Intelligent dangerous rock identification and stability analysis system based on non-contact measurement

Abstract

The invention discloses a dangerous rock intelligent identification and stability analysis system based on non-contact measurement. The dangerous rock intelligent identification and stability analysis system comprises a survey acquisition subsystem and a central control subsystem, the surveying and collecting subsystem is used for conducting multi-view close photogrammetry on dangerous rock through an unmanned aerial vehicle, collecting dangerous rock three-dimensional point cloud data and uploading the dangerous rock three-dimensional point cloud data to the center control subsystem. The central control subsystem is used for carrying out point cloud processing and structural plane clustering processing according to the collected dangerous rock three-dimensional point cloud data, judging dangerous rock types and whether the dangerous rock has potential damage according to processing results of the point cloud processing and the structural plane clustering processing, calculating stability coefficients of different dangerous rock types when the potential damage exists, and sending the stability coefficients to the central control subsystem; and obtaining a stability analysis result of the dangerous rock according to the stability coefficients of the different dangerous rock types. According to the method, the operation efficiency is improved through unmanned exploration, and the accuracy of dangerous rock identification and stability analysis is improved based on the dangerous rock three-dimensional point cloud model and a prefabricated algorithm.

Description

Technical Field

[0001] The present invention relates to the field of geological disaster research, and in particular to a dangerous rock intelligent identification and stability analysis system based on non-contact measurement. Background Art

[0002] As one of the typical geological disasters in mountainous areas, rock collapse has the characteristics of wide distribution, strong suddenness, strong catastrophicity and high concealment. Conventional surveying methods are difficult to measure effectively and accurately. Since the mountains in the western region are generally high and steep, they provide natural conditions for the breeding of collapse disasters. The Party and the State will place disaster prevention, mitigation and relief in a more prominent position, and the importance of rock collapse prevention and control is becoming increasingly prominent.

[0003] Dangerous rock refers to rock mass that has not collapsed yet but has the main conditions for collapse and has shown precursory phenomena. Dangerous rock mass is in an unstable state. Under the influence of gravity, earthquake, fissure water pressure and other inducing factors, geological disasters such as sliding and collapse may occur at any time. Mountainous areas are high and steep, with high mountains and deep valleys. There are a large number of dangerous rocks, and collapse disasters may occur at any time. Traditional contact survey methods are difficult to conduct efficient and accurate measurements on peaks that are often hundreds of meters high. They often rely on the experience of engineers and have certain errors. With the rise of non-contact measurement technologies such as drones and lidar, new solutions have been provided for the survey and identification of dangerous rocks. However, there is currently no mature and systematic dangerous rock identification solution, so an accurate, efficient and intelligent dangerous rock intelligent identification method and stability analysis method are urgently needed. Summary of the invention

[0004] In view of the above-mentioned deficiencies in the prior art, the present invention provides a dangerous rock intelligent identification and stability analysis system based on non-contact measurement. Through non-contact measurement technology, improved KNN algorithm and GASAK algorithm are used to realize the identification and cluster analysis of dangerous rock structural surfaces, and the stereographic projection method is used to make a preliminary stability judgment on the dangerous rocks, thereby improving the accuracy and efficiency of dangerous rock identification and stability analysis.

[0005] In order to solve the above technical problems, the present invention adopts the following technical solutions:

[0006] A dangerous rock intelligent identification and stability analysis system based on non-contact measurement comprises a survey and acquisition subsystem and a central control subsystem; the survey and acquisition subsystem is used to perform multi-view close-up photogrammetry of dangerous rocks through unmanned aerial vehicles, collect three-dimensional point cloud data of dangerous rocks, and upload the data to the central control subsystem; the central control subsystem is used to perform point cloud processing and structural surface clustering processing according to the collected three-dimensional point cloud data of dangerous rocks, judge the type of dangerous rocks and whether there is potential damage to the dangerous rocks according to the processing results of the point cloud processing and the structural surface clustering processing, and calculate the stability coefficients of different dangerous rock types when there is potential damage, and obtain the stability analysis results of the dangerous rocks according to the stability coefficients of the different dangerous rock types.

[0007] As a preferred solution, the central control subsystem includes a point cloud processing module, a structural surface processing module, a type judgment module and a stability analysis module;

[0008] The point cloud processing module is used to process the dangerous rock three-dimensional point cloud data collected by the survey and acquisition subsystem based on the improved KNN algorithm, so as to adaptively determine the number of field points of each target point, realize the plane fitting and normal vector calculation of the dangerous rock point cloud set, and construct a dangerous rock three-dimensional point cloud model;

[0009] The structural surface processing module is used to improve the K-means algorithm according to the genetic simulated annealing algorithm, construct the GASAK algorithm to perform cluster analysis on the point cloud set plane processed by the point cloud processing module, identify the dominant grouping of the structural surface point cloud and the structural surface parameters of each group, and the structural surface parameters include structural surface occurrence information;

[0010] The type judgment module is used to identify the dangerous rock structure surface characteristics according to the dangerous rock three-dimensional point cloud model and the structural surface parameters, and obtain the dangerous rock type judgment result according to the dangerous rock structure surface characteristics;

[0011] The stability analysis module is used to perform qualitative analysis on dangerous rocks based on the dangerous rock type judgment result obtained by the type judgment module, based on the stereographic projection method, combined with the structural surface parameters obtained by the structural surface processing module, to determine whether the dangerous rocks have potential damage. When it is determined that the dangerous rocks have potential damage, the stability coefficients of different dangerous rock types are calculated to quantitatively evaluate the stability of the dangerous rocks.

[0012] As a preferred solution, the specific processing flow of the point cloud processing module is as follows:

[0013] S101, input the three-dimensional point cloud data of the dangerous rock, and determine the maximum value D of the field point number D according to the density distribution of the three-dimensional point cloud data max and minimum value D min , based on the improved KNN algorithm, the number of domain points D of the target point in the point cloud space is adaptively determined;

[0014] S102: for any target point p in the dangerous rock three-dimensional point cloud data space i =(x i ,y i ,z i ), according to the determined number of points in the area D, obtain D-1 nearest points in the area of ​​the target point, and form a point cloud set {P} with the target point, and calculate the center point p′=(x′, y′, z′) of the point cloud set {P};

[0015] S103, a matrix formed by the three-dimensional coordinates of the target point area point cloud in the point cloud set {P} And the matrix composed of the three-dimensional coordinates of the center point of the target point area in the point cloud set {P} Constructing the covariance matrix T is the transposition symbol;

[0016] S104, performing eigenvalue decomposition on the covariance matrix to obtain eigenvalues ​​λ=(λ1,λ2,λ3) and corresponding eigenvectors μ=(μ1,μ2,μ3); the relational expression of eigenvalue decomposition satisfies λμ=Mμ, λ1,λ2,λ3 and μ1,μ2,μ3 respectively correspond to the eigenvalues ​​and eigenvectors in the three coordinate dimensions of the point cloud representing the target point area;

[0017] S105, selecting the eigenvector corresponding to the minimum eigenvalue as the normal vector of the plane of the point cloud set {P}, and using the normal vector to establish the plane equation of each point in the point cloud set {P};

[0018] S106, based on the plane equation of each point, the least square method is used to fit and adjust the equation parameters so that the sum of the squares of the vertical distances from all points in the point cloud set {P} to the plane is minimized, thereby determining the target point p i =(x i ,y i ,z i ) on the plane;

[0019] S107. Through the processing of steps S102 to S106, normal vector calculation and plane fitting are performed on each target point in the dangerous rock three-dimensional point cloud data respectively, and the plane where each target point is located is determined, so as to construct a dangerous rock three-dimensional point cloud model.

[0020] As a preferred solution, in step S106, the least square method is used to fit and adjust the equation parameters so that the objective function of minimizing the sum of the squares of the vertical distances from all points in the point cloud set {P} to the plane is:

[0021]

[0022] d n =|ax n +by n +czn +d|;

[0023] Where D represents the number of domain points contained in the point cloud set {P}; d e is the sum of the squares of the vertical distances from all target points in the point cloud to the plane, d n is any nth point p in the point cloud set {P} n =(x n ,y n ,z n ) to the plane, n = 1, 2, …, D; a, b, c are the components of the unit normal vector corresponding to the plane in the directions of the x, y, and z coordinate axes; d represents the distance from the origin to the plane.

[0024] As a preferred solution, the plane fitting of the dangerous rock point cloud set also includes: introducing characteristic index entropy, calculating the linear index, surface index and dispersion index through characteristic values ​​to quantify the spatial distribution characteristics of the point cloud and identify the plane area in the point cloud; the equation of the characteristic index entropy is:

[0025] E=L λ ln(L λ )+P λ ln(P λ )+S λ ln(S λ );

[0026] L λ =(λ1+λ2) / λ1;

[0027] P λ =(λ2-λ3) / λ1;

[0028] S λ =λ3 / λ1;

[0029] Where E is the characteristic exponential entropy, L λ is the linear index, P λ is the surface index, S λ is the dispersion index, S λ +P λ +L λ =1.

[0030] As a preferred solution, the specific processing flow of the structural surface processing module is as follows:

[0031] S201, obtain the objective function of structural surface clustering:

[0032] Where J is the objective function, that is, the sum of the distances from each structural surface to the cluster center, d 2 (N i ,K l) represents the structural surface Ni to the cluster center K l The distance, N i is the i-th structural surface, i = 1, 2, 3…, h, h represents the number of structural surfaces of the l-th cluster set; K l is the cluster center of the lth cluster set, l = 1, 2, 3…, L, L is the number of cluster sets;

[0033] S202, using genetic simulated annealing algorithm to improve K-means algorithm to minimize the objective function and complete the clustering analysis of the point cloud plane.

[0034] As a preferred solution, in step S202, the step of using a genetic simulated annealing algorithm to improve a K-means algorithm to minimize an objective function based on the similarity measure between the structural planes to complete the cluster analysis of the point cloud plane includes:

[0035] S2021. Initialize the genetic algorithm parameters, simulated annealing algorithm parameters and K-means clustering algorithm parameters of the GASAK algorithm, set the maximum number of iterations Maxgen, the initial temperature T0, the temperature cooling coefficient t, and the termination stable temperature T end ;The value range of temperature cooling coefficient t is (0.5,1);

[0036] S2022. Randomly generate an initial cluster center matrix, set the fitness function of the GASAK algorithm and the membership function of each cluster center;

[0037] S2023, set the initial loop count variable gen=0;

[0038] S2024. The clustering set of the structure surface point cloud grouping is used as the individuals of the population, the population is randomly selected, and the membership u of each individual is calculated according to the membership function and the comfort function. il and individual fitness f i ;

[0039] S2025, after performing a crossover operation on the randomly selected population, a mutation operation is performed, and the newly generated individuals are recalculated according to the fitness function and the membership function to obtain a new membership u i ' l and the new fitness f i ′;

[0040] S2026. Determine whether the new fitness is greater than the old individual fitness. If it is satisfied, the newly generated individual replaces the old individual. If it is not satisfied, the following probability formula of the simulated annealing algorithm is used to receive the newly generated cluster center and replace the old individual:

[0041]

[0042] Where P is the probability of receiving the newly generated clustering center, e is the base of the natural logarithm, and f i ' is the fitness of the newly generated individual after the current annealing iteration update of the annealing algorithm, and f i is the fitness of the old individual before the current annealing iteration update of the annealing algorithm. K is the Boltzmann constant, and T gen is the annealing temperature of the current annealing iteration of the annealing algorithm;

[0043] S2027. Determine whether gen < Maxgen. If so, increment gen by 1 and return to step S2024 to update the individuals in the population; otherwise, perform S2028;

[0044] S2028. Determine whether T gen < T end is satisfied. If so, end the GASAK algorithm and output the optimal solution; otherwise, perform the cooling operation, update T gen = t·T gen and return to step S2023.

[0045] As an optimal solution, the membership degree u il of the individual and the individual fitness f i are respectively:

[0046]

[0047] Among them, in the formula, K l is the clustering center of the l-th clustering set as an individual, N i is the i-th structural plane in the l-th clustering set as an individual, i = 1, 2, 3…, h, where h represents the number of structural planes in the l-th clustering set, and l = 1, 2, 3…, L; L is the number of clustering sets; d 2 (N i , K l ) represents the distance from the structural plane N i to the clustering center K l ; m is the weight distribution coefficient.

[0048] As an optimal solution, in the stability analysis module, based on the stereographic projection method, combined with the structural plane parameters obtained by the structural plane processing module, the specific process of qualitatively analyzing the dangerous rock and judging whether there is potential damage to the dangerous rock is as follows:

[0049] Firstly, the three-dimensional structural surface parameters are projected onto the two-dimensional stereographic plane to obtain the stereographic projection map of the structural surface. Then, the relationship and spatial distribution of the structural surfaces are identified according to the stereographic projection map, and the failure mode area of ​​the dangerous rock is analyzed. The failure mode includes sliding failure and toppling failure. Finally, by evaluating the position of the pole of the intersection line of the structural surface on the stereographic projection map, it is judged whether the dangerous rock is located in the failure mode area. If the dangerous rock is located in the failure mode area, it is determined that the dangerous rock has potential failure.

[0050] As a preferred solution, the dangerous rock type judgment results include: compression-shear sliding type, tension-shear overturning type and combined falling type;

[0051] The stability coefficient of different dangerous rock types is calculated by the following formula:

[0052] Calculation method of stability factor of compression-shear-slip dangerous rock:

[0053]

[0054] Calculation method of stability factor of tensile-shear overturning dangerous rock:

[0055] The tipping point is on the inside:

[0056] The tipping point is outside:

[0057] Calculation method of stability factor of compound falling dangerous rock:

[0058]

[0059] Where W is the deadweight of the dangerous rock mass, P is the horizontal seismic force, β is the inclination angle of the main control structure surface, C, is the shear strength parameter, Q is the water pressure on the main control structure surface of the dangerous rock, e1 is the vertical height of the main control structure surface filled with water, a is the horizontal distance from the center of gravity of the dangerous rock mass to the exposed surface of the base, and f lk is the standard value of tensile strength of dangerous rock mass, H is the vertical height of dangerous rock mass, h1 is the vertical height of the through section of the main control structure surface, l b is the contact length between the dangerous rock mass and the base, f 0k is the standard value of the tensile strength between the dangerous rock mass and the base, P is the horizontal seismic force, and h0 is the vertical distance from the center of gravity of the dangerous rock mass to the exposed surface of the base.

[0060] Compared with the prior art, the present invention has the following technical effects:

[0061] (1) The present invention uses drone close-up photogrammetry technology to achieve dangerous rock survey, which greatly reduces the manpower and material costs of traditional surveys, improves the personal safety of surveyors, and enhances the accuracy and breadth of dangerous rock surveys. At the same time, drones can be reused for a long time, which improves the efficiency of surveys. On the basis of conventional close-up photography routes, the present invention performs multi-perspective compensation shooting on some structural surfaces, eliminates blind spots in the field of vision, and achieves multi-perspective and high-precision acquisition of slope rock mass information, thereby establishing a more accurate three-dimensional slope model.

[0062] (2) The present invention realizes intelligent identification of dangerous rocks and stability judgment by combining the improved KNN algorithm, GASAK algorithm and stereographic projection method in the central control subsystem, thus forming a complete system from dangerous rock survey, data collection, dangerous rock identification and stability analysis, realizing intelligent identification of dangerous rocks and stability judgment.

[0063] (3) The present invention uploads the three-dimensional point cloud data obtained by the UAV survey to the central control subsystem for processing. Through unmanned survey and real-time data transmission, the central control subsystem can process and obtain the stability of the dangerous rock as soon as the survey is completed, so that the protection personnel can propose corresponding prevention and control measures more timely according to the stability assessment results, thereby helping to improve the prevention and control efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0064] In order to make the purpose, technical solution and advantages of the invention clearer, the present invention will be further described in detail below with reference to the accompanying drawings, in which:

[0065] Figure 1 This is a functional structure diagram of a dangerous rock intelligent identification and stability analysis system based on non-contact measurement disclosed by the present invention;

[0066] Figure 2 This is a schematic diagram of non-contact measurement by a drone according to an embodiment of the present invention;

[0067] Figure 3 Schematic diagram of dangerous rock types in an embodiment of the present invention. DETAILED DESCRIPTION

[0068] In order to make the purpose, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. The components of the embodiments of the present invention generally described and shown in the drawings here can be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of the present invention provided in the drawings is not intended to limit the scope of the claimed invention, but only represents selected embodiments of the present invention. Based on the embodiments in the present invention, all other embodiments obtained by ordinary technicians in this field without making creative work belong to the scope of protection of the present invention.

[0069] The present invention will be further described in detail below in conjunction with the accompanying drawings.

[0070] Existing dangerous rock survey and identification mainly rely on traditional contact survey methods. It is difficult to conduct efficient and accurate measurements on peaks that are often hundreds of meters high. The identification of dangerous rock stability often relies on the experience of engineers and is prone to errors. With the rise of non-contact measurement technologies such as drones and lidar, new solutions have been provided for dangerous rock survey and identification, but there is currently no mature and systematic dangerous rock identification solution. In response to the above problems and shortcomings, the present invention obtains three-dimensional point cloud data through non-contact measurement technology, uses an improved KNN algorithm to identify dangerous rock structural surfaces, and uses the GASAK algorithm to achieve dominant structural surface grouping, extract key parameters such as structural surface trace length and spacing, and finally judge the stability of the identified dangerous rocks based on the stereographic projection method, thereby improving the accuracy of stability identification.

[0071] like Figure 1 As shown, this embodiment discloses a dangerous rock intelligent identification and stability analysis system based on non-contact measurement, including a survey and acquisition subsystem and a central control subsystem;

[0072] The survey and acquisition subsystem is used to perform multi-angle close-up photogrammetry of dangerous rocks through drones, collect three-dimensional point cloud data of dangerous rocks, and upload it to the central control subsystem;

[0073] The central control subsystem is used to perform point cloud processing and structural surface clustering processing based on the collected three-dimensional point cloud data of dangerous rocks, and judge the type of dangerous rocks and whether the dangerous rocks have potential damage based on the processing results of the point cloud processing and structural surface clustering processing, and when potential damage exists, calculate the stability coefficients of different dangerous rock types, and obtain the stability analysis results of the dangerous rocks based on the stability coefficients of the different dangerous rock types.

[0074] In the present invention, Figure 2As shown in the figure, drones are used to survey dangerous rock point cloud data, and laser radar is used to perform multi-angle close-up photogrammetry of the target dangerous rock. That is, on the basis of conventional routes, the shooting distance is kept unchanged, supplementary routes are added, and multi-angle compensation shooting is performed on local areas, thereby clearing the blind spots of the field of vision and supplementing local images, so that the collected dangerous rock three-dimensional point cloud data is more accurate and more comprehensive, avoiding the risk of manual measurement. When the drone completes data collection, it is uploaded to the central control subsystem in real time through the cloud to achieve rapid response and timely analysis of dangerous rock point cloud data, thereby obtaining the stability analysis results of dangerous rocks, so that protection personnel can propose corresponding prevention and control measures more timely based on the stability assessment results, helping to improve prevention and control efficiency.

[0075] In specific implementation, the central control subsystem includes a point cloud processing module, a structural surface processing module, a type judgment module and a stability analysis module. The central control subsystem uses the improved KNN algorithm, the GASAK algorithm and the stereographic projection method to realize the intelligent identification of dangerous rocks and the judgment of stability, thus forming a complete system from dangerous rock survey, data collection, dangerous rock identification and stability analysis, realizing the intelligent identification of dangerous rocks and the judgment of stability. Each module is introduced in detail below.

[0076] The point cloud processing module is used to process the dangerous rock three-dimensional point cloud data collected by the survey and acquisition subsystem based on the improved KNN algorithm to adaptively determine the number of field points for each target point, realize plane fitting and normal vector calculation of the dangerous rock point cloud set, and construct a dangerous rock three-dimensional point cloud model.

[0077] Specifically, the specific processing flow of the point cloud processing module is as follows:

[0078] S101, input the three-dimensional point cloud data of the dangerous rock, and determine the maximum value D of the field point number D according to the density distribution of the three-dimensional point cloud data max and minimum value D min , based on the improved KNN algorithm, the number of domain points D of the target point in the point cloud space is adaptively determined.

[0079] S102: for any target point p in the dangerous rock three-dimensional point cloud data space i =(x i ,y i ,z i ), according to the determined number of points in the area D, obtain D-1 nearest points in the area of ​​the target point, and form a point cloud set {P} with the target point, and calculate the center point p'=(x', y', z') of the point cloud set {P}. The calculation formula of the center point p' is: p n =(x n ,y n ,z n) is any nth point in the point cloud set {P}, n = 1, 2,…, D.

[0080] S103, a matrix formed by the three-dimensional coordinates of the target point area point cloud in the point cloud set {P} And the matrix composed of the three-dimensional coordinates of the center point of the target point area in the point cloud set {P} Constructing the covariance matrix T is the transposition symbol;

[0081] S104, perform eigenvalue decomposition on the covariance matrix to obtain eigenvalues ​​λ=(λ1,λ2,λ3) and corresponding eigenvectors μ=(μ1,μ2,μ3); the relationship of eigenvalue decomposition satisfies λμ=Mμ, λ1,λ2,λ3 and μ1,μ2,μ3 respectively correspond to the eigenvalues ​​and eigenvectors in the three coordinate dimensions of the point cloud representing the target point area.

[0082] S105 . Select the eigenvector corresponding to the minimum eigenvalue as the normal vector of the plane of the point cloud set {P}, and use the normal vector to establish the plane equation of each point in the point cloud set {P}.

[0083] S106, based on the plane equation of each point, the least square method is used to fit and adjust the equation parameters so that the sum of the squares of the vertical distances from all points in the point cloud set {P} to the plane is minimized, thereby determining the target point p i =(x i ,y i ,z i ) on the plane.

[0084] S107. Through the processing of steps S102 to S106, normal vector calculation and plane fitting are performed on each target point in the dangerous rock three-dimensional point cloud data respectively, and the plane where each target point is located is determined, so as to construct a dangerous rock three-dimensional point cloud model.

[0085] In order to improve the calculation accuracy, the characteristic index entropy is introduced in the plane fitting of the dangerous rock point cloud set. The linear index, surface index and dispersion index are obtained by eigenvalue calculation to quantify the spatial distribution characteristics of the point cloud and identify the plane area in the point cloud.

[0086] The equation for the characteristic exponential entropy is:

[0087] E=L λ ln(L λ )+P λ ln(P λ )+S λ ln(S λ );

[0088] L λ =(λ1+λ2) / λ1;

[0089] P λ =(λ2-λ3) / λ1;

[0090] S λ =λ3 / λ1;

[0091] Where E is the characteristic exponential entropy, L λ is the linear index, P λ is the surface index, S λ is the dispersion index, S λ +P λ +L λ =1.

[0092] The coplanarity of the point cloud in the target point area is calculated according to the following formula:

[0093]

[0094] Assume the plane equation is:

[0095] ax+by+cz+d=0;

[0096] Where: a, b, c are the components of the unit normal vector corresponding to the plane in the directions of each coordinate axis, and a 2 +b 2 +c 2 = 1. d represents the distance from the origin to the plane, d = -(ax′+by′+cz′), (x′, y′, z′) is the three-dimensional coordinate of the center point p′ of the point cloud set {P}.

[0097] In order to make the fitted plane the best plane, it is necessary to satisfy the minimum sum of the squares of the vertical distances from all points in the point cloud set {P} to the plane, that is:

[0098]

[0099] d n =|ax n +by n +cz n +d|;

[0100] Where D represents the number of domain points contained in the point cloud set {P}; d e is the sum of the squares of the vertical distances from all target points in the point cloud to the plane, d n is any nth point p in the point cloud set {P} n =(x n ,y n ,z n ) to the plane, n = 1, 2,…, D.

[0101] In specific applications, the least squares method can be used to fit and adjust the equation parameters to achieve the goal of minimizing the sum of the squares of the vertical distances from all points in the point cloud set {P} to the plane, thereby determining a target point p i =(x i ,y i ,z i ) is located. By using the processing method of steps S102 to S106 above, normal vector calculation and plane fitting can be performed on each target point in the dangerous rock three-dimensional point cloud data to determine the plane where each target point is located; then, the plane information of each target point is used to construct a dangerous rock three-dimensional point cloud model.

[0102] In the present invention, by using the improved KNN algorithm, the field point number K value can be adaptively determined according to the field spatial characteristics of each target point, thereby improving the sensitivity to the local characteristics of the point cloud data, making the plane fitting of the point cloud set more accurate. Secondly, the algorithm dynamically adjusts the K value based on the point cloud density distribution, optimizes the plane fitting process, and enhances the recognition ability of complex dangerous rock structures. At the same time, the characteristic index entropy is introduced, and the quantitative accuracy of the point cloud spatial distribution characteristics is further improved through eigenvalue analysis, which helps to more accurately identify and analyze the structural surface of dangerous rocks, thereby making the dangerous rock intelligent identification and stability analysis system more efficient and accurate when dealing with dangerous rock surveys in high-altitude steep mountainous areas.

[0103] The structural surface processing module is used to improve the K-means algorithm based on the genetic simulated annealing algorithm, construct a GASAK algorithm to perform cluster analysis on the point cloud set plane obtained by the point cloud processing module, identify the dominant grouping of the structural surface point cloud and the structural surface parameters of each group, and the structural surface parameters include structural surface occurrence information.

[0104] The GASAK algorithm is a structural surface clustering method obtained by improving the K-means algorithm through the genetic simulated annealing algorithm. According to the space vector angle formula, we know that the normal vectors n of two structural surfaces can be i 、n j The acute angle θ is:

[0105] θ=arccos|n i ·n j T |;

[0106] Then the distance between the unit normal vectors d(n i ,n j )for:

[0107]

[0108] Then the similarity measure of two structural surfaces i and j is The expression is as follows:

[0109]

[0110] For the classic K-means clustering, assume that the structural surface is divided into L groups, each group is a cluster set, K l is the cluster center of the lth cluster set as an individual, N i is the i-th structural surface in the l-th cluster as an individual, i = 1, 2, 3…, h, h represents the number of structural surfaces in the l-th cluster, l = 1, 2, 3…, L; the cluster center K l for:

[0111]

[0112] If you use u ij Indicates the structural surface N i For cluster center K l The membership degree is expressed by the objective function J, which represents the sum of the distances from each structural surface to the cluster center. 2 (N i ,K j ) represents the structural surface N i To cluster center K l The distance, then u ij and J can be expressed as:

[0113]

[0114] Where: m is the weight distribution coefficient, generally m=2.

[0115] In order to minimize the objective function J, the distance from each parameter of the structural surface to the cluster center must be as small as possible, which is achieved through the genetic annealing algorithm.

[0116] Therefore, specifically, the specific processing flow of the structural surface processing module is as follows:

[0117] S201, obtain the objective function of structural surface clustering:

[0118] Where J is the objective function, that is, the sum of the distances from each structural surface to the cluster center, d 2 (N i ,K l ) represents the structural surface N i To cluster center K l The distance, N i is the i-th structural surface, i = 1, 2, 3…, h, h represents the number of structural surfaces of the l-th cluster set; K l is the cluster center of the lth cluster set, l = 1, 2, 3…, L, L is the number of cluster sets;

[0119] S202, using genetic simulated annealing algorithm to improve K-means algorithm to minimize the objective function and complete the clustering analysis of the point cloud plane.

[0120] Specifically, in step S202, the step of using a genetic simulated annealing algorithm to improve a K-means algorithm to minimize an objective function based on the similarity measure between the structural planes to complete the cluster analysis of the point cloud plane includes:

[0121] S2021. Initialize the genetic algorithm parameters, simulated annealing algorithm parameters and K-means clustering algorithm parameters of the GASAK algorithm, set the maximum number of iterations Maxgen, the initial temperature T0, the temperature cooling coefficient t, and the termination stable temperature T end ;The value range of temperature cooling coefficient t is (0.5,1);

[0122] S2022. Randomly generate an initial cluster center matrix, set the fitness function of the GASAK algorithm and the membership function of each cluster center;

[0123] S2023, set the initial loop count variable gen=0;

[0124] S2024. The clustering set of the structure surface point cloud grouping is used as the individuals of the population, the population is randomly selected, and the membership u of each individual is calculated according to the membership function and the comfort function. il and individual fitness f i ;

[0125] S2025, after performing a crossover operation on the randomly selected population, a mutation operation is performed, and the newly generated individuals are recalculated according to the fitness function and the membership function to obtain a new membership u i ' l and the new fitness f i ′;

[0126] S2026. Determine whether the new fitness is greater than the old individual fitness. If it is satisfied, the newly generated individual replaces the old individual. If it is not satisfied, the following probability formula of the simulated annealing algorithm is used to receive the newly generated cluster center and replace the old individual:

[0127]

[0128] Where P is the probability of receiving a newly generated cluster center, e is the base of the natural logarithm, and f i ' is the fitness of the newly generated individuals after the current annealing iteration of the annealing algorithm, f i is the fitness of the old individual before the current annealing iteration of the annealing algorithm, K is the Boltzmann constant, T gen is the annealing temperature of the current annealing iteration of the annealing algorithm;

[0129] S2027. Determine whether gen < Maxgen. If so, increment gen by 1, return to step S2024, and update the individuals in the population; otherwise, proceed to S2028;

[0130] S2028. Determine whether T gen <T end is satisfied. If so, end the GASAK algorithm and output the optimal solution; otherwise, perform a cooling operation, update T gen = t·T gen , and return to step S2023.

[0131] Among them, the membership degree u il of the individual and the individual fitness f i are respectively:

[0132]

[0133] Among them, in the formula, K l is the clustering center of the l-th clustering set as an individual, N i is the i-th structural plane in the l-th clustering set as an individual, i = 1, 2, 3…, h, where h represents the number of structural planes in the l-th clustering set, l = 1, 2, 3…, L, and L is the number of clustering sets; d 2 (N i , K l ) represents the distance from the structural plane N i to the clustering center K l ; m is the weight distribution coefficient, and generally m = 2.

[0134] In the present invention, the GASAK algorithm combines the advantages of the genetic algorithm and the simulated annealing algorithm, optimizes the K-means clustering process, and effectively improves the clustering quality and accuracy. Through the local search ability of simulated annealing and the global search ability of the genetic algorithm, the GASAK algorithm can more accurately identify the dangerous rock structural planes and maintain high clustering performance even under complex and variable geological conditions. Secondly, the algorithm realizes the efficient grouping of the structural plane point clouds by calculating the included angle and similarity measure between the normal vectors of the structural planes, provides key structural plane parameters for the intelligent identification and stability analysis of dangerous rocks, and effectively avoids local optimal solutions through the iterative optimization process of the GASAK algorithm.

[0135] The type judgment module is used to identify the dangerous rock structural plane characteristics according to the dangerous rock three-dimensional point cloud model and the structural plane parameters, and obtain the dangerous rock type judgment result according to the dangerous rock structural plane characteristics.

[0136] Specifically, the dangerous rock type judgment results include: compression-shear sliding type, tension-shear overturning type and combined falling type; the judgment principle used by the type judgment module is:

[0137] The compression-shear slip type: The main control structure surface of the dangerous rock mass is generally controlled by the weak surface of the stratum with a gentle inclination angle or the unloading tension structure surface in the slope. The inclination angle of the main control structure surface is very small, usually ≤45°, and the center of gravity is located inside the main control structure surface. Under the influence of gravity and external loads, the main control structure surface deteriorates and penetrates, and gradually slips, which manifests as compression-shear instability as a whole.

[0138] The tension-shear overturning type: the main controlling structural surface is a steep (side) slope unloading tension type, the inclination angle is usually >25°, the lower end of the main controlling structural surface is relatively backward relative to the rock cavity or the free surface, under the influence of gravity and external loads, the dangerous rock mass generally rotates around the lower end of the main controlling structural surface or the intersection point of the bottom of the dangerous rock mass and the free surface, thereby causing toppling damage.

[0139] The composite fall type is controlled by two main control structural surfaces, including a nearly horizontal first main control structural surface and a second main control structural surface with an inclination angle of less than 80°. The first main control structural surface mainly suffers from tensile cracking failure, and the second main control structural surface mainly suffers from compressive shear failure. Due to the influence of tension, the cracks on the first main control structural surface gradually expand. After extending to a certain range, the dangerous rock mass begins to slide and deform along the second main control structural surface. When the deformation critical value is reached, the dangerous rock falls as a whole due to instability.

[0140] In a specific implementation, the type judgment module can recognize the dangerous rock types of the above-mentioned compression-shear sliding type, tension-shear overturning type and compound falling type by matching the classification data table with the program coding, or by using the classification network tools commonly used in the prior art for implementing data feature classification, such as the SVM (Support Vector Machine) classifier or BP (Back Propagation) neural network trained with big data in advance. Thus, the dangerous rock type judgment result is identified and determined according to the three-dimensional point cloud model of the dangerous rock and the structural surface parameters.

[0141] The stability analysis module is used to perform qualitative analysis on dangerous rocks based on the dangerous rock type judgment result obtained by the type judgment module, based on the stereographic projection method, combined with the structural surface parameters obtained by the structural surface processing module, to determine whether the dangerous rocks have potential damage. When it is determined that the dangerous rocks have potential damage, the stability coefficients of different dangerous rock types are calculated to quantitatively evaluate the stability of the dangerous rocks.

[0142] In the stability analysis module, based on the stereographic projection method, the dangerous rock is qualitatively analyzed in combination with the structural surface parameters obtained by the structural surface processing module. The specific process for judging whether the dangerous rock has potential damage is as follows: first, the three-dimensional structural surface parameters are projected onto the two-dimensional stereographic plane to obtain the stereographic projection diagram of the structural surface; then, the relationship and spatial distribution of the structural surfaces are identified according to the stereographic projection diagram, and the failure mode area of ​​the dangerous rock is analyzed and obtained, and the failure mode includes sliding failure and tipping failure; finally, by evaluating the position of the pole of the intersection line of the structural surface on the stereographic projection diagram, it is judged whether the dangerous rock is located in the failure mode area; if the dangerous rock is located in the failure mode area, it is determined that the dangerous rock has potential damage.

[0143] Specifically, the principle of the stereographic projection method for qualitative analysis of the stability of dangerous rocks is as follows: using a sphere as a projection tool, the three-dimensional structural surface is projected onto the two-dimensional stereographic plane of the sphere in the form of points and lines, and the angles and combination relationships between the structural surfaces are displayed on the plane, which is used to determine the cutting direction, sliding direction, failure mode and boundary conditions of the slope rock mass, and then determine the stability of the slope rock mass. According to the relationship between the inclination angle of a single structural surface, the slope angle, the friction angle of the slope rock mass and the inclination angle of the intersection line of each group of structural surfaces, the failure mode of the slope rock mass can be generally divided into two types: sliding failure and toppling failure. Sliding failure is divided into plane sliding and wedge sliding according to the number of structural surfaces cutting the rock mass. Plane sliding is the rock block sliding failure caused by the cutting of a group of main control structural surfaces; wedge sliding is the rock mass forming a wedge under the cutting of two groups of structural surfaces, and the wedge sliding causes the failure. Toppling failure is divided into direct toppling and curved toppling according to the failure mode and the size of the inclination angle.

[0144] First, draw the damage zone where sliding and tipping may occur on the stereographic projection plane, and then judge the stability of the dangerous rock according to whether the poles of the structural surface and the combined intersection line between the structural surfaces fall into the sliding zone or the tipping zone. When the pole of a structural surface or the intersection line of the structural surface falls into the sliding zone or the tipping zone, it indicates that the plane represented by the structural surface or the wedge represented by the intersection line of the structural surface has the potential risk of sliding damage or the possibility of tipping damage. Corresponding to the potential damage judgment of the dangerous rock, if the pole of the structural surface or the intersection line of the structural surface of the dangerous rock falls into the damage mode area, it is judged that the dangerous rock has potential damage.

[0145] In the present invention, according to the judgment result of the type judgment module, the stability coefficient of different types of dangerous rocks is calculated to quantitatively evaluate the stability of dangerous rocks, such as Figure 3 As shown in the figure, the specific calculation method of the stability coefficient of compression-shear-slip type dangerous rock is:

[0146]

[0147] Calculation method of stability factor of tensile-shear overturning dangerous rock:

[0148] The tipping point is on the inside:

[0149] The tipping point is outside:

[0150] Calculation method of stability factor of compound falling dangerous rock:

[0151]

[0152] Where W is the deadweight of the dangerous rock mass, P is the horizontal seismic force, β is the inclination angle of the main control structure surface, C, is the shear strength parameter, Q is the water pressure on the main control structure surface of the dangerous rock, Q = 0.5γ w e1 2 l q , l q is the length of the dangerous rock mass along the steep cliff or steep slope, e1 is the vertical height of the main control structure surface filled with water, a is the horizontal distance from the center of gravity of the dangerous rock mass to the exposed surface of the base, and f lk is the standard value of tensile strength of dangerous rock mass, H is the vertical height of dangerous rock mass, h1 is the vertical height of the through section of the main control structure surface, l b is the contact length between the dangerous rock mass and the base, f 0k is the standard value of the tensile strength between the dangerous rock mass and the base, P is the horizontal seismic force, and h0 is the vertical distance from the center of gravity of the dangerous rock mass to the exposed surface of the base.

[0153] In the present invention, the stability of dangerous rocks is analyzed by stereographic projection method, and the three-dimensional structural surface parameters are projected onto a two-dimensional stereographic plane, which simplifies the visualization and analysis process of complex geological structures, so that the structural characteristics and spatial relationships of dangerous rocks can be intuitively understood; secondly, the stereographic projection method effectively identifies potential sliding and dumping damage areas by showing the angles and combination relationships between structural surfaces, and provides a qualitative basis for judging the stability of dangerous rocks; at the same time, in the present invention, combined with quantitative analysis, the stability coefficient of dangerous rocks is calculated, which further improves the accuracy of the evaluation.

[0154] In summary, the present invention adopts unmanned aerial vehicle to survey dangerous rock point cloud data, uses laser radar to perform multi-perspective close-up photogrammetry of target dangerous rocks, and performs multi-angle compensation shooting of local areas, so as to eliminate blind spots in the field of vision and supplement local images to make the collected three-dimensional point cloud data of dangerous rocks more accurate and comprehensive, avoiding the risk of manual measurement; using the improved KNN algorithm, the field point number K value can be adaptively determined according to the field spatial characteristics of each target point, thereby improving the sensitivity to the local characteristics of the point cloud data and making the plane fitting of the point cloud set more accurate; using the GASAK algorithm to combine the advantages of the genetic algorithm and the simulated return algorithm, the K-means clustering process is optimized, and the clustering quality and accuracy are effectively improved; the stability of dangerous rocks is analyzed by the stereographic projection method, and the three-dimensional structural surface parameters are projected onto the two-dimensional stereographic plane, which simplifies the visualization and analysis process of complex geological structures, so that the structural characteristics and spatial relationships of dangerous rocks can be intuitively understood, and potential sliding and dumping damage areas can be effectively identified, providing a qualitative basis for judging the stability of dangerous rocks.

[0155] Through the dangerous rock intelligent identification and stability analysis system based on non-contact measurement disclosed by the present invention, comprehensive and accurate surveys of dangerous rocks in the western region can be carried out, and the operation efficiency is improved and the operation cost is saved through unmanned and intelligent operations. Based on the dangerous rock three-dimensional point cloud model and the preset algorithm program, accurate dangerous rock body geometric information and structural surface parameter information can be obtained, and the dangerous rock type can be judged, and the stability of the dangerous rock body can be qualitatively analyzed and quantitatively calculated, forming a full process plan for dangerous rock survey, parameter extraction, stability judgment, and disaster prevention in the western mountainous areas.

[0156] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention rather than to limit it. Although the present invention has been described with reference to the preferred embodiments of the present invention, it should be understood by those skilled in the art that various changes may be made in form and details without departing from the spirit and scope of the present invention as defined in the appended claims.

Claims

1. A dangerous rock intelligent identification and stability analysis system based on non-contact measurement, characterized in that: Including survey and acquisition subsystem and central control subsystem; The survey and acquisition subsystem is used to perform multi-angle close-up photogrammetry of dangerous rocks through drones, collect three-dimensional point cloud data of dangerous rocks, and upload it to the central control subsystem; The central control subsystem is used to perform point cloud processing and structural surface clustering processing based on the collected three-dimensional point cloud data of dangerous rocks, and judge the type of dangerous rocks and whether the dangerous rocks have potential damage based on the processing results of the point cloud processing and structural surface clustering processing, and when potential damage exists, calculate the stability coefficients of different dangerous rock types, and obtain the stability analysis results of the dangerous rocks based on the stability coefficients of the different dangerous rock types.

2. The dangerous rock intelligent identification and stability analysis system based on non-contact measurement according to claim 1 is characterized in that: The central control subsystem includes a point cloud processing module, a structural surface processing module, a type judgment module and a stability analysis module; The point cloud processing module is used to process the dangerous rock three-dimensional point cloud data collected by the survey and acquisition subsystem based on the improved KNN algorithm, so as to adaptively determine the number of field points of each target point, realize the plane fitting and normal vector calculation of the dangerous rock point cloud set, and construct a dangerous rock three-dimensional point cloud model; The structural surface processing module is used to improve the K-means algorithm according to the genetic simulated annealing algorithm, construct the GASAK algorithm to perform cluster analysis on the point cloud set plane processed by the point cloud processing module, identify the dominant grouping of the structural surface point cloud and the structural surface parameters of each group, and the structural surface parameters include structural surface occurrence information; The type judgment module is used to identify the dangerous rock structure surface characteristics according to the dangerous rock three-dimensional point cloud model and the structural surface parameters, and obtain the dangerous rock type judgment result according to the dangerous rock structure surface characteristics; The stability analysis module is used to perform qualitative analysis on dangerous rocks based on the dangerous rock type judgment result obtained by the type judgment module, based on the stereographic projection method, combined with the structural surface parameters obtained by the structural surface processing module, to determine whether the dangerous rocks have potential damage. When it is determined that the dangerous rocks have potential damage, the stability coefficients of different dangerous rock types are calculated to quantitatively evaluate the stability of the dangerous rocks.

3. The dangerous rock intelligent identification and stability analysis system based on non-contact measurement according to claim 2 is characterized in that: The specific processing flow of the point cloud processing module is as follows: S101, input the three-dimensional point cloud data of the dangerous rock, and determine the maximum value D of the field point number D according to the density distribution of the three-dimensional point cloud data max and minimum value D min , based on the improved KNN algorithm, the number of domain points D of the target point in the point cloud space is adaptively determined; S102: for any target point p in the dangerous rock three-dimensional point cloud data space i =(x i ,y i ,z i ), according to the determined number of points in the area D, obtain D-1 nearest points in the area of ​​the target point, and form a point cloud set {P} with the target point, and calculate the center point p′=(x′,y′,z′) of the point cloud set {P}; S103, a matrix formed by the three-dimensional coordinates of the target point area point cloud in the point cloud set {P} And the matrix composed of the three-dimensional coordinates of the center point of the target point area in the point cloud set {P} Constructing the covariance matrix T is the transposition symbol; S104, performing eigenvalue decomposition on the covariance matrix to obtain eigenvalues ​​λ=(λ1,λ2,λ3) and corresponding eigenvectors μ=(μ1,μ2,μ3); the relational expression of eigenvalue decomposition satisfies λμ=Mμ, λ1,λ2,λ3 and μ1,μ2,μ3 respectively correspond to the eigenvalues ​​and eigenvectors in the three coordinate dimensions of the point cloud representing the target point area; S105, selecting the eigenvector corresponding to the minimum eigenvalue as the normal vector of the plane of the point cloud set {P}, and using the normal vector to establish the plane equation of each point in the point cloud set {P}; S106, based on the plane equation of each point, the least square method is used to fit and adjust the equation parameters so that the sum of the squares of the vertical distances from all points in the point cloud set {P} to the plane is minimized, thereby determining the target point p i =(x i ,y i ,z i ) on the plane; S107. Through the processing of steps S102 to S106, normal vector calculation and plane fitting are performed on each target point in the dangerous rock three-dimensional point cloud data respectively, and the plane where each target point is located is determined, so as to construct a dangerous rock three-dimensional point cloud model.

4. The dangerous rock intelligent identification and stability analysis system based on non-contact measurement according to claim 3 is characterized in that: In step S106, the least square method is used to fit and adjust the equation parameters so that the objective function of minimizing the sum of the squares of the vertical distances from all points in the point cloud set {P} to the plane is: d n =|ax n +by n +cz n +d|; Where D represents the number of domain points contained in the point cloud set {P}; d e is the sum of the squares of the vertical distances from all target points in the point cloud to the plane, d n is any nth point p in the point cloud set {P} n =(x n ,y n ,z n ) to the plane, n = 1, 2, …, D; a, b, c are the components of the unit normal vector corresponding to the plane in the directions of the x, y, and z coordinate axes; d represents the distance from the origin to the plane.

5. The dangerous rock intelligent identification and stability analysis system based on non-contact measurement according to claim 3 is characterized in that: The plane fitting of the rockfall point clouds further includes: introducing the characteristic index entropy, calculating the linear index, planar index, and dispersion index through eigenvalues to quantify the spatial distribution characteristics of the point clouds, and identifying the planar regions in the point clouds; the equation of the characteristic index entropy is: E=L λ ln(L λ )+P λ ln(P λ )+S λ ln(S λ ); L λ =(λ1+λ2) / λ1; P λ =(λ2-λ3) / λ1; S λ =λ3 / λ1; Where E is the characteristic exponential entropy, L λ is the linear index, P λ is the surface index, S λ is the dispersion index, S λ +P λ +L λ =1.

6. The dangerous rock intelligent identification and stability analysis system based on non-contact measurement according to claim 2 is characterized in that: The specific processing flow of the structural plane processing module is as follows: S201, obtain the objective function of structural surface clustering: Where J is the objective function, that is, the sum of the distances from each structural surface to the cluster center, d 2 (N i ,K l ) represents the structural surface Ni to the cluster center K l The distance, N i is the i-th structural surface, i = 1, 2, 3…, h, h represents the number of structural surfaces of the l-th cluster set; K l is the cluster center of the lth cluster set, l = 1, 2, 3…, L, L is the number of cluster sets; S202. Improve the K-means algorithm using the genetic simulated annealing algorithm to minimize the objective function and complete the clustering analysis of the plane of the point cloud set.

7. The dangerous rock intelligent identification and stability analysis system based on non-contact measurement according to claim 6 is characterized in that: In step S202, the step of improving the K-means algorithm using the genetic simulated annealing algorithm to minimize the objective function and complete the clustering analysis of the plane of the point cloud set according to the similarity measure between the structural planes includes: S2021. Initialize the genetic algorithm parameters, simulated annealing algorithm parameters and K-means clustering algorithm parameters of the GASAK algorithm, set the maximum number of iterations Maxgen, the initial temperature T0, the temperature cooling coefficient t, and the termination stable temperature T end ;The value range of temperature cooling coefficient t is (0.5,1); S2022. Randomly generate an initial clustering center matrix, and set the fitness function of the GASAK algorithm and the membership function of each clustering center. S2023. Set the initial loop count variable gen = 0. S2024. The clustering set of the structure surface point cloud grouping is used as the individuals of the population, the population is randomly selected, and the membership u of each individual is calculated according to the membership function and the comfort function. il and individual fitness f i ; S2025, after performing a crossover operation on the randomly selected population, a mutation operation is performed, and the newly generated individuals are recalculated according to the fitness function and the membership function to obtain a new membership u i ' l and the new fitness f i ′; S2026. Judge whether the new fitness is greater than the old individual fitness. When it is satisfied, the newly generated individual replaces the old individual; when it is not satisfied, the following probability formula of the simulated annealing algorithm is used to receive the newly generated clustering center and replace the old individual: Where P is the probability of receiving a newly generated cluster center, e is the base of the natural logarithm, and f i ' is the fitness of the newly generated individuals after the current annealing iteration of the annealing algorithm, f i is the fitness of the old individual before the current annealing iteration of the annealing algorithm, K is the Boltzmann constant, T gen is the annealing temperature of the current annealing iteration of the annealing algorithm; S2027. Judge whether gen < Maxgen. If so, increment gen by 1, return to step S2024, and update the individuals in the population; otherwise, perform S2028. S2028, determine whether T is satisfied gen <T end If yes, then the GASAK algorithm ends and the optimal solution is output; otherwise, the cooling operation is performed and T is updated. gen =t·T gen , return to step S2023.

8. The dangerous rock intelligent identification and stability analysis system based on non-contact measurement according to claim 7 is characterized in that: The membership degree u of the individual il and individual fitness f i They are: Among them, K l is the cluster center of the lth cluster set as an individual, N i is the i-th structural surface in the l-th cluster set as an individual, i = 1, 2, 3…, h, h represents the number of structural surfaces in the l-th cluster set, l = 1, 2, 3…, L; L is the number of cluster sets; d 2 (N i ,K l ) represents the distance from the structural surface Ni to the cluster center Kl; m is the weight distribution coefficient.

9. The dangerous rock intelligent identification and stability analysis system based on non-contact measurement according to claim 2 is characterized in that: In the stability analysis module, based on the stereographic projection method, combined with the structural plane parameters obtained by the structural plane processing module, the specific process of qualitatively analyzing the rockfall and judging whether there is potential damage to the rockfall is as follows: First, project the three-dimensional structural plane parameters onto the two-dimensional stereographic plane to obtain the stereographic projection map of the structural plane. Then, according to the stereographic projection map, identify the mutual relationship and spatial distribution of the structural planes, and analyze to obtain the failure mode area of the rockfall. The failure modes include sliding failure and toppling failure. Finally, by evaluating the position of the pole of the intersection line of the structural planes on the stereographic projection map, judge whether the rockfall is located in the failure mode area. If the rockfall is located in the failure mode area, it is determined that the rockfall has potential damage.

10. The dangerous rock intelligent identification and stability analysis system based on non-contact measurement according to claim 1 is characterized in that: The judgment results of the rockfall types include: compression-shear sliding type, tension-shear overturning type, and composite falling type; The calculation of the stability coefficients of different rockfall types is calculated by the following formula: Calculation method for the stability coefficient of compression-shear sliding type rockfall: Calculation method for the stability coefficient of tension-shear overturning type rockfall: The tipping point is on the inside: The tipping point is outside: Calculation method for the stability coefficient of composite falling type rockfall: Where W is the deadweight of the dangerous rock mass, P is the horizontal seismic force, β is the inclination angle of the main control structure surface, C, is the shear strength parameter, Q is the water pressure on the main control structure surface of the dangerous rock, e1 is the vertical height of the main control structure surface filled with water, a is the horizontal distance from the center of gravity of the dangerous rock mass to the exposed surface of the base, and f lk is the standard value of tensile strength of dangerous rock mass, H is the vertical height of dangerous rock mass, h1 is the vertical height of the through section of the main control structure surface, l b is the contact length between the dangerous rock mass and the base, f 0k is the standard value of the tensile strength between the dangerous rock mass and the base, P is the horizontal seismic force, and h0 is the vertical distance from the center of gravity of the dangerous rock mass to the exposed surface of the base.

Images