Slope monitoring and early warning method and device, storage medium and electronic equipment

Abstract

The present disclosure provides a slope monitoring and early warning method and device, a storage medium and an electronic device, and relates to the technical field of reservoir safety. The method comprises: acquiring multi-modal data of different monitoring points in a target slope area; performing principal component analysis on the multi-modal data of each monitoring point to obtain corresponding slope feature scores; constructing a geographical space continuous model of the target slope area, determining the reference feature scores of the spatial neighborhood of each monitoring point in the geographical space continuous model, and determining the local deviation degree of each monitoring point according to the slope feature scores and the reference feature scores; performing spatial correlation analysis using the local deviation degrees to obtain the global deviation degree of the target slope area, and generating slope early warning information according to the global deviation degree. The present disclosure can accurately evaluate the slope stability by deeply integrating the multi-modal data.

Description

TECHNICAL FIELD

[0001] The present disclosure relates to the technical field of reservoir safety, in particular, to a slope monitoring and early warning method and device, a storage medium and an electronic device. BACKGROUND

[0002] Reservoir slope early warning is one of the important means of reservoir safety guarantee. The slope early warning usually adopts a technical means of sensor monitoring combined with early warning judgment algorithm. The sensors used include image sensors, ultrasonic detection sensors, vibration monitoring sensors, etc. The characteristic information collected includes displacement, crack monitoring, soil humidity, etc., which are all factors that can directly or indirectly affect slope geological disasters.

[0003] Further, each sensor is independently processed as a unit, and only in the subsequent analysis stage, different types of data are simply superimposed or compared. This processing method is difficult to fully mine the potential correlation between multi-modal data, ignores the internal coupling relationship of different modal data in time, space and physical characteristics, and limits the comprehensive perception ability of slope early warning.

[0004] Therefore, it is urgent to provide a more efficient and scientific method to deeply process multi-modal data in reservoir slope monitoring, thereby improving the accuracy and reliability of slope monitoring and early warning.

[0005] It should be noted that the information disclosed in the above background section is only used to strengthen the understanding of the background of the present disclosure, and therefore can include information that does not constitute prior art known to those of ordinary skill in the art. SUMMARY

[0006] The purpose of the embodiments of the present disclosure is to provide a slope monitoring and early warning method and device, a storage medium and an electronic device, thereby at least to some extent solving the problem of low accuracy and reliability of slope monitoring and early warning in the related art due to the difficulty in fully mining the potential correlation between multi-modal data.

[0007] According to a first aspect of the embodiments of the present disclosure, a slope monitoring and early warning method is provided, comprising:

[0008] Obtaining multi-modal data of different monitoring points in a target slope area;

[0009] Performing principal component analysis on the multi-modal data of each monitoring point to obtain corresponding slope feature scores;

[0010] Constructing a geographical space continuous model of the target slope area, determining reference feature scores of the spatial neighborhood of each monitoring point in the geographical space continuous model, and determining the local deviation degree of each monitoring point according to the slope feature scores and the reference feature scores;

[0011] The local deviation degrees are used for spatial correlation analysis to obtain a global deviation degree of the target slope region, and slope early warning information is generated according to the global deviation degree.

[0012] In an example embodiment of the present disclosure, the multi-modal data of each monitoring point is subjected to principal component analysis to obtain a corresponding slope feature score, including:

[0013] A standardized matrix is generated according to the multi-modal data of each monitoring point, and slope feature data corresponding to each modal data is extracted from the standardized matrix;

[0014] A pairing priority between any two modal data is determined according to a preset combination rule, and a target slope feature data pair is determined according to the pairing priority;

[0015] A covariance matrix is constructed according to a mean difference between the target slope feature data pair;

[0016] The covariance matrix is subjected to eigenvalue decomposition, and a principal component matrix is determined based on the eigenvalues obtained by the decomposition;

[0017] The standardized matrix and the principal component matrix are subjected to matrix operation to obtain a slope feature score of each monitoring point.

[0018] In an example embodiment of the present disclosure, the covariance matrix is subjected to eigenvalue decomposition, and a principal component matrix is determined based on the eigenvalues obtained by the decomposition, including:

[0019] The covariance matrix is subjected to eigenvalue decomposition to obtain a characteristic vector and a corresponding eigenvalue of each principal component direction;

[0020] A target principal component direction is determined according to the size of the eigenvalue, and the principal component matrix is generated according to the characteristic vector of the target principal component direction.

[0021] In an example embodiment of the present disclosure, the reference feature score of the spatial neighborhood of each monitoring point in the geographic spatial continuous model is determined, including:

[0022] The slope feature scores of each monitoring point in the geographic spatial continuous model are subjected to spatial interpolation to obtain a reference feature score of the spatial neighborhood of each monitoring point.

[0023] In an example embodiment of the present disclosure, the local deviation degrees are used for spatial correlation analysis to obtain a global deviation degree of the target slope region, including:

[0024] Based on geographical distribution information of each monitoring point in the target slope region, a graph structure is constructed with monitoring points as nodes and spatial adjacency relationship between monitoring points as edges, local deviation degree of each monitoring point is taken as a node feature vector, and initial graph data is generated according to the graph structure and the node feature vector;

[0025] The initial graph data is input into a trained graph neural network model, neighbor features of each node are aggregated by the graph neural network model, and an updated node feature vector is obtained;

[0026] Based on the updated node feature vector, spatial weighted deviation values of each node are calculated in combination with edge weight information, and a node deviation feature map of the target slope region is generated according to the spatial weighted deviation values;

[0027] The node deviation feature map is globally aggregated to obtain a global deviation degree of the target slope region.

[0028] In an exemplary embodiment of the present disclosure, the generation of slope warning information according to the global deviation degree comprises:

[0029] A plurality of warning threshold values are set according to historical monitoring data of the target slope region;

[0030] The global deviation degree is judged by levels according to the plurality of warning threshold values, and a current warning level of the target slope region is determined;

[0031] Based on the current warning level, corresponding slope warning information is generated in combination with a risk trigger condition;

[0032] The slope warning information at least includes a warning level, a risk region, a risk prediction time and a recommended countermeasure; and the risk trigger condition at least includes a rainfall trigger analysis result, an earthquake trigger analysis result and a static instability analysis result.

[0033] In an exemplary embodiment of the present disclosure, the acquisition of multi-modal data of different monitoring points in the target slope region comprises:

[0034] Multi-modal sensors are arranged according to topographic features and risk distribution of the target slope region; wherein the multi-modal sensors include micro-vibration sensors, acoustic wave sensors, displacement sensors, rainfall sensors, inclination sensors and thermal imaging sensors;

[0035] Multi-modal data of each monitoring point is collected by the multi-modal sensors; wherein the multi-modal data includes micro-vibration signals, acoustic wave signals, ground displacement data, rainfall data, inclination data of slope nodes and ground temperature data in the target slope region.

[0036] According to a second aspect of the embodiments of the present disclosure, a slope monitoring and early warning device is provided, comprising:

[0037] A multi-modal data acquisition module is configured to acquire multi-modal data of different monitoring points in a target slope region.

[0038] A feature score determination module is configured to perform principal component analysis on the multi-modal data of each monitoring point to obtain a corresponding slope feature score.

[0039] A deviation degree determination module is configured to construct a geographical space continuous model of the target slope region, determine reference feature scores of spatial neighborhoods of each monitoring point in the geographical space continuous model, and determine local deviation degrees of each monitoring point according to the slope feature scores and the reference feature scores.

[0040] An early warning information generation module is configured to perform spatial correlation analysis using the local deviation degrees to obtain a global deviation degree of the target slope region, and generate slope early warning information according to the global deviation degree.

[0041] According to a third aspect of the present disclosure, a computer-readable storage medium is provided, and the computer-readable storage medium stores a computer program. When the computer program is executed by a processor, any step of the slope monitoring and early warning method in the first aspect is implemented.

[0042] According to a fourth aspect of the present disclosure, an electronic device is provided, comprising:

[0043] A processor and a memory are provided. The memory stores computer readable instructions. When the computer readable instructions are executed by the processor, any step of the slope monitoring and early warning method in the first aspect is implemented.

[0044] The technical solutions provided by the embodiments of the present disclosure can have the following beneficial effects:

[0045] The slope monitoring and early warning method provided in the example embodiments of the present disclosure comprises: acquiring multi-modal data of different monitoring points in a target slope region; performing principal component analysis on the multi-modal data of each monitoring point to obtain corresponding slope feature scores; constructing a geographical space continuous model of the target slope region, determining reference feature scores of the spatial neighborhood of each monitoring point in the geographical space continuous model, and determining the local deviation degree of each monitoring point according to the slope feature scores and the reference feature scores; performing spatial correlation analysis using the local deviation degrees to obtain the global deviation degree of the target slope region, and generating slope early warning information according to the global deviation degree. On the one hand, by performing principal component analysis on the multi-modal data, the correlation information between features can be extracted from multiple dimensions, thereby realizing deep integration of complex multi-modal data, making full use of the internal coupling relationship of different modal data, and avoiding the potential information loss problem caused by single data analysis in related technologies. On the other hand, by determining the spatial neighborhood and reference feature scores of the monitoring points through the geographical space continuous model, the relative state of the monitoring points in the local environment can be dynamically reflected, the local abnormality degree can be accurately quantified, and the potential risk area can be quickly located. Further, by integrating the local deviation degrees through spatial correlation analysis, the overall deviation representation of the target slope region is formed, thereby providing quantitative evaluation of the global risk. On the other hand, the early warning information is dynamically generated by using the global deviation degree, so that the slope monitoring system can quickly respond to the change of the slope environment and has the real-time identification ability of the sudden risk, thereby providing a more scientific and effective solution for landslide early warning.

[0046] It should be understood that the foregoing general description and the following detailed description are only exemplary and explanatory, and cannot limit the present disclosure. BRIEF DESCRIPTION OF DRAWINGS

[0047] The drawings incorporated into the specification and forming a part thereof illustrate embodiments consistent with the present disclosure and, together with the specification, serve to explain the principles of the present disclosure. Obviously, the drawings in the following description are only some embodiments of the present disclosure, and other drawings can be obtained by those skilled in the art without creative labor based on these drawings.

[0048] Figure 1 A system architecture diagram of a slope monitoring and early warning method to which the embodiments of the present disclosure can be applied is shown.

[0049] Figure 2 A flowchart of a slope monitoring and early warning method in the embodiments of the present disclosure is shown.

[0050] Figure 3 A flowchart of determining slope feature scores of monitoring points in the embodiments of the present disclosure is shown.

[0051] Figure 4 A flow diagram of determining a global deviation degree of a target slope region in an embodiment of the present disclosure is shown.

[0052] Figure 5 A schematic diagram of a slope monitoring and early warning device in an embodiment of the present disclosure is shown.

[0053] Figure 6 A structural schematic diagram of an electronic device suitable for implementing an embodiment of the present disclosure is shown.

[0054] In the drawings, identical or corresponding reference signs indicate identical or corresponding parts. DETAILED DESCRIPTION

[0055] The terminology used in this specification is for the purpose of describing particular embodiments only and is not intended to limit the specification. As used in this specification, the singular forms "a," "an" and "the" are intended to include the plural forms as well, unless the context clearly indicates otherwise. It will be further understood that the terms "comprises" and / or "comprising," when used in this specification, specify the presence of stated features, integers, steps, operations, elements, and / or components, but do not preclude the presence or addition of one or more other features, integers, steps, operations, elements, components, and / or groups thereof.

[0056] It should be understood that although the terms first, second, third, etc. can be used herein to describe various information, these terms are not intended to denote a particular order or hierarchy. These terms are used only to distinguish one from another. For example, a first information can be termed a second information, and similarly, a second information can be termed a first information, without departing from the scope of the present specification. As used herein, the term "if' can be interpreted to mean "when" or "responsive to the determination" or "in response to the determination" or "upon the determination" or "in response to the establishment" depending on the context.

[0057] Figure 1 A system architecture schematic diagram of a slope monitoring and early warning method to which an embodiment of the present disclosure can be applied is shown.

[0058] As Figure 1 shown, the system architecture 100 can include one or more of a terminal device such as a smart phone 101, a portable computer 102, a desktop computer 103, a network 104, and a server 105. The network 104 is a medium for providing a communication link between the terminal device and the server 105. The network 104 can include various connection types, such as wired, wireless communication links, or fiber optic cables, etc.

[0059] The terminal device can be any electronic device with data processing capabilities, having a display screen for displaying multimodal data, slope characteristic scores, reference characteristic scores, local deviations, global deviations, and slope warning information from different monitoring points within the target slope area. Such electronic devices include, but are not limited to, the aforementioned desktop computers, portable computers, smartphones, and tablet computers.

[0060] It should be understood that Figure 1 The number of terminal devices, networks and servers in the embodiment is merely illustrative. Any number of terminal devices, networks and servers may be provided as needed. For example, the server 105 may be a server cluster consisting of multiple servers.

[0061] The slope monitoring and early warning method provided in the embodiments of the present disclosure can be executed by a terminal device, and accordingly, the slope monitoring and early warning device can be installed in the terminal device. However, those skilled in the art will readily appreciate that the slope monitoring and early warning method provided in the embodiments of the present disclosure can also be executed by server 105, and accordingly, the slope monitoring and early warning device can also be installed in server 105, and this exemplary embodiment does not specifically limit this.

[0062] The present disclosure first provides a slope monitoring and early warning method. The method is described below using server execution as an example.

[0063] Figure 2 The flowchart of a slope monitoring and early warning method in an embodiment of the present disclosure is schematically shown. Figure 2 As shown, the method may include steps S210 to S240:

[0064] Step S210, acquiring multimodal data of different monitoring points in the target slope area;

[0065] Step S220, performing principal component analysis on the multimodal data of each monitoring point to obtain a corresponding slope characteristic score;

[0066] Step S230: constructing a geospatial continuity model of the target slope area, determining a reference characteristic score of a spatial neighborhood of each monitoring point in the geospatial continuity model, and determining a local deviation degree of each monitoring point based on the slope characteristic score and the reference characteristic score;

[0067] Step S240 , performing spatial correlation analysis using the local deviation degrees to obtain a global deviation degree of the target slope area, and generating slope warning information according to the global deviation degree.

[0068] The slope monitoring and early warning method in this exemplary embodiment will be described in detail below.

[0069] In step S210, multimodal data of different monitoring points in the target slope area are obtained.

[0070] In the present embodiment, the target slope area refers to a slope geological area that needs to be monitored and studied. For example, the target slope area can be a slope area where cracks, settlement, or displacement occur, and landslides or rockfalls are likely to occur. It can also be a slope area where the strata are loose and there are weak soil layers or fault zones. It can also be a slope area with a large slope gradient and low stability. This disclosure does not limit this. Multimodal data refers to data obtained by deploying multiple types of sensors and acquisition equipment to conduct comprehensive monitoring of the target slope area.

[0071] For example, multimodal sensors can be deployed based on the terrain characteristics and risk distribution of the target slope area. These sensors include micro-vibration sensors, acoustic wave sensors, displacement sensors, rainfall sensors, tilt sensors, and thermal imaging sensors. For example, based on the terrain characteristics of the target slope area and known landslide or collapse risk points, sensors can be deployed preferentially in high-risk areas. The sensor density can be adjusted based on the importance and risk level of the area to ensure that sensors cover different areas of the slope.

[0072] Specifically, micro-vibration sensors can be deployed in key monitoring areas, such as around cracks, at the top and toe of the slope, to capture subtle vibration signals from the slope and identify crack expansion or landslide precursors. Acoustic wave sensors can be deployed in the middle of the slope and in areas prone to landslide activity to monitor acoustic signals generated by landslides or rockfall, thereby identifying landslide triggering events. Displacement sensors can be deployed at key points on the slope surface, such as crack edges, to monitor changes in horizontal and vertical displacement and determine deformation trends. Rainfall sensors can be deployed at the top of the slope and at representative hydrological locations within the area, recording local rainfall intensity and accumulated rainfall to analyze the potential for rainfall-induced landslides. Inclination sensors can be embedded within the slope or installed at locations prone to slippage. They measure the slope's inclination angle and changes, which can indicate landslide precursors. Thermal imaging sensors can be deployed on the slope surface, covering cracks and areas of thermal anomalies, to monitor surface temperature distribution and identify groundwater activity or abnormal heat sources.

[0073] Multimodal sensors are then used to collect multimodal data from each monitoring point. This multimodal data includes microvibration signals, acoustic signals, surface displacement data, rainfall data, slope node inclination data, and surface temperature data within the target slope area. Microvibration signals, including acceleration, spectral characteristics, and vibration energy, are collected by microvibration sensors. By capturing rapid vibration changes, they can identify local slope instability phenomena such as crack propagation and slip initiation. Acoustic signals, including the frequency and intensity of landslides or rockfalls, can be collected by acoustic sensors to identify landslide triggering events and provide dynamic information about rockfall or slope sliding processes. Surface displacement data, including horizontal and vertical displacement changes at surface points, can be collected by surface displacement sensors for long-term monitoring of slope deformation. Precipitation data, including real-time rainfall intensity and cumulative rainfall, is collected by rain gauge sensors and used to analyze the impact of hydrological factors on slope instability. Slope node inclination data, including the inclination value and rate of change, provides information about the slope's inclination trend and assists in landslide risk assessment. A slope node can be a specific location on the slope or the location of a rockfall. This data is collected using an inclination sensor. Surface temperature data, including the slope surface temperature distribution and temperature changes, is collected using a thermal imaging sensor.

[0074] It is important to note that the data collected by each sensor is recorded with a unified timestamp to ensure time synchronization of data from different modalities. Furthermore, data from different modalities are labeled according to the spatial coordinates of the sensor locations to form a unified monitoring network. When preprocessing the collected multimodal data, filtering algorithms can be used to denoise the microvibration and acoustic signals. Standardization methods are used to unify the dimensions of data from different modalities. The preprocessed data is then structured and stored to form standardized data tables divided by monitoring points.

[0075] Through the joint deployment and coordination of multimodal sensors, comprehensive physical, meteorological and thermal information of the target slope area can be obtained, providing reliable data support for subsequent slope stability analysis and slope early warning.

[0076] In step S220, principal component analysis is performed on the multimodal data of each monitoring point to obtain a corresponding slope characteristic score.

[0077] In the embodiments of this specification, principal component analysis is performed on the multimodal data of the monitoring points to extract key features reflecting slope stability from the multimodal data. Optionally, the covariance matrix can be calculated by evaluating the correlation between the modal data, and the principal components that can be used to express the comprehensive characteristics of the multimodal data can be determined based on the covariance matrix. The slope characteristic score obtained by the final analysis is the projection value of the principal component in the low-dimensional space, which reflects the comprehensive stability of the monitoring point.

[0078] Reference Figure 3 As shown in FIG. 6, the process of performing principal component analysis on the multi-modal data of each monitoring point to obtain the corresponding slope feature score can include steps S310 to S350:

[0079] In step S310, a standardized matrix is generated according to the multi-modal data of each monitoring point, and the slope feature data corresponding to each modal data is extracted from the standardized matrix;

[0080] The standardized data table obtained in step S210 is used to generate a standardized matrix, denoted as Z = {Z ij}, where i is the monitoring point number, j is the type of modal data, and each row in the matrix represents the standardized data of a certain modal. Based on this, each modal data can be extracted from the standardized matrix Z of the ith monitoring point as a slope feature sub-matrix Z j , where i is the monitoring point number, j is the type of modal data, and each row in the matrix represents the standardized data of a certain modal. Based on this, each modal data can be extracted from the standardized matrix Z of the ith monitoring point as a slope feature sub-matrix Z j , where i is the monitoring point number, j is the type of modal data, and each row in the matrix represents the standardized data of a certain modal. Based on this, each modal data can be extracted from the standardized matrix Z of the ith monitoring point as a slope feature sub-matrix Z

[0081] In step S320, the pairing priority between any two modal data is determined according to a preset combination rule, and the target slope feature data pair is determined according to the pairing priority;

[0082] In an example embodiment, the combination rule can be defined according to the triggering mechanism and correlation of landslides. Specifically, the abnormal signal of micro-vibration can be an early sign of instability of the ground structure, and the change of inclination can further reflect the displacement trend of the slope, and the combination of the two can help to capture the early features before the landslide occurs, so the micro-vibration data and the inclination data can be paired. Rainfall is one of the main causes of landslides, and the change of ground displacement is often directly related to the intensity and cumulative amount of rainfall, so by pairing the ground displacement data and the rainfall data, the real-time impact of rainfall on slope stability can be analyzed. Changes in ground temperature can affect the evaporation of water in the slope and the strength of the soil, and thus affect the change of inclination, so by pairing the inclination data and the ground temperature data, the indirect impact of temperature changes on slope stability can be revealed. The crack expansion before the landslide can be accompanied by abnormal signals of low-frequency sound waves and micro-vibration, so by pairing the micro-vibration data and the sound wave data, hidden changes in the slope cracks can be detected. Rapid changes in inclination are often a precursor to ground displacement, so by pairing the ground displacement data and the inclination data, the core features in the overall instability process of the slope can be captured.

[0083] Further, the preset combination rule can determine the pairing priority between any two modal data, i.e., the pairing priority between micro-vibration data and inclination data, between ground surface displacement data and rainfall data, between inclination data and ground surface temperature data, between micro-vibration data and acoustic wave data, and between ground surface displacement data and inclination data is higher, and the modal data pair with higher pairing priority is taken as the target slope feature data pair.

[0084] The preset combination rule closely combines the physical connection between modal data and risk logic, can improve the correlation of data analysis, and provides more accurate and comprehensive support for landslide early warning.

[0085] Step S330, constructing a covariance matrix according to the mean difference between the target slope feature data pairs;

[0086] For the target slope feature data pair set D={(Z j ,Z k )} of the i th monitoring point, Z j represents the slope feature data corresponding to the j th modal data, and Z k represents the slope feature data corresponding to the k th modal data. For each data pair (Z j ,Z k ), the mean difference between them is calculated.

[0087] (1)

[0088] wherein, , are the mean values of the j th modal data and the k th modal data, respectively.

[0089] Then, the covariance between the j th modal data and the k th modal data is calculated according to formula (2).

[0090] (2)

[0091] wherein, is the mean difference between the j th modal data and the k th modal data, used to adjust the weight of the covariance, , are the standard deviations of the j th modal data and the k th modal data, respectively, representing the overall volatility of each modal data.

[0092] According to formula (1) and (2), the covariance of all target slope feature data pairs can be calculated, and a covariance matrix is constructed according to all covariances. The covariance The greater, the stronger correlation between the changes of the two modal data, i.e., the data values change consistently. For slope monitoring, the data pair with a large covariance indicates that the two modal data have a strong coupling relationship in the development process of the landslide and can jointly describe a certain landslide feature or stage.

[0093] For example, the sudden increase point of the micro-vibration signal and the point where the inclination change rate increases simultaneously may indicate the accelerated development stage of the landslide. After the rainfall exceeds a certain threshold, the displacement rate significantly increases, which may reflect the soil sliding caused by rainfall. In the short term after the end of the high-temperature drought period, the inclination change increases, which may be related to soil swelling and strength reduction. The low-frequency signal of the acoustic wave is enhanced, and at the same time, the frequency range of the micro-vibration is expanded, which may reflect the process of surface crack expansion. The inclination changes sharply in a certain region, and at the same time, the displacement rate suddenly increases, which may reflect the critical state of the landslide.

[0094] In step S340, eigenvalue decomposition is performed on the covariance matrix, and a principal component matrix is determined based on the eigenvalues obtained by the decomposition;

[0095] Exemplarily, the eigenvalue decomposition is performed on the covariance matrix to obtain the eigenvector of each principal component direction and the corresponding eigenvalue. Then, the target principal component direction is determined according to the size of the eigenvalue, and the principal component matrix is generated according to the eigenvector of the target principal component direction. For example, the eigenvector direction corresponding to the first K largest eigenvalues is selected as the target principal component direction, and the principal component matrix is composed of the eigenvectors of the first K largest eigenvalues.

[0096] In step S350, matrix operation is performed on the standardized matrix and the principal component matrix to obtain the slope feature score of each monitoring point.

[0097] The matrix operation is performed on the standardized matrix and the principal component matrix to obtain the slope feature score matrix. In the slope feature score matrix, a higher slope feature score indicates that the region is in a stable state, and a lower slope feature score indicates that the region has a risk of landslide or rockfall.

[0098] The disclosed embodiments can effectively capture the dynamic changes of the slope region by integrating the features of multi-modal data, thereby providing an important scientific basis for landslide prediction and early warning.

[0099] In step S230, a geospatial continuous model of the target slope region is constructed, a reference feature score of the spatial neighborhood of each monitoring point in the geospatial continuous model is determined, and a local deviation degree of each monitoring point is determined according to the slope feature score and the reference feature score.

[0100] The geospatial continuous model is constructed based on the monitoring point data in the target slope region, in combination with the terrain, geology, and sensor distribution. For example, the target slope region can be divided into a continuous geospatial grid by using a finite element grid or a point cloud interpolation method. The local deviation degree of the monitoring point quantifies the deviation degree of the slope feature score of the monitoring point relative to the reference feature score of the spatial neighborhood, and is used to identify the abnormality of the monitoring point.

[0101] The spatial position and slope feature score of each monitoring point in the target slope region are known, and the spatial neighborhood of each monitoring point is determined by using the constructed geospatial continuous model. For example, all reference points within a specified radius from each monitoring point are selected as neighborhood points. For another example, a plurality of reference points closest to each monitoring point are selected as neighborhood points. For another example, the neighborhood is defined according to the trend of the slope surface terrain, such as along the main landslide direction or the fracture line, and the present disclosure does not limit this.

[0102] Exemplarily, the slope feature scores of the monitoring points in the geospatial continuous model are spatially interpolated to obtain the reference feature scores of the spatial neighborhood of the monitoring points. For example, by using the slope feature scores of the monitoring points, the reference feature scores of the spatial neighborhood of the monitoring points are spatially predicted by methods such as Kriging interpolation or Gaussian process regression, so as to obtain the reference feature scores of all neighborhood points in the corresponding spatial neighborhood.

[0103] After obtaining the reference feature scores of the spatial neighborhood of the monitoring points, the local deviation degree of each monitoring point can be determined according to the slope feature score and the reference feature score, and has:

[0104] (3)

[0105] wherein, is the local deviation degree of each monitoring point, is the slope feature score of the i-th monitoring point, is the reference feature score of the spatial neighborhood of the i-th monitoring point, which can be, for example, a weighted average of the reference feature scores of all neighborhood points in the spatial neighborhood of the i-th monitoring point.

[0106] By constructing the geospatial continuous model, calculating the reference feature scores of the spatial neighborhood of each monitoring point, and quantifying the local deviation degree based on the difference between the slope feature score and the reference feature score, the spatial variation of the slope feature can be effectively captured, and accurate quantitative analysis means is provided for landslide dynamic identification, risk assessment, and early warning.

[0107] In step S240, spatial correlation analysis is performed by using the local deviation degrees to obtain a global deviation degree of the target slope region, and slope early warning information is generated according to the global deviation degree.

[0108] In the examples of this specification, the purpose of spatial correlation analysis is to integrate local deviations into a global deviation, assessing the overall slope stability by analyzing the spatial relationships between monitoring points. The global deviation can be defined as an instability index for the entire target slope region, calculated based on multiple local deviations.

[0109] refer to Figure 4 As shown, the process of determining the global deviation degree of the target slope area may include steps S410 to S440:

[0110] Step S410: Based on the geographical distribution information of each monitoring point in the target slope area, a graph structure is constructed with the monitoring points as nodes and the spatial adjacency relationships between the monitoring points as edges. The local deviation degree of each monitoring point is used as a node feature vector, and initial graph data is generated based on the graph structure and the node feature vector.

[0111] Exemplarily, the specific location coordinates of each monitoring point are recorded, such as longitude and latitude or three-dimensional spatial coordinates. The adjacency relationship between the monitoring points is defined according to the specific location coordinates of each monitoring point, such as using Euclidean distance to define adjacent nodes. The monitoring points are regarded as nodes, and the adjacency relationship between the monitoring points is regarded as edges, and an undirected weighted graph G=(V, E, A) is generated, where V is a node set, E is an edge set, and A is edge weight information, such as an adjacency matrix composed of the inverse of the distance. The local deviation degree of each monitoring point is used as a node feature vector to form a node feature matrix. The initial graph data includes an adjacency matrix and a node feature matrix, which represent the spatial distribution of the current state of the target slope area.

[0112] Step S420: Input the initial graph data into the trained graph neural network model, aggregate the neighbor features of each node through the graph neural network model, and obtain an updated node feature vector;

[0113] The graph neural network model uses a message-passing mechanism to enable each node to exchange information with its neighbors, thereby updating the node feature vector. Specifically, when the trained graph neural network model is used to process the initial graph data, for each node, the corresponding node feature matrix and neighbor features are aggregated to obtain the updated node feature vector.

[0114] For example, the node feature vector is updated according to formula (4):

[0115] (4)

[0116] in, is the updated node feature vector of node i, is the activation function, is the linear transformation matrix of node features, is the node feature matrix of node i before updating, is a linear transformation matrix of neighbor features, is the neighbor feature of node i, is an edge weight, reflecting the relationship strength between node i and node j, is the number of nodes.

[0117] Through the superposition of multi-layer graph neural networks, a larger range of spatial correlation can be captured. Finally, the updated node feature vector is obtained, and at this time, the feature vector of each node has integrated the feature information of its neighbors.

[0118] In step S430, based on the updated node feature vector, the spatial weighted deviation value of each node is calculated in combination with the edge weight information, and the node deviation feature map of the target slope region is generated according to the spatial weighted deviation value of each node.

[0119] Exemplarily, for node i, the spatial weighted deviation value of the node can be calculated based on the updated node feature vector and the edge weight information , and has:

[0120] (5)

[0121] Further, the spatial weighted deviation value of each node is calculated according to formula (5), and the spatial weighted deviation value reflects the deviation degree of the node after integrating its own state and the influence of the neighborhood, which is a further aggregation of the node features.

[0122] In step S440, the node deviation feature map is globally aggregated to obtain the global deviation degree of the target slope region.

[0123] The global deviation degree of the target slope region can be obtained by globally aggregating the spatial weighted deviation values of all nodes in the node deviation feature map, and has:

[0124] (6)

[0125] wherein, is the global deviation degree of each monitoring point, is the weight of node i, is the spatial weighted deviation value of node i, is the number of nodes.

[0126] The global deviation degree can reflect the current stability condition of the target slope region. The higher the value of the global deviation degree, the higher the instability of the target slope region, and the greater the potential landslide risk or rockslide risk.

[0127] ​After determining the global deviation degree of the target slope area, a multi-level warning threshold can be set based on historical monitoring data for the target slope area. The global deviation degree is then graded using these thresholds to determine the current warning level for the target slope area. Based on the current warning level and combined with risk trigger conditions, corresponding slope warning information is generated. This slope warning information includes at least the warning level, risk area, risk prediction time, and recommended response measures. Risk trigger conditions include at least rainfall trigger analysis results, earthquake trigger analysis results, and static instability analysis results.

[0128] It is important to note that the risk prediction time in slope warning information can include the time the warning information was generated, the time the risk occurred (or the time of emergency response), etc., and this disclosure does not limit this. The risk area can be the entire slope area, a specific high-risk area, or a monitoring point number.

[0129] For example, a multi-level warning threshold is set based on the global deviation degree corresponding to the historical monitoring data collected in the historical landslide events. , the warning level is green warning, indicating that the global deviation degree is within the normal range and there is no landslide risk; if , the warning level is yellow, indicating that there is a slight landslide risk, and monitoring needs to be strengthened and relevant departments should be notified to pay attention; if , the warning level is orange, indicating a high possibility of landslides, and it is recommended to take preventive measures, such as evacuating personnel and vehicles, and blocking areas; if , the warning level is red, indicating that the risk of landslide is extremely high and requires emergency response, such as emergency evacuation, full blockade of the area, and initiation of landslide emergency response.

[0130] The current warning level of the target slope area is determined by comparing the current global deviation degree of the target slope area with the set multi-level warning threshold. For example, if the current global deviation degree of the target slope area is , the corresponding warning level is orange warning.

[0131] It should be noted that risks such as slope landslides are usually related to external triggering conditions, such as rainfall intensity, earthquake fluctuations, etc. Therefore, in the embodiments of this specification, on the basis of determining the warning level, a comprehensive analysis can be conducted in combination with the risk triggering conditions, such as predicting the risk time point, which helps to generate more accurate slope warning information.

[0132] For example, in combination with real-time rainfall and predicted rainfall trend, the cumulative curve is used to predict the time point when the rainfall reaches the threshold. The future rainfall trend can be obtained through the rainfall prediction data of the meteorological department. The real-time monitored rainfall is superimposed with the future rainfall prediction, the cumulative rainfall at a certain time in the future is calculated, and the time when the cumulative rainfall reaches the set threshold is the time range when the landslide risk may occur. Similarly, by monitoring the seismic wave or vibration frequency, the time range that may cause the landslide is identified, and the critical instability time can also be calculated in combination with the current slope inclination and stress state, which is not described in detail in the embodiments of the present specification.

[0133] In addition, the slope early warning information can be published through multiple channels, such as short message, APP (application) push, broadcast system, etc. For different levels of early warning, appropriate publication range can also be selected, for example, the coverage range is expanded for red early warning.

[0134] By performing the slope monitoring and early warning method provided by the present disclosure, on the one hand, the correlation information between features can be extracted from multiple dimensions by performing principal component analysis on multi-modal data, thereby realizing deep integration of complex multi-modal data, making full use of the internal coupling relationship of different modal data, and avoiding the potential information loss problem caused by single data analysis in the related art; on the other hand, the spatial neighborhood of the monitoring point and the reference feature score are determined by the geographic spatial continuous model, the relative state of the monitoring point in the local environment is dynamically reflected, the local anomaly degree is accurately quantified, and the potential risk area is quickly located; further, the local deviation degree is integrated by spatial correlation analysis to form the overall deviation representation of the target slope area, thereby providing quantitative evaluation of global risk; on the other hand, the early warning information is dynamically generated by using the global deviation degree, so that the slope monitoring system can quickly respond to the change of the slope environment and has real-time identification ability for sudden risks, thereby providing a more scientific and effective solution for landslide early warning.

[0135] In the example embodiment, a slope monitoring and early warning device is also provided. Referring to Figure 5 As shown in the figure, the slope monitoring and early warning device 500 can include a multi-modal data acquisition module 510, a feature score determination module 520, a deviation degree determination module 530, and an early warning information generation module 540, wherein:

[0136] The multi-modal data acquisition module 510 is configured to acquire multi-modal data of different monitoring points in a target slope area;

[0137] The feature score determination module 520 is configured to perform principal component analysis on the multi-modal data of each monitoring point to obtain corresponding slope feature scores;

[0138] The deviation degree determination module 530 is configured to construct a geospatial continuous model of the target slope region, determine reference feature scores of spatial neighborhoods of each of the monitoring points in the geospatial continuous model, and determine local deviation degrees of each of the monitoring points according to the slope feature scores and the reference feature scores.

[0139] The early warning information generation module 540 is configured to perform spatial correlation analysis on the local deviation degrees to obtain a global deviation degree of the target slope region, and generate slope early warning information according to the global deviation degree.

[0140] In an example embodiment of the present disclosure, the feature score determination module 520, when performing principal component analysis on the multi-modal data of each of the monitoring points to obtain corresponding slope feature scores, is specifically configured to:

[0141] generate a standardized matrix from the multi-modal data of each of the monitoring points, and extract slope feature data corresponding to each modality from the standardized matrix;

[0142] determine a pairing priority between any two modalities according to a preset combination rule, and determine a target slope feature data pair according to the pairing priority;

[0143] construct a covariance matrix according to mean value differences between the target slope feature data pair;

[0144] perform eigenvalue decomposition on the covariance matrix, and determine a principal component matrix based on eigenvalues obtained by the decomposition;

[0145] perform matrix operation on the standardized matrix and the principal component matrix to obtain the slope feature scores of each of the monitoring points.

[0146] In an example embodiment of the present disclosure, the feature score determination module 520, when performing eigenvalue decomposition on the covariance matrix and determining a principal component matrix based on eigenvalues obtained by the decomposition, is specifically configured to:

[0147] perform eigenvalue decomposition on the covariance matrix to obtain a feature vector and a corresponding eigenvalue of each principal component direction;

[0148] determine a target principal component direction according to the size of the eigenvalue, and generate the principal component matrix according to a feature vector of the target principal component direction.

[0149] In an example embodiment of the present disclosure, the deviation degree determination module 530, when determining reference feature scores of spatial neighborhoods of each of the monitoring points in the geospatial continuous model, is specifically configured to:

[0150] The slope feature scores of each monitoring point in the geospatial continuous model are spatially interpolated to obtain reference feature scores of a spatial neighborhood of each monitoring point.

[0151] In an example embodiment of the present disclosure, the early warning information generation module 540, when performing spatial correlation analysis using the local deviation degrees to obtain the global deviation degree of the target slope region, is specifically configured to:

[0152] Based on the geographical distribution information of each monitoring point in the target slope region, a graph structure is constructed with monitoring points as nodes and spatial adjacency relationships between monitoring points as edges, the local deviation degree of each monitoring point is taken as a node feature vector, and initial graph data is generated according to the graph structure and the node feature vector;

[0153] The initial graph data is input into a trained graph neural network model, and neighbor features of each node are aggregated by the graph neural network model to obtain an updated node feature vector;

[0154] Based on the updated node feature vector, spatial weighted deviation values of each node are calculated in combination with edge weight information, and a node deviation feature map of the target slope region is generated according to the spatial weighted deviation values;

[0155] The node deviation feature map is globally aggregated to obtain the global deviation degree of the target slope region.

[0156] In an example embodiment of the present disclosure, the early warning information generation module 540, when generating slope early warning information according to the global deviation degree, is specifically configured to:

[0157] Setting multi-level early warning thresholds according to historical monitoring data of the target slope region;

[0158] Classifying and judging the global deviation degree according to the multi-level early warning thresholds to determine a current early warning level of the target slope region;

[0159] Generating corresponding slope early warning information based on the current early warning level in combination with risk trigger conditions;

[0160] The slope early warning information at least includes early warning level, risk area, risk prediction time and recommended countermeasures; and the risk trigger conditions at least include rainfall trigger analysis results, earthquake trigger analysis results and static instability analysis results.

[0161] In an example embodiment of the present disclosure, the multi-modal data acquisition module 510, when acquiring multi-modal data of different monitoring points in the target slope region, is specifically configured to:

[0162] Deploy multimodal sensors based on the terrain characteristics and risk distribution of the target slope area; wherein the multimodal sensors include micro-vibration sensors, acoustic wave sensors, displacement sensors, rainfall sensors, tilt sensors, and thermal imaging sensors;

[0163] The multimodal data of each monitoring point is collected by the multimodal sensor; wherein the multimodal data includes micro-vibration signals, acoustic wave signals, surface displacement data, rainfall data, slope node inclination data and surface temperature data in the target slope area.

[0164] The specific details of each module of the above-mentioned slope monitoring and early warning device have been described in detail in the corresponding slope monitoring and early warning method, so they will not be repeated here.

[0165] The exemplary embodiments of the present disclosure further provide a computer-readable storage medium having stored thereon a program product capable of implementing the methods described above in this specification. In some possible implementations, various aspects of the present disclosure may also be implemented in the form of a program product comprising program code that, when executed on an electronic device, causes the electronic device to perform the steps described in the "Exemplary Methods" section above according to various exemplary embodiments of the present disclosure.

[0166] The program product may be a portable compact disc read-only memory (CD-ROM) and include program code, and may be run on an electronic device, such as a personal computer. However, the program product of the present disclosure is not limited thereto. In this document, a readable storage medium may be any tangible medium that contains or stores a program that can be used by or in conjunction with an instruction execution system, apparatus, or device.

[0167] The program product may utilize any combination of one or more readable media. The readable medium may be a readable signal medium or a readable storage medium. The readable storage medium may be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, device, or component, or any combination thereof. More specific examples (a non-exhaustive list) of readable storage media include: an electrical connection having one or more wires, a portable disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination thereof.

[0168] A computer readable signal medium can include a propagated data signal with computer readable program code embodied therein, for example, in baseband or as part of a carrier wave. Such a propagated signal can take any of a variety of forms, including, but not limited to, electro-magnetic, optical, or any suitable combination thereof. A computer readable signal medium can be any computer readable medium that can be involved in

[0169] The code can be transmitted in any form in any medium, including, but not limited to, radio frequency (RF), wireless, wire line, optical, or any suitable combination of the foregoing.

[0170] The program code can be implemented in any of a variety of programming languages, including object-oriented programming languages such as Java, C++, or the like, and conventional procedural programming languages, such as the "C" programming language or similar programming languages. The program code can execute entirely on the user's computing device, partly on the user's computing device, as a stand-alone software package, partly on the user's computing device and partly on a remote computing device or entirely on the remote computing device or server. In the latter scenario, the remote computing device can be connected to the user's computing device through any type of network, including a local area network (LAN) or a wide area network (WAN), or the connection can be made to an external computing device, such as through the Internet using an Internet Service Provider.

[0171] In addition, an example embodiment of the present disclosure further provides an electronic device capable of implementing the above-mentioned slope monitoring and early warning method.

[0172] The electronic device 600 according to this embodiment of the present disclosure will be described below with reference to Figure 6 Figure 6 The electronic device 600 shown is merely an example, and should not bring any limitation to the function and use range of the embodiments of the present disclosure.

[0173] As shown in Figure 6 The electronic device 600 is in the form of a general computing device. The components of the electronic device 600 can include, but are not limited to, the at least one processing unit 610, the at least one storage unit 620, a bus 630 connecting different system components, including the storage unit 620 and the processing unit 610, and a display unit 640.

[0174] ​The storage unit 620 stores program code that can be executed by the processing unit 610 to cause the processing unit 610 to perform the steps described in the above "Example Methods" section of this specification in accordance with various example implementations of the present disclosure. For example, the processing unit 610 can perform the method steps of the example implementations of the present disclosure.

[0175] The storage unit 620 can include a readable medium that can be in the form of volatile storage such as random access memory (RAM) 621 and / or cache memory 622, and also can include non-volatile storage such as read only memory (ROM) 623.

[0176] The storage unit 620 also can include a program / utility 624 having a set of programs / modules 625, including but not limited to, an operating system, one or more application programs, other program modules, and program data, each of which can include implementation of a network environment, alone or in combination.

[0177] The bus 630 can represent one or more of several types of bus structures, including a storage bus or bus controller, a peripheral bus, a graphics acceleration port, a processing unit, or a local bus using any of a variety of bus architectures.

[0178] The electronic device 600 also can communicate with one or more external devices 700 such as a keyboard or pointing device, a Bluetooth device, etc.; other devices that enable a user to interact with the electronic device 600; and / or one or more devices that enable the electronic device 600 to communicate with one or more other computing devices. Such communication can occur via an input / output (I / O) interface 650. Still yet, the electronic device 600 can communicate with one or more networks such as a local area network (LAN), a general area network (WAN), and / or the public network, such as the Internet, via a network adapter 660. As depicted, the network adapter 660 communicates with the other components of the electronic device 600 via the bus 630. It should be appreciated that although the network adapter 660 is depicted as a single component, the network adapter 660 can comprise two or more components that work together to facilitate communications between the electronic device 600 and one or more other computing devices. It should be appreciated that although not shown, other hardware and / or software components that can be used in conjunction with the electronic device 600 include but are not limited to, microcode, device drivers, redundant processing units, external disk drive arrays, RAID systems, tape drives, and data archival storage systems, etc.

[0179] Through the above description of the embodiments, those skilled in the art can easily understand that the example embodiments described herein can be implemented by software, or by software in combination with necessary hardware. Therefore, the technical solutions according to the embodiments of the disclosure can be embodied in the form of a software product. The software product can be stored in a non-volatile storage medium (which can be a CD-ROM, a USB, a mobile hard disk, or the like) or a network, and includes a number of instructions to make a computing device (which can be a personal computer, a server, a terminal device, or a network device) execute the methods according to the embodiments of the disclosure.

[0180] In addition, the above-described drawings are only schematic illustrations of the processes included in the method according to the exemplary embodiments of the disclosure, and are not intended for limiting purposes. It is easy to understand that the processes shown in the above-described drawings do not indicate or limit the time sequence of the processes. In addition, it is also easy to understand that the processes can be executed synchronously or asynchronously, for example, in a plurality of modules.

[0181] Through the above description of the embodiments, those skilled in the art can easily understand that the example embodiments described herein can be implemented by software, or by software in combination with necessary hardware. Therefore, the technical solutions according to the embodiments of the disclosure can be embodied in the form of a software product. The software product can be stored in a non-volatile storage medium (which can be a CD-ROM, a USB, a mobile hard disk, or the like) or a network, and includes a number of instructions to make a computing device (which can be a personal computer, a server, a terminal device, or a network device) execute the methods according to the embodiments of the disclosure.

[0182] Other embodiments of the disclosure will be apparent to those skilled in the art from consideration of the specification and practice of the disclosure disclosed herein. This application is intended to cover any variations, uses, or adaptations of the disclosure and includes what is present in the prior art and those that are present in the following claims. The specification and examples are to be regarded as exemplary only, with the true scope and spirit of the disclosure being indicated by the following claims.

[0183] It should be understood that the present disclosure is not limited to the precise structures described above and illustrated in the drawings and that various modifications and changes can be made without departing from the scope thereof. The scope of the present disclosure is limited only by the appended claims.

Claims

1. A slope monitoring and early warning method, characterized in that: include: Acquire multimodal data from different monitoring points within the target slope area; generating a standardized matrix based on the multimodal data of each monitoring point, and extracting slope characteristic data corresponding to each modal data from the standardized matrix; Determining the pairing priority between any two modal data according to a preset combination rule, and determining the target slope characteristic data pair according to the pairing priority; constructing a covariance matrix based on the mean difference between the target slope characteristic data pairs; Performing eigenvalue decomposition on the covariance matrix, and determining a principal component matrix based on the eigenvalues ​​obtained by the decomposition; Performing matrix operations on the standardized matrix and the principal component matrix to obtain the slope characteristic score of each monitoring point; Constructing a geospatial continuous model of the target slope area; the geospatial continuous model is a model constructed based on monitoring point data within the target slope area, combined with topography, geology, and sensor distribution; Performing spatial interpolation on the slope characteristic scores of each monitoring point in the geographic spatial continuity model to obtain a reference characteristic score of a spatial neighborhood of each monitoring point, and determining a local deviation degree of each monitoring point based on the slope characteristic score and the reference characteristic score; wherein the local deviation degree is used to quantify the degree of deviation of the slope characteristic score of each monitoring point relative to the reference characteristic score of the spatial neighborhood; Based on the geographical distribution information of each monitoring point in the target slope area, a graph structure is constructed with the monitoring points as nodes and the spatial adjacency relationship between the monitoring points as edges, the local deviation degree of each monitoring point is used as a node feature vector, and initial graph data is generated according to the graph structure and the node feature vector; Inputting the initial graph data into the trained graph neural network model, aggregating the neighbor features of each node through the graph neural network model to obtain an updated node feature vector; Based on the updated node feature vectors, the spatial weighted deviation value of each node is calculated in combination with the edge weight information, and a node deviation feature map of the target slope area is generated according to each of the spatial weighted deviation values; The node deviation feature graph is globally aggregated to obtain a global deviation degree of the target slope area, and slope warning information is generated according to the global deviation degree.

2. The slope monitoring and early warning method according to claim 1 is characterized in that: The performing eigenvalue decomposition on the covariance matrix and determining the principal component matrix based on the eigenvalues ​​obtained by the decomposition includes: Performing eigenvalue decomposition on the covariance matrix to obtain the eigenvector and corresponding eigenvalue of each principal component direction; The target principal component direction is determined according to the magnitude of the eigenvalue, and the principal component matrix is ​​generated according to the eigenvector of the target principal component direction.

3. The slope monitoring and early warning method according to claim 1 is characterized in that: Generating slope warning information according to the global deviation degree includes: Setting multi-level warning thresholds based on historical monitoring data of the target slope area; Performing graded judgment on the global deviation degree according to the multi-level warning threshold value to determine the current warning level of the target slope area; Based on the current warning level and combined with risk triggering conditions, corresponding slope warning information is generated; Among them, the slope warning information at least includes the warning level, risk area, risk prediction time and recommended response measures; the risk triggering conditions at least include rainfall triggering analysis results, earthquake triggering analysis results and static instability analysis results.

4. The slope monitoring and early warning method according to claim 1 is characterized in that: The obtaining of multimodal data from different monitoring points within the target slope area includes: Deploy multimodal sensors based on the terrain characteristics and risk distribution of the target slope area; wherein the multimodal sensors include micro-vibration sensors, acoustic wave sensors, displacement sensors, rainfall sensors, tilt sensors, and thermal imaging sensors; The multimodal data of each monitoring point is collected by the multimodal sensor; wherein the multimodal data includes micro-vibration signals, acoustic wave signals, surface displacement data, rainfall data, slope node inclination data and surface temperature data in the target slope area.

5. A slope monitoring and early warning device, characterized in that: The slope monitoring and early warning method according to any one of claims 1 to 4 is applied, wherein the device comprises: Multimodal data acquisition module, used to obtain multimodal data from different monitoring points in the target slope area; a characteristic score determination module, configured to perform principal component analysis on the multimodal data of each monitoring point to obtain a corresponding slope characteristic score; a deviation degree determination module, configured to construct a geospatial continuity model of the target slope area, determine a reference characteristic score of a spatial neighborhood of each monitoring point in the geospatial continuity model, and determine a local deviation degree of each monitoring point based on the slope characteristic score and the reference characteristic score; The warning information generating module is used to perform spatial correlation analysis using the local deviation degrees to obtain the global deviation degree of the target slope area, and generate slope warning information according to the global deviation degree.

6. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the slope monitoring and early warning method according to any one of claims 1 to 4 is implemented.

7. An electronic device, characterized in that: include: processor; and A memory storing computer-readable instructions, wherein the computer-readable instructions, when executed by the processor, implement the slope monitoring and early warning method according to any one of claims 1 to 4.

Images