A landslide intelligent identification method based on easy-disaster geological structure

Abstract

The present application relates to the technical field of geological disaster intelligent identification, and particularly relates to a landslide intelligent identification method based on easy-disaster geological structure, comprising: acquiring data, determining an upper constraint area, a lower exposed area and a dip candidate section; constructing a soft interlayer directional correlation graph; determining a through state according to a connection state of the candidate soft interlayer structure chain, a stable rock layer blocking state and a connection state with the lower exposed area, and determining a candidate sliding path according to the through state; determining a critical rainfall condition of the candidate sliding path by adjusting simulated rainfall conditions; and generating a landslide identification result. In the present application, the technical scheme of determining the dip candidate section by combining the occurrence characteristics of the slope, the occurrence characteristics of the soft interlayer and the spatial constraint relationship between the soft interlayer section and the lower exposed area solves the problem that it is difficult to accurately distinguish the soft interlayer section related to the bedding sliding from the complex slope geological structure.

Description

Technical Field

[0001] This invention relates to the field of intelligent geological disaster identification technology, and in particular to an intelligent landslide identification method based on disaster-prone geological structures. Background Technology

[0002] In rock slopes controlled by weak interlayers, the weak interlayers are typically distributed along a certain dip and angle. When the dip of the weak interlayer is close to the slope aspect and extends downslope towards the toe or free area, it may form a dip-side geological structure that controls slope sliding. Existing landslide identification methods usually identify weak interlayers and potential sliding zones within the slope based on topographic data, geological survey data, borehole data, outcrop data, or geophysical exploration results.

[0003] However, weak interlayers in complex slopes are often characterized by multiple spatially separated weak interlayer segments with varying attitudes, influenced by factors such as overburden obstruction, outcrop interruption, limited detection range, and geological structural changes. Existing technologies for identifying these weak interlayer segments typically focus on the dip, angle, or spatial location of individual segments, failing to comprehensively assess the slope's attitude characteristics, the interlayer's attitude characteristics, and the spatial constraints between the weak interlayer segments and the slope toe, free face, or exposed interlayer area. This makes it difficult to determine whether each weak interlayer segment matches the overall downslope direction and whether it possesses the spatial conditions to extend down the slope along the downslope direction. Summary of the Invention

[0004] To overcome the above shortcomings, this invention provides a landslide intelligent identification method based on disaster-prone geological structures, aiming to improve the problem of accurately distinguishing weak interlayer segments related to bedding sliding from complex slope geological structures.

[0005] In a first aspect, the present invention provides the following technical solution: a landslide intelligent identification method based on disaster-prone geological structures, comprising the following steps: S1. Obtain slope spatial data, weak interlayer structure data and rainfall data of the target slope, and determine the upper restraint zone, lower exposed zone and dip-side candidate segments based on the slope spatial data and weak interlayer structure data. S2. Generate candidate segment pairs based on the spatial structural relationship between the dip candidate segments, extract the segment association features of the candidate segment pairs, input the segment association features into the interlayer connection discrimination model to obtain the connection discrimination value, and establish a directional association relationship between dip candidate segments that meet the connection conditions based on the connection discrimination value and the blocking state of stable rock strata, and construct a weak interlayer directional association map. S3. Search for candidate weak interlayer structural chains extending from the upper constrained area to the lower exposed area in the weak interlayer orientation correlation diagram. Determine the penetration state based on the connection state of the candidate weak interlayer structural chains, the blocking state of stable rock strata, and the connection state with the lower exposed area. Determine candidate slip paths based on the penetration state. S4. Determine the infiltration state under rainfall based on the rainfall data and the spatial structure parameters of the candidate slip path, update the hydraulic state parameters and shear strength parameters of the candidate slip path based on the infiltration state, determine the stability index of the candidate slip path, and determine the critical rainfall conditions of the candidate slip path by adjusting the simulated rainfall conditions. S5. Determine the target landslide path based on the connectivity status, stability index and critical rainfall conditions of each candidate landslide path, and generate landslide identification results based on the target landslide path.

[0006] By adopting the above technical solution, firstly, candidate dip-side segments are determined based on the slope spatial data and weak interlayer structure data of the target slope, and the spatial extension direction of the disaster-prone geological structure is limited by the upper constraint zone and the lower exposed zone; further, the interlayer connection discrimination model is used to determine the connection of the spatially separated dip-side candidate segments, while the stable rock layer blocking state is introduced to constrain the connection results, and a weak interlayer directional correlation map from the upper part of the slope to the lower part of the slope is constructed; on this basis, candidate weak interlayer structure chains are searched and the connection state and stable rock layer blocking are considered. The state of the landslide is determined by the state of the landslide and the connection state of the lower part, and candidate landslide paths are obtained. Then, the hydraulic state parameters, shear strength parameters and stability indices of the candidate landslide paths are combined with the rainfall infiltration process to determine the critical rainfall conditions of each candidate landslide path. Therefore, it is possible to realize the correlation reconstruction of spatially discontinuous weak interlayers, the identification of continuous landslide structures, the analysis of the stability state of landslide paths under rainfall, and the determination of target landslide paths, and obtain landslide identification results that are interconnected with the potential landslide paths, landslide space range, weak points of the path and critical rainfall conditions of the target slope.

[0007] The present invention has the following beneficial effects: 1. In this invention, a technical solution is adopted to determine the dip-parallel candidate segments by combining the slope attitude characteristics, the weak interlayer attitude characteristics, and the spatial constraint relationship between the weak interlayer segments and the lower exposed area. This achieves the technical effect of screening disaster-prone geological structures that match the potential bedding-parallel sliding direction from the target slope. Compared with the existing technical solution that mainly identifies landslide-prone areas based on topographic slope, surface morphology, or a single geological attribute, this solution solves the problem that it is difficult to accurately distinguish the weak interlayer segments related to bedding-parallel sliding from complex slope geological structures.

[0008] 2. In this invention, a technical solution is adopted to determine the connection discrimination value between weak interlayer segments based on segmented association features and interlayer connection discrimination model, and to establish directional association relationship in combination with the blocking state of stable rock strata. This achieves the technical effect of reconstructing spatially separated weak interlayer segments and forming a directional association map of weak interlayers. Compared with the existing technical solution of connecting weak interlayer segments based only on spatial distance, similarity of attitude or human experience, this solves the problem that it is easy to incorrectly connect segments that are actually blocked by stable rock strata, or to omit weak interlayer segments with potential continuity.

[0009] 3. In this invention, a technical solution is adopted to search for candidate weak interlayer structural chains extending from the upper constrained area to the lower exposed area in the weak interlayer directional correlation diagram, and to determine the continuity state by comprehensively considering the connection state, the stable rock layer blocking state, and the lower connection state. This achieves the technical effect of identifying complete candidate slip paths from local weak interlayer segments. Compared with the existing technical solutions that evaluate individual weak interlayers, local structural planes, or isolated geological anomaly areas separately, this solution solves the problem that it is difficult to reflect the directional transmission relationship between multiple weak interlayer segments and whether they jointly constitute a continuous slip structure.

[0010] 4. In this invention, a technical solution is adopted to correlate the rainfall infiltration state with the hydraulic state parameters, shear strength parameters, and stability indices of candidate slip paths, and to determine the critical rainfall conditions and target slip paths by adjusting simulated rainfall conditions. This achieves the technical effect of analyzing the state changes of candidate slip paths under rainfall and forming landslide identification results that include target slip paths, slip space identification information, path weak location identification information, and critical rainfall conditions. Compared with the existing technology that separates geological structure identification from rainfall stability analysis, or judges landslide probability solely based on rainfall information, this invention solves the problem of difficulty in establishing the correspondence between disaster-prone geological structures, rainfall infiltration, strength parameter changes, and potential slip paths. Attached Figure Description

[0011] Figure 1 This is a flowchart of a landslide intelligent identification method based on disaster-prone geological structures proposed in this invention. Figure 2 This is a flowchart of the dip-side candidate segment determination process for a landslide intelligent identification method based on disaster-prone geological structures proposed in this invention. Figure 3 This is a flowchart of the construction process of the directional correlation map of weak interlayers in a landslide intelligent identification method based on disaster-prone geological structures proposed in this invention. Figure 4 This is a flowchart of the candidate sliding path determination process for a landslide intelligent identification method based on disaster-prone geological structures proposed in this invention. Figure 5 This is a flowchart illustrating the landslide identification result generation process of a landslide intelligent identification method based on disaster-prone geological structures proposed in this invention. Detailed Implementation

[0012] The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0013] This invention provides a landslide intelligent identification method based on disaster-prone geological structures, such as... Figure 1 As shown, it includes the following steps: S1. Obtain slope spatial data, weak interlayer structure data and rainfall data of the target slope, and determine the upper restraint zone, lower exposed zone and dip-side candidate segments based on the slope spatial data and weak interlayer structure data. Furthermore, such as Figure 2 As shown, the upper restraint zone, lower exposure zone, and dip-following candidate segments are determined, including: Determine the slope attitude characteristics and slope spatial boundaries of the target slope based on slope spatial data; The upper constrained zone and the lower exposed zone are determined based on the spatial boundaries of the slope. Based on the slope attitude characteristics, the interlayer attitude characteristics of the weak interlayer segments, and the spatial constraint relationship between the weak interlayer segments and the lower exposed area, dip-side candidate segments are determined from the weak interlayer segments. Specifically, slope spatial data includes at least one of the following: digital elevation model, 3D laser point cloud data, UAV photogrammetry data, topographic survey data, orthophoto, and 3D slope model.

[0014] Data on weak interlayer structures include at least one of the following: spatial location, extent of extension, dip, dip angle, thickness, burial depth, lithology, hydrological parameters, and mechanical parameters.

[0015] Rainfall data includes at least one of the following: current rainfall, historical rainfall, rainfall intensity, cumulative rainfall, and rainfall duration.

[0016] Slope spatial data and weak interlayer structure data were converted to a unified spatial coordinate system and elevation datum, and rainfall data were converted to a unified time datum. Outlier removal, ground point classification, and slope reconstruction were performed on the slope spatial data; the weak interlayer structure data were segmented according to spatial continuity and changes in interlayer attitude, forming multiple weak interlayer segments.

[0017] Each weak interlayer segment is assigned a corresponding segment identifier and is associated with at least the spatial surface, center position, upper end position, lower end position, dip, dip angle, thickness, and burial depth. When the same weak interlayer manifests as multiple spatially separated planar structures due to overburden obstruction, blind spots in detection, outcrop interruption, or missing survey data, each planar structure is treated as an independent weak interlayer segment. The slope aspect and slope angle at each location of the target slope are calculated based on the slope spatial data, and the slope aspect and slope angle are used as slope attitude characteristics.

[0018] In one implementation, the upper constrained zone and the lower exposed zone are determined using the normalized elevation of each location on the target slope: ; in, Indicates spatial location Normalized elevation at the location; Represents the three-dimensional spatial position vector within the target slope; Indicates spatial location The actual elevation at the location; and These represent the minimum and maximum elevations within the target slope area, respectively.

[0019] Normalized elevation satisfies: Regions with normalized elevation greater than or equal to the upper elevation threshold are selected as candidate regions for the upper constraint region, and regions with normalized elevation less than or equal to the lower elevation threshold are selected as candidate regions for the lower exposed region.

[0020] In this scheme, the upper elevation threshold is preferably 0.8, and the lower elevation threshold is preferably 0.2. When the actual slope rear edge, slope top, slope toe, free boundary, or weak interlayer exposure boundary has been obtained, the actual boundary is preferred to determine the upper constrained area and the lower exposed area.

[0021] For the For each weak interlayer segment, calculate the minimum directional difference between the dip of the weak interlayer segment and the slope aspect of its location: ; in, Indicates the first The minimum directional difference between the dip of a weak interlayer segment and the corresponding slope aspect; Indicates the first A tendency towards weak interlayer segmentation; This indicates the slope direction of the section containing the weak interlayer. .

[0022] Based on the relationship between the dip angle of the weak interlayer segment and the corresponding slope angle, determine its dip angle matching state; extend the weak interlayer segment spatially along the downdip direction, and determine whether the extension direction is towards or reaches the lower exposed area.

[0023] When the directional difference of the weak interlayer segment meets the preset directional condition, the dip angle of the interlayer meets the preset dip angle condition, and the downdip direction of the interlayer is towards the lower exposed area, the weak interlayer segment is determined as a dip-side candidate segment.

[0024] In this scheme, the preset directional difference threshold is preferably 30°; the dip angle of the interlayer is preferably greater than 5°, and not greater than the sum of the corresponding slope angle and 10°. The specific threshold can be calibrated based on the attitude data of historical bedding landslides in the target area.

[0025] By using S1, candidate dip segments with spatial location, relative slope location, and interlayer attitude attributes are obtained, and these candidate dip segments are used as the basis for candidate segment generation and interlayer connection discrimination in S2.

[0026] S2. Generate candidate segment pairs based on the spatial structural relationship between dip candidate segments, extract the segment association features of the candidate segment pairs, input the segment association features into the interlayer connection discrimination model to obtain the connection discrimination value, and establish directional association relationships between dip candidate segments that meet the connection conditions based on the connection discrimination value and the blocking state of stable rock strata, and construct a directional association map of weak interlayers. Furthermore, such as Figure 3 As shown, candidate segment pairs are generated and segment association features are extracted, including: Based on the spatial distribution relationship of each dip candidate segment relative to the upper restraint area and the lower exposure area, the starting segment and the target segment located on the dip side of the starting segment are determined, and the starting segment and the target segment form a candidate segment pair; Based on the geometric association state, attitude association state, and medium association state between two dip candidate segments in the candidate segment pair, geometric association features, attitude association features, and medium association features are extracted respectively. Segmented correlation features are generated based on geometric correlation features, attitude correlation features, and medium correlation features; Specifically, any dip candidate segment obtained in S1 is taken as the starting segment, and a spatial search area is set on the downdip side of the starting segment. Other dip candidate segments located within the spatial search area are taken as target segments.

[0027] In this scheme, the spatial search area is preferably a three-dimensional fan-shaped area that unfolds along the downward tilting direction of the initial segment, and the angle between the search direction and the downward tilting direction of the initial segment is preferably no greater than 45°.

[0028] When the target segment is located on the downsloping side of the starting segment, and the average elevation of the target segment is lower than the average elevation of the starting segment, a candidate segment pair is formed by the starting segment and the target segment.

[0029] For candidate segment pairs Extract geometric correlation features. The shortest spatial distance between two weak interlayer segments is expressed as: ; in, Indicates the first The and the first The shortest spatial distance between each candidate segment with a forward inclination; and These represent two segmented sets of points on the three-dimensional surface. and These represent the three-dimensional position vectors of the two point sets, respectively. This represents the Euclidean norm.

[0030] Geometric correlation features may also include at least one of the following: elevation difference between adjacent ends, degree of overlap of directional projections, interval of dip projections, degree of thickness variation, and distance of fitted surface.

[0031] The attitude correlation features include at least one of the following: the dip difference between the two segments, the dip angle difference, the angle between the downdip directions, the angle between the normal vectors of the fitted surface, and the degree of consistency between the candidate connection direction and the downdip direction of the interlayer.

[0032] The medium association characteristics include at least one of the following: the type of spacer medium, the thickness of the spacer medium, the integrity of the rock mass, the degree of joint and fracture development, the shear strength of the spacer medium, and the tangential relationship between the spacer medium and the candidate connection direction.

[0033] For the Normalize the segmented association features of the items: ; in, Indicates candidate segment pairs The Original association features; This represents the normalized association features; and They represent the first The minimum and maximum values ​​of each feature in the training data; This indicates a positive number that prevents the denominator from being zero. In this plan, Preferred .

[0034] The geometric correlation feature vector, attitude correlation feature vector, and medium correlation feature vector are defined as follows: ; Concatenate the column vectors of the three types of features in a fixed order: ,in: ; in, Indicates candidate segment pairs The segmented associated feature column vector; , and These represent the dimensions of geometric correlation features, attitude correlation features, and media correlation features, respectively. This represents the total dimension of the segmented association features. All candidate segment pairs and historical training samples use the same feature types, arrangement order, and vector dimensions.

[0035] Furthermore, the mezzanine connection discrimination model is obtained through the following steps: Obtain historical weak interlayer segments confirmed by geological survey data; The same segment association feature extraction method as the candidate segment pairs is used to extract the training association features of historical weak interlayer segment pairs; Based on the historical weak interlayer segmentation, set corresponding connection markers to determine whether they belong to the same continuous weak interlayer; A classification model is trained based on training association features and connection labels to obtain a mezzanine connection discrimination model; Specifically, geological survey data includes at least one of the following: borehole data, trench data, continuous outcrop data, geophysical exploration data, construction excavation data, and landslide surface survey data of landslides that have occurred.

[0036] When historical weak interlayer segments are confirmed by geological survey to belong to the same continuous weak interlayer, the corresponding connection mark is set to 1; when historical segments are separated by stable rock layers or belong to different weak interlayers, the corresponding connection mark is set to 0.

[0037] for The training feature matrix is ​​represented by the historical training samples as follows: ; The corresponding linker vector is represented as: ; in, Represents the training feature matrix; Indicates the first Training associated feature vectors for each historical segment pair; Indicates transpose; Represents the training label vector; Indicates the first The connection markers for each historical segment pair.

[0038] Classification models can be logistic regression, support vector machine, random forest, gradient boosting tree, neural network, or ensemble classification models.

[0039] In this scheme, the gradient boosting tree model is preferred for the mezzanine connection discrimination model; when the number of training samples is small, the random forest model is preferred.

[0040] The trained mezzanine connectivity discriminant model is represented as follows: ,in: ; in, Indicates candidate segment pairs The connection discrimination value; This represents the completed training of the mezzanine connectivity discriminant model; Represents the set of model parameters; This represents the segmented correlation feature vector of the candidate segment pair.

[0041] Furthermore, a directional correlation is established based on the connectivity discrimination value and the blocking state of stable rock strata, including: The stable rock layer blocking state is determined based on the geometric, structural, and mechanical blocking characteristics of the interlayers between candidate segments; When the stable rock layer blocking state meets the preset blocking conditions, no directional correlation relationship is established between the corresponding dip-side candidate segments; When the stable rock layer blocking state does not meet the preset blocking conditions, and the connection discrimination value meets the preset connection conditions, a directional association relationship from the starting segment to the target segment is established between the corresponding dip-side candidate segments; Construct a weak interlayer directional correlation diagram based on the candidate segments and directional correlations; Specifically, geometric blocking features include at least one of the following: stable rock layer thickness, continuous length, spatial coverage, and tangential relationship with candidate connection directions.

[0042] Structural blocking characteristics include at least one of the following: the integrity of the interstitial rock mass, the degree of joint and fissure development, the rock mass quality grade, and the distribution of fracture zones.

[0043] Mechanical blocking characteristics include at least one of the following: interlayer cohesion, internal friction angle, uniaxial compressive strength, and equivalent shear strength.

[0044] The overall blocking value of stable rock strata is expressed as: ; And satisfy: ; in, Indicates candidate segment pairs The comprehensive blocking value of stable rock strata; , and These represent the geometric blocking component, the structural blocking component, and the mechanical blocking component, respectively. , and These represent the corresponding weights.

[0045] All blocking components and the overall blocking value of stable rock strata were normalized to between 0 and 1.

[0046] In this scheme, when historical calibration data is lacking, , and The preferred value is one-third.

[0047] when When the threshold is greater than or equal to the preset blocking threshold, no corresponding directional association is established; when... Less than the preset blocking threshold, and When the threshold is greater than or equal to the preset connection threshold, a directional association relationship is established from the starting segment to the target segment.

[0048] In this scheme, the preset connection threshold is preferably 0.6, and the preset blocking threshold is preferably 0.7.

[0049] The orientation correlation diagram of the weak interlayer is represented as follows: ; in, This represents a directional correlation diagram of a weak interlayer. This represents the set of nodes corresponding to the candidate segments with the inclination. This represents the set of directed edges corresponding to directional associations. Indicates that by node Pointing to node The directed edge.

[0050] S2 converts the dip-side candidate segments obtained in S1 into a weak interlayer directional correlation diagram with connection direction, connection discrimination value and stable rock layer blocking value, and uses the directional correlation diagram as input for S3.

[0051] S3. Search for candidate weak interlayer structural chains extending from the upper constrained area to the lower exposed area in the weak interlayer directional correlation diagram. Determine the penetration state based on the connection state of the candidate weak interlayer structural chains, the blocking state of stable rock strata, and the connection state with the lower exposed area. Determine the candidate slip path based on the penetration state. Furthermore, such as Figure 4 As shown, the search for candidate weak sandwich structure chains includes: The candidate segments with dip direction located in the upper constraint area are determined as the starting candidate segments, and the candidate segments with dip direction located in the lower exposure area or satisfying the preset spatial relationship with the lower exposure area are determined as the ending candidate segments. Starting from the initial candidate segment, a directed path search is performed along the directed association relationships in the weak interlayer directed association graph; The directed paths are filtered based on the path direction constraints and path blocking constraints of the searched directed paths relative to the target slope. The directed path from the starting candidate segment to the ending candidate segment is determined as the candidate weak sandwich structure chain; Specifically, the candidate segments for tilting that are located within the upper constraint zone, intersect with the upper constraint zone, or whose distance from the upper constraint zone meets the preset upper distance condition are determined as the starting candidate segments.

[0052] The dip candidate segments located within the lower outcrop area, intersecting with the lower outcrop area, or whose distance from the lower outcrop area meets the preset lower distance condition are determined as termination candidate segments.

[0053] In this scheme, the preset upper distance and the preset lower distance are preferably no greater than half of the average length of the candidate dip segment.

[0054] Starting from each initial candidate segment, a depth-first search method with pruning conditions is used to search for paths along the directed edges in the weak interlayer oriented association graph.

[0055] No. The candidate weak sandwich structure chain is represented as follows: ; Its set of directed edges is represented as: ; in, Indicates the first Candidate weak sandwich structure chain; This indicates the number of nodes contained in the chain structure; Represents the first in the structural chain One node; This represents the set of directed edges contained in the structural chain.

[0056] The elevation change between adjacent nodes is expressed as: ; in, Indicates the first The first in the structural chain The node to the first Elevation change of each node; and These represent the elevations of the segment center points corresponding to adjacent nodes.

[0057] The path direction constraints include the overall elevation decreasing towards the lower part of the slope, the overall path direction being consistent with the downward slope direction, the turning angle of adjacent directed sides not exceeding the preset path turning angle, and not repeatedly passing through the same nodes in the same path.

[0058] In this scheme, the preferred path turning angle is 45°.

[0059] Path blocking constraints include paths that do not pass through directed edges where the overall blocking value of stable rock strata reaches a preset blocking threshold, and paths that do not cross confirmed intact stable rock strata, anti-dip structural zones, or impassable areas.

[0060] Furthermore, the continuity status of the candidate weak sandwich structure chain is determined, including: Based on the connection discrimination value corresponding to each directional correlation that constitutes the candidate weak interlayer structure chain and the blocking state of the stable rock layer, the connection state of each directional correlation is determined. Based on the connection status of each directional relationship, determine the overall connection status and bottleneck connection status of the candidate weak sandwich structure chain; The connection status of the candidate weak sandwich structure chain is determined based on the lower connection characteristics between the end of the candidate weak sandwich structure chain and the lower exposed area. The breakthrough state is determined based on the overall connection state, bottleneck connection state, connection state and stable rock layer blocking state, and the candidate weak interlayer structure chain that meets the preset breakthrough condition is determined as the candidate slip path. Specifically, the reliability of a directional association connection is expressed as follows: ; in, Indicates by the first The first candidate segment with a downward inclination points to the first... The reliability of the connection of the directional association of each candidate segment with a forward inclination; Indicates the connection discrimination value; This represents the overall blocking value of stable rock strata.

[0061] For the The overall connectivity of a candidate weak sandwich structure chain is represented as follows: ; in, Indicates the first Average connection reliability of a structural chain; Indicates the first in the structural chain Reliability of a directed edge connection.

[0062] The bottleneck connection status is represented as follows: ,in, Indicates the first The lowest connection reliability in a chain structure.

[0063] The maximum stable rock layer blocking value is expressed as: ,in, Indicates the first The maximum stable rock layer blocking value in a structural chain; This indicates the first step in the structural chain. The stable rock layer blocking value corresponding to a directed edge.

[0064] The connection status value is determined based on the distance between the end of the structural chain and the lower exposed area, the consistency of the extension direction, the exposed length, and the degree of spatial overlap. ,in: , No. The connectivity of a candidate weak sandwich structure chain is expressed as: ; in, Indicates the first The degree of continuity of the structural chain; , , and These represent the influence parameters of the overall connection status, bottleneck connection status, connection status, and blocking status, respectively, and all are positive numbers.

[0065] In this scheme, when historical calibration data is missing, , , and The preferred value is set to 1.

[0066] when: season ,in, This represents the complete blocking threshold. In this scheme, the complete blocking threshold is preferably 0.9.

[0067] when When the threshold value is greater than or equal to a preset penetration threshold, the corresponding candidate weak interlayer structure chain is determined as a candidate slip path. The preset penetration threshold is preferably 0.6.

[0068] Candidate slip paths with information on penetration, connection reliability, stable rock layer blocking state, and spatial path are obtained through S3 and then passed to S4.

[0069] S4. Determine the infiltration state under rainfall based on rainfall data and spatial structure parameters of candidate slip paths. Update the hydraulic state parameters and shear strength parameters of candidate slip paths based on the infiltration state. Determine the stability index of candidate slip paths. Determine the critical rainfall conditions of candidate slip paths by adjusting simulated rainfall conditions. Furthermore, such as Figure 5 As shown, determining the infiltration state under rainfall includes: Determine the rainfall process characteristics of the target slope based on rainfall data; Based on the characteristics of the rainfall process and the hydrological infiltration characteristics of the overlying medium of the candidate slip path, the spatial influence range of rainfall infiltration is determined; The spatial influence range is compared with the spatial location of each dip candidate segment in the candidate slip path to determine the infiltration state corresponding to each dip candidate segment; The change in water state of each candidate segment along the dip is determined based on the infiltration state. Specifically, when the rainfall data represents the rainfall amount at each time step, the effective rainfall effect is expressed as: ; in, Indicates time The effective rainfall effect; Indicates the number of moments before the current moment. Rainfall within a time step; Indicates the number of historical time steps involved in the calculation; Indicates the time step; This parameter represents the time it takes for rainfall to decay in the preceding period.

[0070] When rainfall data is in terms of rainfall intensity, the effective rainfall effect is expressed as: ; in, This represents the average rainfall intensity within the corresponding time step.

[0071] The hydrological infiltration characteristics of the overlying medium of candidate slip paths include at least one of the following: saturated permeability coefficient, initial volumetric water content, saturated volumetric water content, wetting front suction, fracture development degree, and overlying medium thickness.

[0072] When rainfall intensity exceeds the infiltration capacity of the overlying medium and surface water accumulates, the Green-Ampt model is used to determine the cumulative infiltration depth: ; in, This represents the cumulative infiltration depth calculated by the Green-Ampt model; Represents the equivalent saturated permeability coefficient; Indicates the duration of rainfall; Indicates the suction head of the wetting front; It represents the difference between the saturated volumetric moisture content and the initial volumetric moisture content.

[0073] in , Indicates the saturated volumetric water content; This indicates the initial volumetric water content.

[0074] because When both sides of the formula appear simultaneously, the solution is obtained using an iterative method or a numerical root-finding method.

[0075] The actual cumulative infiltration depth is expressed as: ,in, Indicates the actual cumulative infiltration depth; Indicates the end time The cumulative rainfall depth.

[0076] The depth of a moist front is expressed as: ; When the overlying medium contains fissures or high-permeability channels, for the first The depth of the moist front at the corresponding position of each candidate segment with a downward slope is corrected: ,in, This indicates a correction for the depth of the wetted front; This represents the infiltration correction factor.

[0077] Definition of the first The thickness of each candidate dip segment is: ,in, and These represent the burial depths of the upper and lower bounds of the segment, respectively.

[0078] The water saturation of this segment is expressed as: ; in, Indicates time Water saturation; Indicates the initial water saturation; Indicates the response coefficient under water content conditions; This indicates a positive number that prevents the denominator from being zero. ; This means that the result will be limited to between 0 and 1.

[0079] The change in water content is expressed as: .

[0080] Furthermore, the hydraulic state parameters and shear strength parameters are updated, and the critical rainfall conditions are determined, including: Update the corresponding hydraulic state parameters based on the change in water state of each dip candidate segment; Based on the updated hydraulic state parameters and the preset strength weakening relationship, the corresponding shear strength parameters are updated; Stability indices are determined based on the updated hydraulic state parameters, shear strength parameters, spatial structure parameters of candidate slip paths, and stress state of the corresponding rock mass. Using the current rainfall conditions as the initial simulated rainfall conditions, the simulated rainfall conditions are adjusted according to the preset adjustment method, and the infiltration state, hydraulic state parameters, shear strength parameters and stability indexes are re-determined for the adjusted simulated rainfall conditions; The simulated rainfall condition that first causes the stability index to meet the preset critical stability condition is defined as the critical rainfall condition. Specifically, the hydraulic state parameters include at least one of water saturation and pore water pressure; the shear strength parameters include at least one of cohesion and internal friction angle.

[0081] Pore ​​water pressure is expressed as: ; in, Indicates the first Each candidate segment with a forward inclination at time... pore water pressure; Indicates the initial pore water pressure; Indicates the pore water pressure response coefficient; It indicates the density of water.

[0082] Effective cohesion is expressed as: ; in, Indicates effective cohesion; Indicates initial cohesion; Indicates the lower limit of cohesion; This represents the cohesion weakening coefficient.

[0083] The effective internal friction angle is expressed as: ; in, Indicates the effective internal friction angle; Indicates the initial internal friction angle; Indicates the lower limit of the internal friction angle; This represents the coefficient for weakening the internal friction angle.

[0084] In this scheme, the pore water pressure response coefficient, cohesion weakening coefficient, and internal friction angle weakening coefficient are preferably determined by fitting field monitoring data, direct shear test data, or triaxial test data.

[0085] The first The candidate slip paths are divided into Each computational unit has a stability coefficient expressed as: ; in, Indicates time No. Stability coefficients of candidate slip paths; Indicates the number of computational units; Indicates the first Effective cohesion of each computational unit; Indicates the area of ​​the corresponding slip surface; This represents the normal force acting on the slip surface; This represents the average pore water pressure; Indicates the effective internal friction angle; This represents the downward force along the potential sliding direction.

[0086] Using the current rainfall conditions as the initial simulated rainfall conditions, and adjusting them as follows: ,in , Indicates the first Level simulated rainfall conditions; Indicates current rainfall conditions; Indicates the step size for simulating rainfall; This indicates the maximum number of simulation levels.

[0087] In this scheme, the simulated rainfall adjustment step size is preferably 5 mm or 10 mm.

[0088] No. The critical rainfall conditions for the candidate slip paths are expressed as follows: ; in, Indicates the first Critical rainfall conditions for candidate slip paths; This represents the stability coefficient under the corresponding simulated rainfall conditions; This represents the preset critical stability coefficient.

[0089] In this scheme, the preset critical stability coefficient is preferably 1.0.

[0090] When the critical stability condition is not met under all simulated rainfall conditions, the critical rainfall condition is recorded as follows: The current stability coefficient and critical rainfall conditions for each candidate slip path are obtained through S4, and the results are passed to S5.

[0091] S5. Determine the target landslide path based on the connectivity status, stability index, and critical rainfall conditions of each candidate landslide path, and generate landslide identification results based on the target landslide path; Furthermore, the target landslide path is determined and landslide identification results are generated, including: Based on the connectivity, stability indicators, and critical rainfall conditions of each candidate slip path, the candidate slip paths are ranked. The candidate sliding paths that meet the preset selection criteria are determined as the target sliding paths; Based on the candidate dip segments contained in the target slip path and their spatial extension relationship, determine the slip space identification information; Based on the connection status of each orientation relationship in the target slip path and the change status of the shear strength parameters of each dip candidate segment, the weak location identification information of the path is determined; The landslide identification results are generated based on the target slip path, slip space identification information, path weak point identification information, and critical rainfall conditions. Specifically, the current stability coefficients of each candidate slip path are normalized: ; in, This represents the normalized stability coefficient; and These represent the minimum and maximum values ​​of the stability coefficients for all candidate slip paths, respectively. This indicates a positive number that prevents the denominator from being zero.

[0092] The critical rainfall conditions for each candidate slip path were normalized: ; in, This represents the normalized critical rainfall condition; and These represent the minimum and maximum values ​​among all finite critical rainfall conditions, respectively; This indicates a positive number that prevents the denominator from being zero.

[0093] For candidate slip paths that do not reach critical stability within the simulation upper limit, let ; The path sort value is represented as: and satisfy ; in, Indicates the first The sorting value of the candidate slip paths; Indicates the degree of penetration; , and These represent the weights corresponding to connectivity, stability coefficient, and critical rainfall conditions, respectively.

[0094] In this plan, , and The preferred values ​​are 0.4, 0.3, and 0.3, respectively.

[0095] The candidate slip path with the highest path ranking value is determined as the target slip path.

[0096] Based on the dip candidate segments included in the target slip path, the spatial surfaces of each segment are connected and surface fitted to form a potential slip surface; the spatial area enclosed by the potential slip surface, the target slope surface, the upper constraint area, and the lower exposed area is determined as the potential slip region.

[0097] Based on the connection reliability of each directional relationship in the target sliding path, the position with the lowest connection reliability or below the preset weak connection threshold is determined as the weak connection position.

[0098] No. The degree of cohesion weakening of each candidate dipping segment is expressed as follows: ; in, Indicates the degree of weakening of cohesion; This indicates a positive number that prevents the denominator from being zero.

[0099] The degree of weakening of the internal friction angle is expressed as: ; in, Indicates the degree of weakening of the internal friction angle; This indicates a positive number that prevents the denominator from being zero.

[0100] Candidate segments that meet the preset weak conditions, such as the degree of cohesion weakening, the degree of internal friction angle weakening, or the increase in pore water pressure, are identified as weak points in strength.

[0101] The final landslide identification results include at least one of the following: target landslide path, potential landslide surface, potential landslide area, weak connection location, weak strength location, and critical rainfall condition.

[0102] When no candidate segments for tilting are obtained, a scene mismatch identification result is generated; When no candidate weak interlayer structure chain from the upper constrained area to the lower exposed area is found, an identification result of incomplete structure chain is generated. When the connectivity of the candidate weak sandwich structure chains does not reach the preset connectivity threshold, a non-connected structure chain identification result is generated. When a candidate slip path fails to reach the critical stability condition within the preset simulated rainfall range, its critical rainfall condition is marked as being higher than the upper limit of the simulation calculation.

[0103] Finally, it should be noted that the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art can still modify the technical solutions described in the foregoing embodiments or make equivalent substitutions for some of the technical features. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A landslide intelligent identification method based on disaster-prone geological structures, characterized in that, Includes the following steps: S1. Obtain slope spatial data, weak interlayer structure data and rainfall data of the target slope, and determine the upper restraint zone, lower exposed zone and dip-side candidate segments based on the slope spatial data and weak interlayer structure data. S2. Generate candidate segment pairs based on the spatial structural relationship between the dip candidate segments, extract the segment association features of the candidate segment pairs, input the segment association features into the interlayer connection discrimination model to obtain the connection discrimination value, and establish a directional association relationship between dip candidate segments that meet the connection conditions based on the connection discrimination value and the blocking state of stable rock strata, and construct a weak interlayer directional association map. S3. Search for candidate weak interlayer structural chains extending from the upper constrained area to the lower exposed area in the weak interlayer orientation correlation diagram. Determine the penetration state based on the connection state of the candidate weak interlayer structural chains, the blocking state of stable rock strata, and the connection state with the lower exposed area. Determine candidate slip paths based on the penetration state. S4. Determine the infiltration state under rainfall based on the rainfall data and the spatial structure parameters of the candidate slip path, update the hydraulic state parameters and shear strength parameters of the candidate slip path based on the infiltration state, determine the stability index of the candidate slip path, and determine the critical rainfall conditions of the candidate slip path by adjusting the simulated rainfall conditions. S5. Determine the target landslide path based on the connectivity status, stability index and critical rainfall conditions of each candidate landslide path, and generate landslide identification results based on the target landslide path.

2. The landslide intelligent identification method based on disaster-prone geological structures according to claim 1, characterized in that, In S1, the determination of the upper constraint area, the lower exposure area, and the candidate dip segment includes: The slope attitude characteristics and slope spatial boundaries of the target slope are determined based on the slope spatial data. The upper constrained area and the lower exposed area are determined based on the spatial boundary of the slope. Based on the slope attitude characteristics, the interlayer attitude characteristics of the weak interlayer segments, and the spatial constraint relationship between the weak interlayer segments and the lower exposed area, the dip-side candidate segments are determined from the weak interlayer segments.

3. The landslide intelligent identification method based on disaster-prone geological structures according to claim 1, characterized in that, In S2, generating the candidate segment pairs and extracting the segment association features includes: Based on the spatial distribution relationship of each dip candidate segment relative to the upper constraint area and the lower exposure area, a starting segment and a target segment located on the downdip side of the starting segment are determined, and the starting segment and the target segment form the candidate segment pair; Based on the geometric correlation state, attitude correlation state, and medium correlation state between two dip candidate segments in the candidate segment pair, geometric correlation features, attitude correlation features, and medium correlation features are extracted respectively. The segmented association features are generated based on the geometric association features, attitude association features, and medium association features.

4. The landslide intelligent identification method based on disaster-prone geological structures according to claim 1, characterized in that, In S2, the mezzanine connection discrimination model is obtained through the following steps: Obtain historical weak interlayer segments confirmed by geological survey data; The training association features of the historical weak interlayer segment pairs are extracted using the same segment association feature extraction method as the candidate segment pairs. Based on whether the historical weak interlayer segments belong to the same continuous weak interlayer, set corresponding connection markers; The classification model is trained based on the training association features and connection labels to obtain the mezzanine connection discrimination model.

5. The landslide intelligent identification method based on disaster-prone geological structures according to claim 3, characterized in that, In S2, the directional correlation is established based on the connectivity discrimination value and the stable rock layer blocking state, including: The blocking state of the stable rock layer is determined based on the geometric, structural, and mechanical blocking characteristics of the interlayers between the candidate segments; When the stable rock layer blocking state meets the preset blocking conditions, the directional correlation relationship is not established between the corresponding dip-side candidate segments; When the stable rock layer blocking state does not meet the preset blocking condition, and the connection discrimination value meets the preset connection condition, a directional association relationship is established between the corresponding dip-side candidate segments, pointing from the starting segment to the target segment; The weak interlayer orientation correlation diagram is constructed based on the dip candidate segments and the orientation correlation relationship.

6. The intelligent landslide identification method based on disaster-prone geological structures according to claim 1, characterized in that, In S3, the search for the candidate weak sandwich structure chain includes: The candidate segments with dip direction located in the upper constraint area are determined as the starting candidate segments, and the candidate segments with dip direction located in the lower exposure area or satisfying a preset spatial relationship with the lower exposure area are determined as the ending candidate segments. Starting from the initial candidate segment, a directed path search is performed along the directional association relationships in the weak interlayer directional association graph; The directed paths are filtered based on the path direction constraints and path blocking constraints of the searched directed paths relative to the target slope. The directed path from the starting candidate segment to the ending candidate segment is determined as the candidate weak sandwich structure chain.

7. The landslide intelligent identification method based on disaster-prone geological structures according to claim 1, characterized in that, In S3, determining the continuity state of the candidate weak sandwich structure chain includes: Based on the connection discrimination value corresponding to each directional correlation relationship constituting the candidate weak interlayer structure chain and the blocking state of the stable rock layer, the connection state of each directional correlation relationship is determined; Based on the connection status of each directional relationship, determine the overall connection status and bottleneck connection status of the candidate weak sandwich structure chain; The connection state of the candidate weak sandwich structure chain is determined based on the lower connection characteristics between the end of the candidate weak sandwich structure chain and the lower exposed area. The penetration state is determined based on the overall connection state, bottleneck connection state, joint state, and stable rock layer blocking state, and the candidate weak interlayer structure chain that meets the preset penetration conditions is determined as the candidate slip path.

8. The intelligent landslide identification method based on disaster-prone geological structures according to claim 1, characterized in that, In S4, the infiltration state under rainfall is determined, including: The rainfall process characteristics of the target slope are determined based on the rainfall data. Based on the characteristics of the rainfall process and the hydrological infiltration characteristics of the overlying medium of the candidate slip path, the spatial influence range of rainfall infiltration is determined; The spatial influence range is compared with the spatial location of each dip candidate segment in the candidate slip path to determine the infiltration state corresponding to each dip candidate segment; The change in water state of each dip candidate segment is determined based on the infiltration state.

9. The intelligent landslide identification method based on disaster-prone geological structures according to claim 1, characterized in that, In S4, the hydraulic state parameters and shear strength parameters are updated, and the critical rainfall conditions are determined, including: Update the corresponding hydraulic state parameters based on the change in water state of each dip candidate segment; Based on the updated hydraulic state parameters and the preset strength weakening relationship, the corresponding shear strength parameters are updated; The stability index is determined based on the updated hydraulic state parameters, shear strength parameters, spatial structure parameters of candidate slip paths, and stress state of the corresponding rock mass. Using the current rainfall conditions as the initial simulated rainfall conditions, the simulated rainfall conditions are adjusted according to a preset adjustment method, and the infiltration state, hydraulic state parameters, shear strength parameters, and stability index are re-determined for the adjusted simulated rainfall conditions. The simulated rainfall condition that first makes the stability index meet the preset critical stability condition is determined as the critical rainfall condition.

10. The intelligent landslide identification method based on disaster-prone geological structures according to claim 1, characterized in that, In S5, determining the target slip path and generating the landslide identification result includes: Based on the connectivity, stability indicators, and critical rainfall conditions of each candidate slip path, the candidate slip paths are ranked. The candidate sliding paths that meet the preset selection criteria are determined as the target sliding path; Based on the candidate segments with tilt included in the target sliding path and their spatial extension relationship, the sliding space identification information is determined; Based on the connection status of each directional relationship in the target slip path and the change status of the shear strength parameters of each dip candidate segment, the path weak location identification information is determined; The landslide identification result is generated based on the target slip path, slip space identification information, path weak point identification information, and critical rainfall conditions.

Images