collected by the data acquisition module according to an embodiment of the present invention;

[0047] Figure 4 This is a priori distribution diagram of data calculated by the data acquisition module proposed in an embodiment of the present invention; wherein (a) is the priori distribution of loess cohesion; (b) is the priori distribution of loess internal friction angle;

[0048] Figure 5 A three-dimensional real-scene model diagram created by the model building module proposed in an embodiment of the present invention;

[0049] Figure 6 The distribution diagram of failure points and verification points in the failure domain proposed in the embodiment of the present invention;

[0050] Figure 7 is a system block diagram proposed in an embodiment of the present invention;

[0051] Figure 8 : This is a Monte Carlo sampling diagram of the active learning Kriging algorithm of the data processing and fusion module proposed in an embodiment of the present invention; wherein (a) is a sampling diagram of the cohesion of natural loess; (b) is a sampling diagram of the internal friction angle of natural loess; (c) is a sampling diagram of the cohesion of saturated loess; (d) is a sampling diagram of the internal friction angle of saturated loess;

[0052] Figure 9 This is a Monte Carlo sampling pool diagram of the data processing and fusion module proposed in an embodiment of the present invention;

[0053] Figure 10 This is a flow chart of active learning and instability probability of the data processing and fusion module proposed in an embodiment of the present invention;

[0054] Figure 11 The data processing and fusion module of the embodiment of the present invention is a landslide source and a pre-slide DEM map;

[0055] Figure 12 This is a visualization module landslide mudflow range map proposed in an embodiment of the present invention;

[0056] Figure 13 This is a visualization module of the impact of landslide and mudflow disasters on infrastructure proposed in an embodiment of the present invention;

[0057] Figure 14 Schematic diagram of an electronic device according to an embodiment of the present invention. DETAILED DESCRIPTION

[0058] The present invention will be described in detail below with reference to the accompanying drawings and in combination with embodiments. It should be noted that, in the absence of conflict, the embodiments and features in the embodiments of the present application can be combined with each other.

[0059] The following detailed description is an exemplary description, which is intended to provide further detailed description of the present invention. Unless otherwise indicated, all technical terms used in the present invention have the same meaning as those generally understood by those skilled in the art. The terms used in the present invention are only for describing specific embodiments, and are not intended to limit the exemplary embodiments according to the present invention.

[0060] See also Figure 1 This application discloses a multi-scenario dynamic prediction method for chain disasters on the Loess Plateau, including:

[0061] S1: Obtain real-time data and historical data of the area to be measured, and pre-process the historical data to obtain a database;

[0062] S2: Build numerical models and visual terrain models based on real-time data, and determine the parameter thresholds for disaster occurrence based on historical data and numerical models. Use the parameter thresholds as the basis for judging the dynamic evolution of disasters and determine whether to conduct dynamic evolution of disasters.

[0063] S3: When the disaster is dynamically evolving, the real-time landslide stability coefficient, sliding surface, debris flow movement range and accumulation height of the test area are predicted based on real-time data, and the real-time landslide stability coefficient, sliding surface, debris flow movement range and accumulation height of the test area are displayed on the visual terrain model.

[0064] The dynamic prediction method disclosed in this application constructs a numerical model and a visual terrain model based on real-time data, determines the parameter thresholds for disaster occurrence based on historical data and the numerical model; uses the parameter thresholds as the basis for judging the dynamic evolution of disasters to determine whether to carry out dynamic evolution of disasters; realizes the simulation and prediction of geological disasters, dynamically demonstrates the development of disasters and the scope of threats through visual terrain models, and displays the output results of the deduction and prediction modules in the form of graphics, charts or animations, so as to facilitate decision makers to intuitively understand disaster risks.

[0065] In some embodiments, the real-time data of the area to be measured includes: remote sensing satellite images, drone photography images, DEM data maps, real-time loess shear strength parameters, real-time rainfall, and real-time landslide dynamic motion parameters; the historical disaster data includes historical loess shear strength parameters, historical rainfall, and historical landslide dynamic motion parameters; the landslide dynamic motion parameters are pore water pressure ratio coefficient and internal friction angle;

[0066] The means and variances of historical loess shear strength parameters, historical rainfall, and historical landslide dynamic movement parameters are calculated to form a database.

[0067] In some embodiments, constructing a visual terrain model based on real-time data specifically includes:

[0068] S201: Generate terrain contour lines based on the DEM data map and construct a three-dimensional digital elevation model; obtain three-dimensional terrain data based on drone photography images;

[0069] S202: Constructing a visual terrain model based on the three-dimensional digital elevation model, the three-dimensional terrain data and the remote sensing satellite image.

[0070] In some embodiments, the parameter threshold for disaster occurrence is determined based on historical data, real-time data, and numerical models, specifically:

[0071] A first optimization function is constructed based on the mean and variance of the historical loess shear strength parameters, and the threshold value of the loess shear strength parameters in the test area is inferred through the first optimization function;

[0072] A second optimization function is constructed according to historical landslide dynamic motion parameters, and the threshold value of the landslide dynamic motion parameters in the test area is inferred through the second optimization function.

[0073] Further preferably, the first optimization function includes:

[0074] Mean error function:

[0075] f1(θ)=(θ-U) T c -1 (θ-U)

[0076] Stability coefficient error function:

[0077] f2(θ)=FS(θ)-1

[0078] Where f1(θ) represents the mean error; θ is the posterior distribution of the loess shear strength parameter in the deduction area, U is the parameter mean, represents the inverse matrix of the parameter covariance matrix, and T represents the matrix transpose. Assuming no correlation between any two parameters, the covariance matrix c is a diagonal matrix consisting of the squared standard deviations of the shear strength parameters; f2(θ) represents the stability coefficient error; and FS is the landslide stability coefficient calculated using the FLAC3D finite difference strength reduction method for loess parameters.

[0079] Further preferably, the second optimization function is:

[0080] g1(θ′)=[HY(θ′)] T [HY(θ′)]

[0081]

[0082] The specific derivation process is as follows:

[0083] The instability of loess landslides is closely related to the rise of groundwater levels and the increase of pore water pressure. The landslide movement uses the friction rheological model formula built into MASSFLOW as shown above. Due to the liquefaction movement of loess, the loess cohesion c is assumed to be 0. The density of loess is usually 1500kg / m 3 , g is the local acceleration of gravity, usually 9.8m / s 2 ; h is the height of the moving material at any time and position during its movement, so the pore water pressure ratio coefficient λ and the internal friction angle As the variable θ′, and perform threshold inversion on these two parameter values.

[0084]

[0085] Get the mean of θ′ and standard deviation T is the transpose of the matrix. Select multiple representative observation points within the landslide accumulation area to control the spatial accumulation characteristics of the landslide and obtain the final accumulation thickness H observed at the observation points, H = [h1,h2…,h n ] T ,MassFlow simulation is used to reproduce the movement and accumulation process after the landslide instability, and the final accumulation thickness Y(θ′) simulated at the observation point is obtained, which is expressed by the following formula:

[0086] Y(θ′)=[y1(θ′),y2(θ′)...,y n (θ′)] T

[0087] Where y i (θ) represents the final accumulation thickness simulated at the i-th observation point.

[0088] Based on the observed value H and the simulated value Y(θ′), in order to quantify the degree of difference between the two, the first optimization function in the second optimization function is:

[0089] g1(θ′)=[HY(θ′)] T [HY(θ′)]

[0090] Different parameter combinations may lead to similar or even identical simulation results. In theory, the reasonable value of the optimal parameter group should be within the value range of its prior probability interval and relatively close to its prior mean. The second optimization error function in the second optimization function can be expressed as follows:

[0091]

[0092] In some embodiments, the method uses parameter thresholds as a basis for judging the dynamic evolution of disasters, predicts the real-time landslide stability coefficient, sliding surface, debris flow movement range, and accumulation height of the area to be tested, and displays the real-time landslide stability coefficient, sliding surface, debris flow movement range, and accumulation height of the area to be tested on a visual terrain model; including:

[0093] S301: using the Bayesian Monte Carlo Markov chain method for the real-time loess shear strength parameter, the real-time rainfall, and the real-time landslide dynamic motion parameter to obtain the posterior distribution of each parameter; calculating the mean of the real-time loess shear strength parameter, the mean of the real-time rainfall, the mean of the real-time landslide dynamic motion parameter, the variance of the real-time loess shear strength parameter, the variance of the real-time rainfall, and the variance of the real-time landslide dynamic motion parameter based on the posterior distribution of each parameter; and calculating the system failure probability of the landslide when the real-time parameter mean of the real-time loess shear strength parameter is less than the loess shear strength parameter threshold;

[0094] S302: Determine the stability of each landslide based on the system failure probability, and judge whether each landslide has experienced unstable movement based on the stability;

[0095] S303: In the landslide where the instability movement occurs, when the real-time landslide dynamic motion parameter mean is less than the landslide dynamic motion parameter threshold, calculating the sliding surface instability probability of the sliding surface in the landslide;

[0096] S304: extracting the sliding surface with the highest probability of instability in the landslide where instability occurs;

[0097] S305: Obtaining real-time landslide dynamic motion parameters of the extracted sliding surface, and calculating the debris flow movement range and accumulation height when unstable motion occurs;

[0098] S306: Display the sliding surface, the systematic failure probability of the landslide, the movement range and accumulation height of the debris flow on the visual terrain model.

[0099] In some embodiments, the calculation of the system failure probability of landslide is specifically as follows:

[0100]

[0101] f′(θ)=FS(θ)-1

[0102] Where, P f represents the systematic failure probability of the landslide; Q represents the number of samples with f′(θ) < 0, and I represents the total number of sampling times; θ represents the loess shear strength parameter, and f′(θ) represents the landslide stability function; FS is the landslide stability coefficient calculated using the strength reduction method for the loess shear strength parameter.

[0103] In some embodiments, the extraction of the sliding surface with the highest probability of instability in the landslide where instability occurs is performed using the following formula:

[0104] P=Φ(-β)=1-Φ(β)

[0105] Where P represents the probability of landslide failure along the sliding surface; Φ represents the standard normal cumulative distribution function; β represents the distance from the verification point to the origin in the standard normal space. Figure 6 , based on the active learning Kriging proxy model algorithm, MCS simulation is performed, and the points with stability coefficients less than 1 are found among all sampling points, which are called failure points; the distances from all failure points to the origin are calculated in the standard normal space, and the point with the smallest corresponding distance is defined as the verification point; through NATAF transformation, the verification point is converted to physical space and substituted into the above formula to obtain the instability probability corresponding to the verification point (that is, the sliding surface), and the sliding surface with the largest instability probability is extracted.

[0106] Example 1

[0107] In order to protect the safety of life and property of the people on the Loess Plateau, the development and prediction of the loess disaster chain is particularly important. In order to obtain the dynamic development process and prediction of disasters, and then effectively predict the geological chain disasters on the Loess Plateau, this application proposes a multi-scenario dynamic deduction method for the loess disaster chain, taking the "rainfall-wetting-subsidence-landslide-mudflow" disaster chain as an example.

[0108] The method comprises the following steps:

[0109] S1: First, through sensors, satellite remote sensing and UAV LiDAR, remote sensing satellite images, UAV photography, DEM data maps, and historical monitoring data are collected to obtain loess parameters, rainfall and other information in the deduction area; loess parameters, rainfall and other information are statistically analyzed in the module to obtain parameter prior distribution, as well as mean and variance and stored in TXT files for later use; the collected DEM data and image data are processed by calling ARCGIS software and saved as data to be used. The data in this embodiment is the Heifangtai area of Yongjing, and the drone-photographed images and DEM images collected by the data are as follows: Figure 2 As shown in Figure 2, rainfall information is obtained through local monitoring instruments, such as Figure 3 shown.

[0110] The statistical results show that the mean cohesion of natural loess in this area is 28.83 kPa, the standard deviation is 5.45 kPa, the mean internal friction angle is 22.26°, the standard deviation is 4.56°, and the statistical prior distribution of both is normal distribution, as shown in the following example: Figure 4 shown.

[0111] S2: Use ARCGIS to import DEM data, use Arctoolbox to generate terrain contour lines, build a 3D digital elevation model, obtain 3D terrain data based on drone images, and build a visual terrain model based on the 3D digital elevation model, 3D terrain data and remote sensing satellite images for visualization. Figure 5 As shown. Based on the mean and variance of the loess shear strength parameters, namely cohesion and internal friction angle, obtained from the real-time parameter mean and real-time parameter variance obtained from the S1 preprocessing, a first optimization function in the genetic algorithm is constructed, and the shear strength parameter threshold of the deduction area is inverted through the first optimization function. Based on the mean and variance of the landslide dynamic movement parameters, namely pore water pressure ratio coefficient and internal friction angle, obtained from the real-time parameter mean and real-time parameter variance obtained from the S1 preprocessing, a second optimization function in the genetic algorithm is constructed, and the shear strength parameter threshold of the deduction area is inverted through the second optimization function.

[0112] When the real-time parameter mean of the real-time loess shear strength parameter is less than the loess shear strength parameter threshold, the system failure probability of the landslide is calculated according to the parameter posterior distribution; the stability of each landslide is determined according to the system failure probability, and whether each landslide has undergone unstable movement is judged according to the stability; in the landslide where unstable movement has occurred, when the real-time parameter mean of the real-time landslide dynamic motion parameter is less than the landslide dynamic motion parameter threshold, the sliding surface instability probability of the sliding surface in the landslide is calculated according to the parameter posterior distribution; the sliding surface with the highest sliding surface instability probability in the landslide where unstable movement has occurred is extracted; the source area of the landslide movement (i.e., the sliding surface with the highest sliding surface instability probability) is extracted. Figure 11 As shown; the real-time landslide dynamic motion parameters of the extracted sliding surface are obtained, and the debris flow movement range and accumulation height when unstable movement occurs are calculated; the sliding surface, landslide failure probability, debris flow movement range and accumulation height are displayed on the visual terrain model.

[0113] Furthermore, the Markov chain is Figure 8 As shown. A large amount of numerical calculations are required for Bayesian updating, so the system establishes a Kriging active learning proxy model algorithm, and uses active learning training to replace the large amount of computing time spent on numerical simulation. Since Bayesian updating requires a large amount of numerical calculations, the system has a built-in Kriging active learning proxy model algorithm, and uses active learning training to replace the large amount of computing time spent on numerical simulation. The idea of active learning is: to select new sample points by combining the learning function with the current training model. First, the initial sample points must be generated in the initial standard normal space, and then the requirement for selecting new sample points is generally to be located at the junction of the failure area and the safe area, away from existing and calculated sample points, to avoid repeated sampling. According to the active learning function proposed by Echard et al.:

[0114]

[0115] argmin(·) returns the optimal parameter that minimizes the function value during the optimization process, u T , are regarded as variable parameters and need to be substituted into the calculation one by one; It is a functional agent model. For classifiers such as support vector machines, for u represents the optimal new sample point selected from the sample pool T, which is generated by Monte Carlo sampling. Figure 9 As shown in the figure, the stability coefficient is 0 on the limit plane formed by the limit state function, the stability coefficient calculated for the safety point in the safety domain is greater than 1, and the stability coefficient of the failure point in the failure domain is less than 1.

[0116] f′(θ)=FS(θ)-1

[0117] Where θ represents the loess shear strength parameter, and f′(θ) represents the landslide stability function.

[0118] By actively learning the Kriging algorithm proxy model instead of numerical calculation, the parameter posterior distribution is obtained, the mean is calculated and stored, and then the system failure probability of the landslide is calculated based on the Monte Carlo method.

[0119] According to the Monte Carlo method, the system failure probability of landslide can be calculated as:

[0120]

[0121] P f represents the system failure probability of the landslide, Q represents the number of samples with f′(θ)<0, and I represents the total number of sampling times. This algorithm samples a total of 200,000 times, that is, L=200,000. The calculation process is as follows Figure 10 shown.

[0122] In some embodiments, the present application can also use the results of landslide simulations to initiate assessments of the impact of landslide disasters on infrastructure-intensive areas. Dynamic simulations are performed to simulate the evolution of disasters under different scenarios. Based on these simulations, a disaster risk assessment report is generated, including the likelihood of a disaster occurring, the scope of impact, estimated losses, and the extent of damage to infrastructure, providing a scientific basis for emergency response and disaster prevention and mitigation.

[0123] In some embodiments, the visual terrain model can also display the calculation results of each deduction link, such as collapsible settlement, mudflow range and accumulation height, such as Figure 12 Furthermore, the response of different disaster scenarios to infrastructure (houses) in the simulation phase can also be demonstrated, such as Figure 13 .

[0124] The above examples show that the prediction results of this method are excellent. It can more accurately predict the occurrence of disaster chains and provide the failure probability of landslides, unstable sliding surfaces, debris flow movement range and accumulation height.

[0125] Example 2

[0126] like Figure 7 As shown, based on the same inventive concept as the above embodiment, the present invention also provides a multi-scenario dynamic prediction system for chain disasters on the Loess Plateau, comprising:

[0127] A data preprocessing unit is used to obtain real-time data and historical data of the area to be measured, and preprocess the historical data to obtain a database;

[0128] The judgment unit is used to sequentially construct a numerical model and a visual terrain model based on real-time data, and determine the parameter threshold for disaster occurrence based on historical data and the numerical model; the parameter threshold is used as the basis for judging the dynamic evolution of the disaster, and whether to conduct the dynamic evolution of the disaster;

[0129] The prediction unit is used to predict the real-time landslide stability coefficient, sliding surface, debris flow movement range and accumulation height of the test area based on real-time data when the disaster is dynamically evolving, and to display the real-time landslide stability coefficient, sliding surface, debris flow movement range and accumulation height of the test area on a visual terrain model.

[0130] Example 3

[0131] like Figure 14 As shown, the present invention also provides an electronic device 100 for implementing a multi-scenario dynamic prediction method for chain disasters on the Loess Plateau;

[0132] The electronic device 100 includes a memory 101 , at least one processor 102 , a computer program 103 stored in the memory 101 and executable on the at least one processor 102 , and at least one communication bus 104 .

[0133] The memory 101 can be used to store a computer program 103, and the processor 102 implements the steps of any of the above-mentioned methods for dynamic prediction of multi-scenario chain disasters on the Loess Plateau by running or executing the computer program stored in the memory 101 and calling the data stored in the memory 101.

[0134] The memory 101 may mainly include a program storage area and a data storage area, wherein the program storage area may store an operating system, an application program required for at least one function (such as a sound playback function, an image playback function, etc.), etc.; the data storage area may store data (such as audio data) created according to the use of the electronic device 100. In addition, the memory 101 may include a non-volatile memory, such as a hard disk, a memory, a plug-in hard disk, a smart memory card (Smart Media Card, SMC), a secure digital (Secure Digital, SD) card, a flash card (Flash Card), at least one disk storage device, a flash memory device, or other non-volatile solid-state storage device.

[0135] The at least one processor 102 may be a central processing unit (CPU), or may be other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The processor 102 may be a microprocessor or any conventional processor. The processor 102 is the control center of the electronic device 100 and connects various parts of the entire electronic device 100 using various interfaces and lines.

[0136] The memory 101 in the electronic device 100 stores multiple instructions to implement a multi-scenario dynamic prediction method for chain disasters on the Loess Plateau. The processor 102 can execute the multiple instructions to implement:

[0137] S1: Obtain real-time data and historical data of the area to be measured, and pre-process the historical data to obtain a database;

[0138] S2: Build numerical models and visual terrain models based on real-time data, and determine the parameter thresholds for disaster occurrence based on historical data and numerical models. Use the parameter thresholds as the basis for judging the dynamic evolution of disasters and determine whether to conduct dynamic evolution of disasters.

[0139] S3: When the disaster is dynamically evolving, the real-time landslide stability coefficient, sliding surface, debris flow movement range and accumulation height of the test area are predicted based on real-time data, and the real-time landslide stability coefficient, sliding surface, debris flow movement range and accumulation height of the test area are displayed on the visual terrain model.

[0140] Example 4

[0141] If the module / unit integrated in the electronic device 100 is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the present invention implements all or part of the process in the above-mentioned embodiment method, and can also be completed by instructing the relevant hardware through a computer program. The computer program can be stored in a computer-readable storage medium, and the computer program can implement the steps of the above-mentioned method embodiments when executed by the processor. Among them, the computer program includes computer program code, and the computer program code can be in source code form, object code form, executable file or some intermediate form, etc. Computer-readable media may include: any entity or device that can carry computer program code, recording medium, U disk, mobile hard disk, magnetic disk, optical disk, computer memory and read-only memory (ROM, Read-Only Memory).

[0142] It will be understood by those skilled in the art that embodiments of the present invention may be provided as methods, systems, or computer program products. Thus, the present invention may take the form of an entirely hardware embodiment, an entirely software embodiment, or an embodiment combining software and hardware. Furthermore, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to magnetic disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0143] The present invention is described with reference to flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to embodiments of the present invention. It should be understood that each process and / or block in the flowcharts and / or block diagrams, as well as combinations of processes and / or blocks in the flowcharts and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowcharts and / or block diagrams. Figure 1 a process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0144] These computer program instructions may also be stored in a computer readable memory that can direct a computer or other programmable data processing device to work in a specific manner, so that the instructions stored in the computer readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 a process or multiple processes and / or boxes Figure 1 The function specified in one or more boxes.

[0145] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operational steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing the instructions executed on the computer or other programmable device for implementing the process. Figure 1 a process or multiple processes and / or boxes Figure 1 A step that specifies a function in one or more boxes.

[0146] Throughout this specification, references to terms such as "one embodiment," "example," or "specific example" indicate that the specific features, structures, materials, or characteristics described in conjunction with that embodiment or example are included in at least one embodiment or example of the present invention. In this specification, schematic representations of these terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in any one or more embodiments or examples.

[0147] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit it. Although the present invention has been described in detail with reference to the above embodiments, ordinary technicians in the field should understand that the specific implementation methods of the present invention can still be modified or replaced by equivalents. Any modification or equivalent replacement that does not depart from the spirit and scope of the present invention should be covered by the scope of protection of the claims of the present invention.

Claims

1. A multi-scenario dynamic prediction method for chain disasters on the Loess Plateau, characterized by: include: Obtain real-time data and historical data of the area to be measured, and pre-process the historical data to obtain a database; Build numerical models and visual terrain models based on real-time data, and determine the parameter thresholds for disaster occurrence based on historical data and numerical models; The parameter threshold is used as the basis for judging the dynamic evolution of disasters to determine whether to conduct dynamic evolution of disasters; When the disaster is dynamically evolving, the real-time landslide stability coefficient, sliding surface, debris flow movement range and accumulation height of the test area are predicted based on real-time data, and the real-time landslide stability coefficient, sliding surface, debris flow movement range and accumulation height of the test area are displayed on the visual terrain model.

2. The multi-scenario dynamic prediction method for chain disasters on the Loess Plateau according to claim 1 is characterized in that: The real-time data of the area to be measured includes: remote sensing satellite images, drone photography images, DEM data maps, real-time loess shear strength parameters, real-time rainfall and real-time landslide dynamic movement parameters; the historical disaster data includes historical loess shear strength parameters, historical rainfall and historical landslide dynamic movement parameters; The means and variances of historical loess shear strength parameters, historical rainfall, and historical landslide dynamic movement parameters are calculated to form a database.

3. The multi-scenario dynamic prediction method for chain disasters on the Loess Plateau according to claim 1 is characterized in that: Build a visual terrain model based on real-time data, including: Generate terrain contour lines based on DEM data and construct a three-dimensional digital elevation model; obtain three-dimensional terrain data based on drone photography images; A visual terrain model is constructed based on three-dimensional digital elevation model, three-dimensional terrain data and remote sensing satellite images.

4. The multi-scenario dynamic prediction method for chain disasters on the Loess Plateau according to claim 2 is characterized in that: The parameter thresholds for disaster occurrence are determined based on historical data, real-time data, and numerical models. Specifically, A first optimization function is constructed based on the mean and variance of the historical loess shear strength parameters, and the threshold value of the loess shear strength parameters in the test area is inferred through the first optimization function; A second optimization function is constructed according to historical landslide dynamic motion parameters, and the threshold value of the landslide dynamic motion parameters in the test area is inferred through the second optimization function.

5. The multi-scenario dynamic prediction method for chain disasters on the Loess Plateau according to claim 2 is characterized in that: The method uses parameter thresholds as a basis for judging the dynamic evolution of disasters, predicts the real-time landslide stability coefficient, sliding surface, debris flow movement range and accumulation height of the test area, and displays the real-time landslide stability coefficient, sliding surface, debris flow movement range and accumulation height of the test area on a visual terrain model; including: S301: using the Bayesian Monte Carlo Markov chain method for the real-time loess shear strength parameter, the real-time rainfall, and the real-time landslide dynamic motion parameter to obtain the posterior distribution of each parameter; calculating the mean of the real-time loess shear strength parameter, the mean of the real-time rainfall, the mean of the real-time landslide dynamic motion parameter, the variance of the real-time loess shear strength parameter, the variance of the real-time rainfall, and the variance of the real-time landslide dynamic motion parameter based on the posterior distribution of each parameter; and calculating the system failure probability of the landslide when the real-time parameter mean of the real-time loess shear strength parameter is less than the loess shear strength parameter threshold; S302: Determine the stability of each landslide based on the system failure probability, and judge whether each landslide has experienced unstable movement based on the stability; S303: In a landslide where unstable motion occurs, when the average value of the real-time landslide dynamic motion parameter is less than the landslide dynamic motion parameter threshold, calculating the sliding surface instability probability of the sliding surface in the landslide; S304: extracting the sliding surface with the highest probability of instability in the landslide where the instability movement occurs; S305: Obtaining real-time landslide dynamic motion parameters of the extracted sliding surface, and calculating the debris flow movement range and accumulation height when unstable motion occurs; S306: Display the sliding surface, the systematic failure probability of the landslide, the movement range and accumulation height of the debris flow on the visual terrain model.

6. A multi-scenario dynamic prediction method for chain disasters on the Loess Plateau according to claim 5, characterized in that: The system failure probability of calculating landslide is specifically: f′(θ)=FS(θ)-1 Where, P f represents the systematic failure probability of the landslide; Q represents the number of samples with f′(θ) < 0, and I represents the total number of sampling times; θ represents the loess shear strength parameter, and f′(θ) represents the landslide stability function; FS is the landslide stability coefficient calculated using the strength reduction method for the loess shear strength parameter.

7. The multi-scenario dynamic prediction method for chain disasters on the Loess Plateau according to claim 5 is characterized in that: The following formula is used to extract the sliding surface with the highest probability of instability in the landslide where instability occurs: P=Φ(-β)=1-Φ(β) Where P represents the probability of landslide failure along the sliding surface; Φ represents the standard normal cumulative distribution function; and β represents the distance from the verification point to the origin in the standard normal space.

8. A multi-scenario dynamic prediction system for chain disasters on the Loess Plateau, characterized by: include: A data preprocessing unit is used to obtain real-time data and historical data of the area to be measured, and preprocess the historical data to obtain a database; A judgment unit is used to sequentially construct a numerical model and a visual terrain model based on real-time data, and determine the parameter threshold for disaster occurrence based on historical data and the numerical model; The parameter threshold is used as the basis for judging the dynamic evolution of disasters to determine whether to conduct dynamic evolution of disasters; The prediction unit is used to predict the real-time landslide stability coefficient, sliding surface, debris flow movement range and accumulation height of the test area based on real-time data when the disaster is dynamically evolving, and to display the real-time landslide stability coefficient, sliding surface, debris flow movement range and accumulation height of the test area on a visual terrain model.

9. An electronic device, characterized in that: The method comprises a processor and a memory, wherein the processor is used to execute a computer program stored in the memory to implement a multi-scenario dynamic prediction method for chain disasters on the Loess Plateau as described in any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that The computer-readable storage medium stores at least one instruction, and when the at least one instruction is executed by the processor, it implements the multi-scenario dynamic prediction method of chain disasters in the Loess Plateau as described in any one of claims 1 to 7.

Images