A method for identifying a failure mode based on a radar of a slope of an open pit mine

Abstract

The application discloses a kind of based on open-pit mine slope radar's damage mode identification method, based on the displacement of corner reflector measured according to the coordinate change of each corner reflector in three-dimensional coordinate system monitored each time by slope radar, according to the size of displacement, the corner reflector of preliminary screening exists abnormal motion;The corner reflector of preliminary screening is carried out clustering analysis, and the corner reflector of all preliminary screening is split into multiple different by corner reflector Abnormal deformation set formed;Build damage mode identification model, respectively to the spatial displacement change characteristics of all the abnormal deformation set by multiple corner reflectors is trained learning, to identify the corresponding relationship between the spatial displacement change characteristics corresponding to each damage mode and damage mode, to determine the four damage modes corresponding to abnormal deformation set;The application determines the corresponding relationship between the spatial displacement change characteristics of slope and damage mode, so that the damage mode corresponding to different regions can be managed by partition targeting.

Description

Technical Field

[0001] This invention relates to the field of slope damage mode classification and identification technology, specifically to a damage mode identification method based on open-pit mine slope radar. Background Technology

[0002] Slope radar monitoring technology can remotely monitor slope deformation. Before radar monitoring, corner reflectors are usually placed on the slope. The radar equipment remotely scans the surface of the slope monitoring area and uses the strong reflection signal from the corner reflectors to calculate the high-precision displacement of the entire area. Monitoring personnel can then determine whether there is a possibility of slope deformation in the current period.

[0003] Most existing technologies determine the displacement and deformation characteristics of each corner reflector by repeatedly monitoring the corner reflector with a set serial number using radar. The corner reflectors are then clustered based on these displacement and deformation characteristics to form a cluster of corner reflectors that correspond to the same area and have the same spatial displacement and deformation trend. However, most existing technologies determine the deformation stage of the slope based on the core displacement characteristics of the cluster, such as the four modes of initial deformation, isotropic deformation, accelerated deformation, and landslide, but lack the identification of the slope failure mode. Therefore, it is very likely that after slope deformation is detected, incorrect or even harmful engineering measures will be adopted, which may accelerate the occurrence of landslides. Summary of the Invention

[0004] Therefore, this invention provides a method for identifying damage modes of open-pit mine slopes based on radar, which effectively solves the technical problem that the existing technology lacks the identification of slope damage mode categories, and therefore may lead to the adoption of incorrect or even harmful engineering measures after slope deformation is detected, thus accelerating landslides.

[0005] To solve the above-mentioned technical problems, the present invention specifically provides the following technical solution: a method for identifying damage patterns on open-pit mine slopes based on radar, comprising the following steps: Step 100: Construct a three-dimensional coordinate system for the target monitoring area, and grid the corner reflectors within the target monitoring area. Calculate the displacement of each corner reflector based on the coordinate changes of each corner reflector detected by the slope radar in the three-dimensional coordinate system. Based on the magnitude of the displacement, preliminarily screen out corner reflectors exhibiting abnormal movement. Step 200: Perform cluster analysis on the initially screened corner reflectors, grouping anomalous corner reflector points that are spatially continuous and have similar motion characteristics into clusters, and splitting all the initially screened corner reflectors into multiple sets of anomalous deformations composed of different corner reflectors. Step 300: Construct a damage mode recognition model. Train and learn the spatial displacement change features of all abnormal deformation sets composed of multiple corner reflectors to identify the correspondence between the spatial displacement change features and the damage mode for each damage mode, so as to determine that the abnormal deformation set corresponds to four damage modes: circular arc sliding, planar sliding, wedge sliding, or tilting deformation.

[0006] Furthermore, in step 100, the method for initially screening out corner reflectors exhibiting abnormal motion is as follows: When the slope radar monitors the target monitoring area, the corner reflectors are identified sequentially and the serial number i is defined. The initial three-dimensional coordinate values ​​(x1i, y1i, z1i) of each corner reflector are determined. The real-time three-dimensional coordinate values ​​(xji, yji, zji) of the corner reflectors are also determined each time during the monitoring process. The three-dimensional displacement change vector (dXi, dYi, dZi) corresponding to each corner reflector during each monitoring is also determined. Obtain the three-dimensional displacement change vector (dXi, dYi, dZi) of each corner reflector after decomposition in the radar line-of-sight direction; Based on long-term background noise analysis, a total displacement threshold is set: S_total = ; And set the minimum displacement rate: V_total = / t; Based on the cumulative displacement of each corner reflector i during the monitoring time, and the displacement rate of the corner reflector i during the monitoring time, it is determined whether the displacement trend of the corner reflector i is significant, and all corner reflectors exceeding the minimum displacement rate and total displacement threshold are marked as abnormal motion points.

[0007] Furthermore, in step 200, the method for performing cluster analysis on the corner reflectors initially selected as anomalous motion points to form an anomalous deformation set composed of corner reflectors is as follows: Construct a motion feature vector for each corner reflector, which consists of multiple feature dimensions: CRi = [ xi, yi, zi, vxi, vyi, vzi, αi, Si]; in, xi, yi, Let zi be the difference between the real-time and initial three-dimensional coordinates of corner reflector i, vxi, vvyi, and vzi be the three-dimensional displacement velocity components of corner reflector i, αi be the vertical displacement direction angle of corner reflector i, and Si be the magnitude of the total displacement rate of corner reflector i. Where Si = ; Perform Z-score normalization or Min-Max normalization on each feature dimension separately; Calculate the overall difference between the two corner reflectors using weighted Euclidean distance: D(CRi, CRr) = ; in, The spatial distance between the two corner reflectors. The kinematic distance between the two corner reflectors and These are the weighting coefficients; First, corner reflectors whose feature vectors change at the same time point are obtained, and these corner reflectors are grouped together. Then, DBSCAN clustering is performed from this group. By setting the neighborhood radius between every two corner reflectors in each abnormal deformation set and the number of cluster points in each abnormal deformation set, corner reflectors that can be classified into the same cluster are selected.

[0008] Furthermore, D_xyz = ; Typically, wx = wy = 1, and wz is adjusted according to the actual situation. Let be the x-axis displacement change vector corresponding to each monitoring of the corner reflector with serial number i. This is the x-axis displacement change vector corresponding to each monitoring of the corner reflector with serial number r; The combined distance includes the difference in velocity vector direction D_direction and the difference in velocity magnitude D_speed between the two corner reflectors, specifically: D_v = ; The velocity vector direction difference D_direction = 1 - cos(θ), where θ is the angle between the vertical displacement directions α of the two corner reflectors. When the directions are exactly the same, the velocity vector direction difference D_direction is 0, and when they are opposite, the velocity vector direction difference D_direction is 2.

[0009] Furthermore, a cluster label is assigned to the corner reflectors within each set of anomalous deformations; Draw the corner reflector for each of the anomalous deformation sets in the three-dimensional coordinate system (X, Y, Z); Calculate the displacement vector field for each of the abnormal deformation sets, and calculate the displacement characteristics of the displacement vector field. Check whether the displacement vectors within the same abnormal deformation set are consistent in direction and magnitude. If the displacement vectors within the same set of abnormal deformations are inconsistent in direction and magnitude, adjust... , , and the neighborhood radius and cluster point number parameters for each of the aforementioned abnormal deformation sets; If the displacement vectors within the same set of abnormal deformations are consistent in both direction and magnitude, then the set of abnormal deformations is considered to belong to the same cluster.

[0010] Furthermore, in step 300, the method for constructing a damage mode recognition model to identify the correspondence between the spatial displacement change characteristics and the damage mode for each damage mode is as follows: Obtain the initial three-dimensional coordinate values ​​(x1i, y1i, z1i) of all corner reflectors within each set of abnormal deformations, as well as the real-time three-dimensional coordinate values ​​(xji, yji, zji) of the corner reflectors. The initial three-dimensional coordinate values ​​(x1i, z1i) of all corner reflectors in each abnormal deformation set, and the real-time three-dimensional coordinate values ​​(xji, zji) of the corner reflectors are integrated into the same three-dimensional coordinate system, that is, the vertical displacement direction angle of each corner reflector in the XZ plane is determined as the evaluation parameter of the failure mode. The vertical displacement direction angle αji is calculated as follows: Calculate the horizontal displacement vector (dXi, dZi) of each corner reflector in the XZ plane, where dE is the X-axis difference between the initial X-axis coordinate value x1i and the real-time X-axis coordinate value xji of each corner reflector, and dN is the Y-axis difference between the initial Y-axis coordinate value y1i and the real-time Y-axis coordinate value yji of each corner reflector. Then, the vertical displacement direction angle αji = atan2(dXi, dZi). Based on the displacement rate of all corner reflectors in the xz plane within each set of anomalous deformations, and the vertical deformation dYi of all corner reflectors in each set of anomalous deformations perpendicular to the slope surface, a displacement direction-vertical deformation curve is established as an evaluation parameter for the failure mode.

[0011] Furthermore, the method for establishing the displacement direction-vertical deformation curve is as follows: The main sliding axis is determined based on the average vertical displacement direction angle of all corner reflectors within each set of abnormal deformations; In a three-dimensional coordinate system, one or more analysis sections are cut in a direction parallel to the main sliding axis; The average vertical deformation of each row of corner reflectors is calculated sequentially along the Z-axis. A two-dimensional coordinate system is constructed with the main sliding axis as the X-axis and the direction perpendicular to the main sliding axis as the Y-axis. The average vertical deformation of each row of corner reflectors is integrated into the two-dimensional coordinate system, and the displacement rate-vertical distance curve is plotted as the displacement direction-vertical deformation curve.

[0012] Furthermore, by combining the vertical displacement direction angle analysis based on each corner reflector in the XZ plane to determine the consistency of the displacement direction, and by determining the gradient change of the vertical deformation dYi along the displacement direction using the displacement direction-vertical deformation curve, the failure mode can be identified. The method for achieving consistency in displacement direction analysis based on the vertical displacement direction angle of each corner reflector in the XZ plane is as follows: The sign of the vertical displacement direction angle is determined, and a two-dimensional coordinate system is constructed with the initial coordinate point of each corner reflector. When the vector field of the vertical displacement direction angle is in the fourth quadrant of the two-dimensional coordinate system, the vertical displacement direction angle is positive, and when the vector field of the vertical displacement direction angle is in the third quadrant of the two-dimensional coordinate system, the vertical displacement direction angle is negative. The number of corner reflectors whose vector field of vertical displacement direction angle is in the third quadrant and the fourth quadrant are counted separately, and the range is calculated. The standard deviation of vertical displacement direction angle is also calculated. When the range is less than the maximum set value and the standard deviation of the vertical displacement direction angle is greater than the minimum stability evaluation value, it means that the displacement direction in this area is inconsistent. When the range is greater than the maximum set value and the standard deviation of the vertical displacement direction angle is less than the minimum stability evaluation value, it means that the displacement direction in this area is consistent.

[0013] Furthermore, when the vertical displacement direction angle of all corner reflectors in the abnormal deformation set is zero, the distribution pattern of the average vertical deformation of different rows of corner reflectors is calculated. When the fitted curve of the displacement direction-vertical deformation curve is inclined upward, the failure mode is classified as tilt deformation. When the displacement directions of the corner reflectors are inconsistent, and when the fitting curve of the displacement direction-vertical deformation curve shows that the average vertical deformation is large in the middle and small at both ends, then the failure mode of the region where the abnormal deformation set is located is circular arc sliding. When the displacement directions of the corner reflectors are consistent, and when the average vertical deformation on the fitting curve of the displacement direction-vertical deformation curve is the same, then the failure mode of the region where the abnormal deformation set is located is planar sliding. When the displacement directions of the corner reflectors are consistent, and when the fitted curve of the displacement direction-vertical deformation curve is inclined downward, then the failure mode of the region where the abnormal deformation set is located is wedge sliding.

[0014] Compared with the prior art, the present invention has the following advantages: This invention identifies the motion feature vector of each corner reflector based on the changes in the reflected signals of the corner reflectors. Cluster analysis is then performed on the motion feature vectors of each corner reflector to group the corner reflectors into an abnormal deformation set corresponding to a deformation region. By analyzing the displacement direction of all corner reflectors in the abnormal deformation set and the vertical deformation of all corner reflectors along the displacement direction perpendicular to the slope surface, the failure mode of the region where the abnormal deformation set is located can be identified. The correspondence between the spatial displacement change characteristics of the slope and the failure mode can be determined, thereby enabling zoned and targeted treatment of the failure modes corresponding to different regions. Attached Figure Description

[0015] To more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings in the following description are merely exemplary, and those skilled in the art can derive other embodiments based on the provided drawings without creative effort.

[0016] Figure 1 This is a flowchart illustrating the destruction mode recognition method in an embodiment of the present invention; Figure 2 This is a schematic diagram of the three-dimensional coordinate system of the target monitoring area in an embodiment of the present invention; Detailed Implementation

[0017] The technical solutions of 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.

[0018] like Figure 1 and Figure 2 As shown, this invention provides a method for identifying damage modes based on open-pit mine slope radar. High-altitude and cold open-pit mines mostly exhibit the following four damage modes: circular arc sliding, planar sliding, wedge-shaped sliding, or tilting deformation. The specific classification of slope changes based on the monitoring results of slope radar can prevent the adoption of incorrect or even harmful engineering measures (such as blindly carrying out large-scale slope cutting for tilting deformation, which may exacerbate instability).

[0019] Furthermore, by designing "targeted" projects based on the damage mechanism, the maximum stability effect can be achieved with the minimum amount of engineering work, avoiding waste of funds. Stability calculations and governance designs are carried out based on the correct mechanism model, which significantly improves the success rate and long-term stability of governance projects. The key parts of potential damage and deformation precursors are clearly identified, making the monitoring point layout and early warning threshold setting more targeted.

[0020] Therefore, when specifically identifying the four damage modes, it is necessary to perform cluster analysis on the monitoring results of slope radar to cluster anomalous change points that are spatially continuous and have similar motion characteristics into groups, forming a set of anomalous deformations composed of corner reflectors. Then, machine learning is performed on the spatial displacement change characteristics of the anomalous deformation set to identify the correspondence between the spatial displacement change characteristics and the damage mode corresponding to each damage mode. This determines whether the anomalous deformation set corresponds to the arc sliding, planar sliding, wedge sliding, or tilting deformation damage mode, thus enabling zoned and targeted treatment of the damage modes corresponding to different areas of the target monitoring area.

[0021] The destruction pattern recognition method includes the following steps: Step 100: Construct a three-dimensional coordinate system for the target monitoring area, and deploy corner reflectors in a grid within the target monitoring area. Calculate the displacement of each corner reflector based on the coordinate changes of each corner reflector detected by the slope radar in the three-dimensional coordinate system. Based on the magnitude of the displacement, preliminarily screen out corner reflectors with abnormal motion. Step 200: Perform cluster analysis on the initially screened corner reflectors, grouping anomalous corner reflector points that are spatially continuous and have similar motion characteristics into clusters, and splitting all the initially screened corner reflectors into multiple sets of anomalous deformations composed of different corner reflectors. Step 300: Construct a failure mode recognition model. Train and learn the spatial displacement change features of all abnormal deformation sets composed of multiple corner reflectors to identify the correspondence between the spatial displacement change features and failure modes corresponding to each failure mode, so as to determine that the abnormal deformation set corresponds to four failure modes: circular arc sliding, planar sliding, wedge sliding, or tilting deformation.

[0022] By placing a corner reflector within the target monitoring area of ​​the slope and a slope radar on the opposite side of the target monitoring area, the slope radar monitoring technology uses radar equipment to remotely scan the surface of the slope monitoring area and calculates the displacement of the corner reflector by acquiring the strong reflection signal of the corner reflector, thereby obtaining a high-precision displacement within the entire area.

[0023] Therefore, in this embodiment, the displacement of the corner reflector is first calculated based on the reflection signal when each corner reflector is monitored, and corner reflectors with abnormal movement are initially screened according to the specific magnitude of the displacement. Cluster analysis is performed on all corner reflectors to form multiple sets of anomalous deformations composed of corner reflectors. Based on the motion characteristic vector of the anomalous deformation sets, four failure modes can be determined for the slope area where each anomalous deformation set is located: circular arc sliding, planar sliding, wedge sliding, or tilting deformation.

[0024] In step 100, the method for initially screening out corner reflectors exhibiting abnormal motion is as follows: When the slope radar monitors the target monitoring area, the corner reflectors identified sequentially are defined with serial number i. The initial three-dimensional coordinate values ​​(x1i, y1i, z1i) of each corner reflector acquired for the first time are determined, as well as the real-time three-dimensional coordinate values ​​(xji, yji, zji) of each corner reflector acquired during the monitoring process. The three-dimensional displacement change vector (dXi, dYi, dZi) corresponding to each corner reflector during each monitoring is determined, where j represents the number of times the slope radar monitors the target monitoring area.

[0025] Obtain the three-dimensional displacement change vector (dXi, dYi, dZi) of each corner reflector in the radar line-of-sight direction, and then decompose it into a three-dimensional displacement coordinate sequence. The three-dimensional coordinate axes of this three-dimensional displacement coordinate sequence represent the displacement velocity value of corner reflector i, the X-axis and Z-axis of the three-dimensional coordinate system in which the corner reflector is located, respectively, so as to determine the distribution position of corner reflector i in the xz plane.

[0026] Based on long-term background noise analysis, a total displacement threshold is set: S_total = ; And set the minimum displacement rate: V_total = / t; Based on the cumulative displacement of each corner reflector i during the monitoring time and the displacement rate of that corner reflector i during the monitoring time, it is determined whether the displacement trend of that corner reflector i is significant, and all corner reflectors that exceed the minimum displacement rate and total displacement threshold are marked as abnormal motion points.

[0027] The goal of this step is to identify corner reflectors that have experienced significant displacement from all corner reflectors, eliminate background noise in stable regions, and mark all corner reflectors that exceed the total displacement threshold and the minimum displacement rate as "abnormal motion points." These corner reflectors will proceed to the next step of analysis, while the remaining corner reflectors are considered "background points" and temporarily set aside.

[0028] In step 200, the method for performing cluster analysis on the corner reflectors initially selected as anomalous motion points to form an anomalous deformation set composed of corner reflectors is as follows: Construct a motion feature vector for each corner reflector, which consists of multiple feature dimensions: CRi = [ xi, yi, zi, vxi, vyi, vzi, αi, Si].

[0029] in, xi, yi, Let zi be the difference between the real-time and initial three-dimensional coordinates of corner reflector i, vxi, vvyi, and vzi be the three-dimensional displacement velocity components of corner reflector i, αi be the vertical displacement direction angle of corner reflector i, and Si be the magnitude of the total displacement rate of corner reflector i. Where Si = .

[0030] Perform Z-score normalization or Min-Max normalization on each feature dimension separately.

[0031] Calculate the overall difference between the two corner reflectors using weighted Euclidean distance: D(CRi, CRr) = ; in, The spatial distance between the two corner reflectors The kinematic distance between the two corner reflectors and These are the weighting coefficients.

[0032] First, identify the corner reflectors whose feature vectors change at the same time point, group these corner reflectors into a cluster, and then perform DBSCAN clustering within this cluster. By setting the neighborhood radius between every two corner reflectors in each abnormal deformation set and the number of cluster points in each abnormal deformation set, corner reflectors that can be classified into the same cluster are selected.

[0033] D_xyz = ; Typically, wx = wy = 1, and wz is adjusted according to the actual situation. Let be the x-axis displacement change vector corresponding to each monitoring of the corner reflector with serial number i. This is the x-axis displacement change vector corresponding to each monitoring of the corner reflector with serial number r; The combined distance includes the difference in velocity vector direction D_direction and the difference in velocity magnitude D_speed between the two corner reflectors, specifically: D_v = ; The velocity vector direction difference D_direction = 1 - cos(θ), where θ is the angle between the vertical displacement directions α of the two corner reflectors. When the directions are exactly the same, the velocity vector direction difference D_direction is 0, and when they are opposite, the velocity vector direction difference D_direction is 2.

[0034] Assign a cluster label to the corner reflectors within each set of anomalous deformations; Draw the corner reflectors for each set of anomalous deformations in a three-dimensional coordinate system (X, Y, Z); Calculate the displacement vector field for each abnormal deformation set, and calculate the displacement characteristics of the displacement vector field. Check whether the displacement vectors within the same abnormal deformation set are consistent in direction and magnitude. If the displacement vectors within the same abnormal deformation set are inconsistent in direction and magnitude, adjust... , , and the neighborhood radius and cluster point number parameters for each abnormal deformation set; If the displacement vectors within the same set of abnormal deformations are consistent in both direction and magnitude, then the set of abnormal deformations is considered to belong to the same cluster.

[0035] The goal of this step is to group together anomalous corner reflectors that are spatially continuous and have similar motion characteristics, with each group potentially corresponding to an independent deformable body.

[0036] Specifically, the clustering can be divided into two directions: Spatial clustering: Density-based clustering algorithms (such as DBSCAN) are used to group outlier CR points that are spatially close together into one class. The advantage of DBSCAN is that it can automatically remove discrete noise points. In pure spatial clustering (such as by distance), two points are "similar" only if they are physically close.

[0037] Kinematic clustering: Based on spatial clustering, attributes such as displacement direction and velocity can be added to constrain the movement of points within the same cluster, that is, the displacement direction and velocity magnitude are similar, thus ensuring that the movement patterns of corner reflectors in the same "abnormal deformation set" are consistent.

[0038] This results in N "abnormal deformation sets", each of which is a set of corner reflectors.

[0039] Cluster analysis is performed using the three-dimensional coordinates (X, Y, Z) of corner reflectors (CRs). The core objective is to automatically identify and group CR points that are spatially adjacent and have similar kinematic behavior (displacement vectors) into the same "deformation unit" or "potentially destructive body".

[0040] Cluster analysis automatically and systematically filters and merges corner reflector sets representing different failure modes from discrete monitoring points, providing clear and structured information input for slope stability assessment.

[0041] After clustering is completed, each cluster becomes an independent analysis unit. Then, the overall characteristics of the cluster can be calculated (such as the variance of directional consistency within the cluster, the displacement-elevation correlation coefficient, and the spatial distribution of displacement magnitude). Based on the overall characteristics of the cluster, the failure mode corresponding to the installation location of the "abnormal deformation set" can be determined.

[0042] If geological damage occurs in the target monitoring area, the slope radar can obtain the core feature vectors within the same abnormal deformation set each time it monitors, namely: the directional consistency αi of all corner reflectors, and the displacement direction-elevation correlation (i.e., the change in vertical distance of corner reflectors at different heights relative to the slope surface along the displacement direction of all corner reflectors).

[0043] In step 300, a damage mode identification model is constructed to identify the correspondence between the spatial displacement change characteristics and the damage mode for each damage mode. The implementation method is as follows: Obtain the initial 3D coordinates (x1i, y1i, z1i) of all corner reflectors within each abnormal deformation set, as well as the real-time 3D coordinates (xji, yji, zji) of the corner reflectors. The initial three-dimensional coordinate values ​​(x1i, z1i) of all corner reflectors in each abnormal deformation set, as well as the real-time three-dimensional coordinate values ​​(xji, zji) of the corner reflectors, are integrated into the same three-dimensional coordinate system. That is, the vertical displacement direction angle of each corner reflector in the XZ plane is determined as the evaluation parameter of the failure mode. The vertical displacement direction angle αji is calculated as follows: Calculate the horizontal displacement vector (dXi, dZi) of each corner reflector in the XZ plane, where dE is the X-axis difference between the initial X-axis coordinate value x1i and the real-time X-axis coordinate value xji of each corner reflector, and dN is the Y-axis difference between the initial Y-axis coordinate value y1i and the real-time Y-axis coordinate value yji of each corner reflector. Then, the vertical displacement direction angle αji = atan2(dXi, dZi). Based on the displacement rate of all corner reflectors in the xz plane within each abnormal deformation set, and the vertical deformation dYi of all corner reflectors in each abnormal deformation set perpendicular to the slope surface, a displacement direction-vertical deformation curve is established as an evaluation parameter for the failure mode.

[0044] By combining the analysis of the vertical displacement direction angle of each corner reflector in the XZ plane to determine the consistency of the displacement direction, and by determining the gradient change of the vertical deformation dYi along the displacement direction using the displacement direction-vertical deformation curve, the failure mode can be identified.

[0045] The method for achieving consistency in displacement direction analysis based on the vertical displacement direction angle of each corner reflector in the XZ plane is as follows: Determine the sign of the vertical displacement direction angle. Construct a two-dimensional coordinate system with the initial coordinate point of each corner reflector. When the vector field of the vertical displacement direction angle is in the fourth quadrant of the two-dimensional coordinate system, the vertical displacement direction angle is positive. When the vector field of the vertical displacement direction angle is in the third quadrant of the two-dimensional coordinate system, the vertical displacement direction angle is negative.

[0046] The number of corner reflectors whose vector field of vertical displacement direction angle is in the third quadrant and the fourth quadrant are counted separately, and the range is calculated. The standard deviation of vertical displacement direction angle is also calculated.

[0047] When the range is less than the maximum set value and the standard deviation of the vertical displacement direction angle is greater than the minimum stability evaluation value, it means that the displacement direction in this region is inconsistent.

[0048] When the range is greater than the maximum set value and the standard deviation of the vertical displacement direction angle is less than the minimum stability evaluation value, it means that the displacement direction in this area is consistent.

[0049] This implementation method mainly transforms discrete corner reflector point data into field information that can reveal the overall motion pattern and geometric shape, which mainly relies on spatial interpolation, vector decomposition, and statistical fitting techniques.

[0050] To represent the vertical displacement direction angular vector field distribution of each abnormal deformation set, quantify the consistency of the motion direction of each point, and identify the dominant sliding direction, this implementation calculates the displacement direction of all corner reflectors within each abnormal deformation set, calculates the vertical displacement direction angle of each corner reflector in the XZ plane, and distinguishes between positive and negative vertical displacement direction angles. When the number of positive and negative vertical displacement direction angles is approximately equal, it means that the vector direction of the displacement vector field is not completely parallel, but rather slightly radial, and therefore it does not belong to planar or plate motion. If the vertical displacement direction angles are concentrated in positive or negative values, based on the standard deviation of the vertical displacement direction angles, it can be determined that the vector direction of the displacement vector field has high parallelism and high directional consistency, and belongs to planar or plate motion.

[0051] In addition, this embodiment also uses the displacement direction-vertical deformation curve as an evaluation parameter for the failure mode. The displacement direction-vertical deformation curve is mainly used to determine the deformation posture that is perpendicular to the slope surface. Since the four failure modes, namely circular arc sliding, planar sliding, wedge sliding and tilting deformation, have different deformation postures, the failure mode can be further identified based on the deformation posture perpendicular to the slope surface.

[0052] Specifically, the method for establishing the displacement direction-vertical deformation curve is as follows: The main sliding axis is determined based on the average vertical displacement direction angle of all corner reflectors within each abnormal deformation set; In a three-dimensional coordinate system, one or more analysis sections are cut in a direction parallel to the main sliding axis; The average vertical deformation of each row of corner reflectors is calculated sequentially along the Z-axis. A two-dimensional coordinate system is constructed with the main sliding axis as the X-axis and the direction perpendicular to the main sliding axis as the Y-axis. The average vertical deformation of each row of corner reflectors is integrated into the two-dimensional coordinate system, and the displacement rate-vertical distance curve is plotted as the displacement direction-vertical deformation curve.

[0053] When the vertical displacement direction angle of all corner reflectors in the abnormal deformation set is zero, the distribution pattern of the average vertical deformation of different rows of corner reflectors is calculated. When the fitted curve of the displacement direction-vertical deformation curve is inclined upward, the failure mode is classified as tilt deformation. The displacement vector of tilt deformation is mainly in the Y direction, but unlike sliding, it is mainly manifested as the "bending" and "rotation" of the rock block. dY increases significantly with the increase of Z-direction elevation. The distribution pattern of the average vertical deformation is larger at the top and smaller at the bottom. The displacement of the corner reflectors at the top of the slope or the upper part of the slope is much greater than the displacement at the toe of the slope.

[0054] When the displacement directions of the corner reflectors are inconsistent, and when the fitting curve of the displacement direction-vertical deformation curve shows that the average vertical deformation is large in the middle and small at both ends, the failure mode of the region where the abnormal deformation set is located is circular arc sliding. The typical characteristics of circular arc sliding are: the curve is approximately normally distributed (Gaussian curve) or parabolic, with a peak value in the middle of the cross-section, and gradually decreasing to the background value towards the rear edge (starting point) and front edge (end point) of the cross-section. Therefore, in typical circular arc sliding, the maximum displacement may also occur in the mid-elevation region.

[0055] When the displacement directions of the corner reflectors are consistent, and the average vertical deformation on the fitting curve of the displacement direction-vertical deformation curve is the same, then the failure mode of the region where the abnormal deformation set is located is planar sliding. When all the corner reflectors located on the sliding surface have the same height in the xz plane displacement vector direction, that is, when all the corner reflectors in the abnormal deformation set have the same displacement direction (the direction and magnitude of the vertical displacement direction angle are the same), and the average vertical deformation of each row of corner reflectors corresponding to different elevations is also the same.

[0056] When the displacement directions of the corner reflectors are consistent, and when the fitted curve of the displacement direction-vertical deformation curve is inclined downward, the failure mode of the region where the abnormal deformation set is located is wedge sliding. All corner reflectors located on the sliding surface have the same displacement vector direction in the xz plane. The upper part of the slope deforms inward along the perpendicular slope surface, while the lower part of the slope deforms outward along the perpendicular slope surface.

[0057] This implementation method, based on the changes in the reflection signals of corner reflectors, can identify the motion feature vector of each corner reflector. Based on the motion feature vector of each corner reflector, cluster analysis is performed to group the corner reflectors into an abnormal deformation set corresponding to a deformation region. By analyzing the displacement direction of all corner reflectors in the abnormal deformation set, as well as the vertical deformation of all corner reflectors along the displacement direction perpendicular to the slope surface, the failure mode of the region where the abnormal deformation set is located is identified. The correspondence between the spatial displacement change characteristics and the failure mode corresponding to each failure mode is determined, thereby enabling zoned targeted treatment of the failure modes corresponding to different regions.

[0058] The above embodiments are merely exemplary embodiments of this application and are not intended to limit this application. The scope of protection of this application is defined by the claims. Those skilled in the art can make various modifications or equivalent substitutions to this application within its substance and scope of protection, and such modifications or equivalent substitutions should also be considered to fall within the scope of protection of this application.

Claims

1. A method for identifying damage patterns on open-pit mine slopes based on radar, characterized in that, Includes the following steps: Step 100: Construct a three-dimensional coordinate system for the target monitoring area, and grid the corner reflectors within the target monitoring area. Calculate the displacement of each corner reflector based on the coordinate changes of each corner reflector detected by the slope radar in the three-dimensional coordinate system. Based on the magnitude of the displacement, preliminarily screen out corner reflectors exhibiting abnormal movement. Step 200: Perform cluster analysis on the initially screened corner reflectors, grouping anomalous corner reflector points that are spatially continuous and have similar motion characteristics into clusters, and splitting all the initially screened corner reflectors into multiple sets of anomalous deformations composed of different corner reflectors. Step 300: Construct a damage mode recognition model. Train and learn the spatial displacement change features of all abnormal deformation sets composed of multiple corner reflectors to identify the correspondence between the spatial displacement change features and the damage mode for each damage mode, so as to determine that the abnormal deformation set corresponds to four damage modes: circular arc sliding, planar sliding, wedge sliding, or tilting deformation.

2. The method for identifying damage patterns based on open-pit mine slope radar according to claim 1, characterized in that, In step 100, the method for initially screening out corner reflectors exhibiting abnormal motion is as follows: When the slope radar monitors the target monitoring area, the corner reflectors are identified sequentially and the serial number i is defined. The initial three-dimensional coordinate values ​​(x1i, y1i, z1i) of each corner reflector are determined. The real-time three-dimensional coordinate values ​​(xji, yji, zji) of the corner reflectors are also determined each time during the monitoring process. The three-dimensional displacement change vector (dXi, dYi, dZi) corresponding to each corner reflector during each monitoring is also determined. Obtain the three-dimensional displacement change vector (dXi, dYi, dZi) of each corner reflector after decomposition in the radar line-of-sight direction; Based on long-term background noise analysis, a total displacement threshold is set: S_total = ; And set the minimum displacement rate: V_total = / t; Based on the cumulative displacement of each corner reflector i during the monitoring time, and the displacement rate of the corner reflector i during the monitoring time, it is determined whether the displacement trend of the corner reflector i is significant, and all corner reflectors exceeding the minimum displacement rate and total displacement threshold are marked as abnormal motion points.

3. The method for identifying damage patterns based on open-pit mine slope radar according to claim 1, characterized in that, In step 200, the method for performing cluster analysis on the corner reflectors initially selected as anomalous motion points to form an anomalous deformation set composed of corner reflectors is as follows: Construct a motion feature vector for each corner reflector, which consists of multiple feature dimensions: CRi = [ xi, yi,, zi, vxi, vyi, vzi, αi, Si]; in, xi, yi,, Let zi be the difference between the real-time and initial three-dimensional coordinates of corner reflector i, vxi, vvyi, and vzi be the three-dimensional displacement velocity components of corner reflector i, αi be the vertical displacement direction angle of corner reflector i, and Si be the magnitude of the total displacement rate of corner reflector i. Where Si = ; Perform Z-score normalization or Min-Max normalization on each feature dimension separately; Calculate the overall difference between the two corner reflectors using weighted Euclidean distance: D(CRi, CRr) = ; in, The spatial distance between the two corner reflectors The kinematic distance between the two corner reflectors and These are the weighting coefficients; First, corner reflectors whose feature vectors change at the same time point are obtained, and these corner reflectors are grouped together. Then, DBSCAN clustering is performed from this group. By setting the neighborhood radius between every two corner reflectors in each abnormal deformation set and the number of cluster points in each abnormal deformation set, corner reflectors that can be classified into the same cluster are selected.

4. The method for identifying damage patterns based on open-pit mine slope radar according to claim 4, characterized in that, D_xyz = ; Typically, wx = wy = 1, and wz is adjusted according to the actual situation. Let be the x-axis displacement change vector corresponding to each monitoring of the corner reflector with serial number i. This is the x-axis displacement change vector corresponding to each monitoring of the corner reflector with serial number r; The combined distance includes the difference in velocity vector direction D_direction and the difference in velocity magnitude D_speed between the two corner reflectors, specifically: D_v = ; The velocity vector direction difference D_direction = 1 - cos(θ), where θ is the angle between the vertical displacement directions α of the two corner reflectors. When the directions are exactly the same, the velocity vector direction difference D_direction is 0, and when they are opposite, the velocity vector direction difference D_direction is 2.

5. A method for identifying damage patterns on open-pit mine slopes based on radar according to claim 5, characterized in that, Assign a cluster label to the corner reflectors within each set of anomalous deformations; Draw the corner reflector for each of the anomalous deformation sets in the three-dimensional coordinate system (X, Y, Z); Calculate the displacement vector field for each of the abnormal deformation sets, and calculate the displacement characteristics of the displacement vector field. Check whether the displacement vectors within the same abnormal deformation set are consistent in direction and magnitude. If the displacement vectors within the same set of abnormal deformations are inconsistent in direction and magnitude, adjust... , , and the neighborhood radius and cluster point number parameters for each of the aforementioned abnormal deformation sets; If the displacement vectors within the same set of abnormal deformations are consistent in both direction and magnitude, then the set of abnormal deformations is considered to belong to the same cluster.

6. The method for identifying damage patterns based on open-pit mine slope radar according to claim 5, characterized in that, In step 300, the method for constructing a damage mode recognition model to identify the correspondence between the spatial displacement change characteristics and the damage mode for each damage mode is as follows: Obtain the initial three-dimensional coordinate values ​​(x1i, y1i, z1i) of all corner reflectors within each set of abnormal deformations, as well as the real-time three-dimensional coordinate values ​​(xji, yji, zji) of the corner reflectors. The initial three-dimensional coordinate values ​​(x1i, z1i) of all corner reflectors in each abnormal deformation set, and the real-time three-dimensional coordinate values ​​(xji, zji) of the corner reflectors are integrated into the same three-dimensional coordinate system, that is, the vertical displacement direction angle of each corner reflector in the XZ plane is determined as the evaluation parameter of the failure mode. The vertical displacement direction angle αji is calculated as follows: Calculate the horizontal displacement vector (dXi, dZi) of each corner reflector in the XZ plane, where dE is the X-axis difference between the initial X-axis coordinate value x1i and the real-time X-axis coordinate value xji of each corner reflector, and dN is the Y-axis difference between the initial Y-axis coordinate value y1i and the real-time Y-axis coordinate value yji of each corner reflector. Then, the vertical displacement direction angle αji = atan2(dXi, dZi). Based on the displacement rate of all corner reflectors in the xz plane within each set of anomalous deformations, and the vertical deformation dYi of all corner reflectors in each set of anomalous deformations perpendicular to the slope surface, a displacement direction-vertical deformation curve is established as an evaluation parameter for the failure mode.

7. The method for identifying damage patterns based on open-pit mine slope radar according to claim 7, characterized in that, The method for establishing the displacement direction-vertical deformation curve is as follows: The main sliding axis is determined based on the average vertical displacement direction angle of all corner reflectors within each set of abnormal deformations; In a three-dimensional coordinate system, one or more analysis sections are cut in a direction parallel to the main sliding axis; The average vertical deformation of each row of corner reflectors is calculated sequentially along the Z-axis. A two-dimensional coordinate system is constructed with the main sliding axis as the X-axis and the direction perpendicular to the main sliding axis as the Y-axis. The average vertical deformation of each row of corner reflectors is integrated into the two-dimensional coordinate system, and the displacement rate-vertical distance curve is plotted as the displacement direction-vertical deformation curve.

8. A method for identifying damage patterns on open-pit mine slopes based on radar according to claim 8, characterized in that, By combining the vertical displacement direction angle analysis based on each corner reflector in the XZ plane to determine the consistency of the displacement direction, and by determining the gradient change of the vertical deformation dYi along the displacement direction using the displacement direction-vertical deformation curve, the failure mode can be identified. The method for achieving consistency in displacement direction analysis based on the vertical displacement direction angle of each corner reflector in the XZ plane is as follows: The sign of the vertical displacement direction angle is determined, and a two-dimensional coordinate system is constructed with the initial coordinate point of each corner reflector. When the vector field of the vertical displacement direction angle is in the fourth quadrant of the two-dimensional coordinate system, the vertical displacement direction angle is positive, and when the vector field of the vertical displacement direction angle is in the third quadrant of the two-dimensional coordinate system, the vertical displacement direction angle is negative. The number of corner reflectors whose vector field of vertical displacement direction angle is in the third quadrant and the fourth quadrant are counted separately, and the range is calculated. The standard deviation of vertical displacement direction angle is also calculated. When the range is less than the maximum set value and the standard deviation of the vertical displacement direction angle is greater than the minimum stability evaluation value, it means that the displacement direction in this area is inconsistent. When the range is greater than the maximum set value and the standard deviation of the vertical displacement direction angle is less than the minimum stability evaluation value, it means that the displacement direction in this area is consistent.

9. A method for identifying damage patterns on open-pit mine slopes based on radar according to claim 9, characterized in that, When the vertical displacement direction angle of all corner reflectors in the abnormal deformation set is zero, calculate the distribution pattern of the average vertical deformation of different rows of corner reflectors. When the fitted curve of the displacement direction-vertical deformation curve is inclined upward, the failure mode is classified as tilt deformation. When the displacement directions of the corner reflectors are inconsistent, and when the fitting curve of the displacement direction-vertical deformation curve shows that the average vertical deformation is large in the middle and small at both ends, then the failure mode of the region where the abnormal deformation set is located is circular arc sliding. When the displacement directions of the corner reflectors are consistent, and when the average vertical deformation on the fitting curve of the displacement direction-vertical deformation curve is the same, then the failure mode of the region where the abnormal deformation set is located is planar sliding. When the displacement directions of the corner reflectors are consistent, and when the fitted curve of the displacement direction-vertical deformation curve is inclined downward, then the failure mode of the region where the abnormal deformation set is located is wedge sliding.

Images