Rock mass informatization model establishment method based on laser point cloud

Abstract

A rock mass informatization model establishment method based on laser point cloud relates to the field of model identification and analysis, and is characterized by comprising the following steps: unmanned aerial vehicle laser data acquisition; establishing an inclination model; re-matching the laser point cloud flight strip; carrying out refined point cloud classification; establishing a triangulation network model; establishing a rock form model; establishing a rock mass information model; establishing a rock mass analysis system; and collapse simulation. The method provided by the embodiment of the invention has the advantages that the unmanned aerial vehicle can be used for carrying the laser radar to carry out close-to-land imitation flight, structures such as collapse, cracks and rockfall of the mountain under the dense forest can be directly measured by using the good vegetation penetrating ability of the laser point cloud, and compared with the original modes such as field investigation and photogrammetry, the method has the advantages that the measurement accuracy is greatly improved. The information obtained by the laser point cloud is more accurate and comprehensive, and the geometric dimension, the steep cliff section shape, the structural surface feature, the lithology, the boundary condition and other features of the dangerous rock mass can be directly confirmed.

Description

Technical Field

[0001] The present invention belongs to the field of model recognition and analysis, and more specifically, relates to a method for establishing a rock mass information model based on laser point cloud based on unmanned aerial vehicle laser radar technology. Background Art

[0002] In recent years, my country's infrastructure construction such as roads and railways has developed rapidly. With the increasing intensity of engineering construction activities, engineering geological problems such as slope stability and dangerous rock collapse have become increasingly prominent. With the continuous advancement of engineering technology, the complexity of the project has also increased. In complex and dangerous geological conditions, traditional survey methods are difficult to guarantee survey accuracy and have certain risks. Most of the dangerous rockfall areas along the highway are characterized by steep mountains, complex geological environments, difficult survey work, and high danger. The investigation and analysis of dangerous rock masses mainly focuses on confirming the significant characteristics of dangerous rock masses such as geometric dimensions, steep cliff section shapes, structural surface characteristics, lithology, and boundary conditions. In order to accurately restore the rock mass morphology, it is necessary to establish a real three-dimensional model to objectively analyze and study the rock mass.

[0003] With the deepening of research on concepts such as BIM, CIM, twin cities, and the metaverse, the technology of building various information models related to architecture has developed rapidly. Nowadays, based on advanced technologies such as oblique photography and lidar, city-level roads, bridges, houses, trees, pipeline facilities, underground spaces and other facilities can be finely modeled with centimeter-level resolution, and the model can be given relevant attribute information for further calculation and analysis. However, there are few technical methods for fine modeling of dangerous rock masses on steep slopes, and it is difficult to study the mesoscopic level of dangerous rock masses on steep slopes, such as the volume of dangerous rock masses and the occurrence of structural surfaces. In the traditional investigation and analysis of high-steep dangerous rock masses, due to the subjective judgment of the surveyors on the scope of dangerous rock masses, the morphology and main control structural surface, the connectivity rate and the degree of opening, the fundamental reason is that the dangerous rock masses are far away, the internal invisible and the influence of the surrounding environment make it difficult to judge the dangerous rock masses. Analogous to BIM, there is an urgent need for a rock mass information model. This model needs to have a complete rock mass engineering information database that is consistent with the actual situation. At the same time, based on the rock mass model, a series of model analyses can be carried out to more accurately predict rock mass changes.

[0004] Around 2020, the technology of using large-load UAVs equipped with high-frequency, long-range laser radars for ground-flight measurement will mature, so we can directly perform aerial laser scanning on dangerous rock masses, obtain the ground conditions under the dense forest, and measure the width and depth of rock cracks. In this way, we can truly obtain the real surface three-dimensional point cloud data of dangerous rock masses on steep slopes. Using high-density point clouds, we can establish an information model of the rock mass. Since there are few relevant literatures on this definition, this article temporarily translates the rock information model into English: Rock Information Modeling, abbreviated as RIM. At present, there is still a lack of a method that uses drones equipped with laser radar to conduct close-up terrain simulation flights. By using the good vegetation penetration ability of laser point clouds, the collapse, cracks, falling rocks and other structures of the mountain under the dense forest can be directly measured. Compared with the original field surveys and photogrammetry, the information obtained by laser point clouds is more accurate and comprehensive, and can directly confirm the geometric dimensions, cliff section shapes, structural surface characteristics, lithology and boundary conditions of dangerous rock masses. Lidar collects high-density laser point cloud data to construct a refined rock mass information model, which can simplify complex field investigation work to indoor model analysis. Compared with manual experience judgment, the simulation model motion simulation method can effectively analyze and predict the collapse and landslide of dangerous rock masses, avoid disasters, and reduce the impact of geological disasters. Summary of the invention

[0005] In order to overcome the drawbacks of existing technical solutions, the use of unmanned aerial vehicles equipped with laser radar for close-up terrain simulation flight is realized. The good vegetation penetration ability of laser point cloud can be used to directly measure the collapse, cracks, falling rocks and other structures of the mountain under the dense forest. Compared with the original field survey and photogrammetry, the information obtained by laser point cloud is more accurate and comprehensive, and can directly confirm the geometric size, cliff section shape, structural surface characteristics, lithology and boundary conditions of the dangerous rock mass. The laser radar collects high-density laser point cloud data to construct a refined rock mass information model, which can simplify complex field investigation work to indoor model analysis. Compared with manual experience judgment, the simulation model motion simulation method can effectively analyze and predict the collapse and landslide of the dangerous rock mass, avoid disasters, and reduce the impact of geological disasters.

[0006] A method for establishing a rock mass information model based on laser point cloud, wherein the method comprises the following steps: UAV laser data acquisition: The UAV is equipped with a high-frequency, long-range laser radar, and performs a ground-flight simulation in the rock mass area, and performs a "well"-shaped flight acquisition technology to collect original point clouds, and separate the ground points under the forest by using the laser's good dense forest penetration ability; Inclined model establishment: The UAV is equipped with a high-definition camera to collect images of the rock mass and establish an inclined photography model by using a close-up photogrammetry method; Laser point cloud flight belt re-matching: For different sorties, the matching of cross-flight flight belt data is relatively poor, and the surface point cloud situation of the mountain around the dangerous rock mass is checked by the relatively smooth surface point cloud of the mountain, and the point cloud is re-accurately matched by the laser emission angle and emission time information; Refined point cloud classification: Refined point cloud classification is performed to distinguish different types of point clouds such as ground points, tree points, low-lying vegetation points, rock wall points, and crack points; Establishment of angle mesh model: Generate rock triangulation model through classified laser point cloud; Establishment of rock morphology model: Combined with the tilt model, further repair the triangulation model to establish the rock morphology model; Establishment of rock information model: On the basis of the rock morphology model, add the rock structure, rock type, density, geological conditions and other related property information to form a rock information model; Establishment of rock analysis system: Add environmental and geological information to the system to form a set of rock information models that are almost consistent with the outside world, and conduct rock change research and collapse time point and disaster degree prediction through simulation methods; Collapse simulation: After setting the rock collapse, further simulate the rock movement trajectory and fragmentation degree to analyze the rock trajectory, simulation characteristics and the impact of landslide, analyze the rock collapse and the harm caused to the surrounding area, make disaster prevention and mitigation measures in advance, and form a geological disaster analysis report.

[0007] A method for establishing a rock mass information model based on laser point cloud, wherein the specific steps of the drone laser data collection are:

[0008] On-site survey: Use a small drone to observe the mountain, analyze the on-site situation, determine the location of suspected dangerous rock mass, cracks and other unfavorable geological bodies, and select a flat, clean and hard site with an area of ​​not less than 5 meters by 5 meters as the drone landing site without any obstructions such as pipe corridors and wires above, and without communication interference during the remote control ground test;

[0009] Route design: Route is laid out along the main rock surface and crack direction, and routes are laid out crosswise in important rock mass areas, and the routes are laid out in a "well" shape;

[0010] Laser equipment parameter setting: To ensure the quality of collected data, point cloud density, overlap, flight altitude, and laser radar are set respectively. The appropriate laser emission frequency, scanning frequency, scanning angle and other parameters are set to ensure that the original point cloud density is greater than 1200 points / ㎡; the flight zone overlap is greater than 50%, and the routes are crossed to ensure that each area has point cloud coverage in the four directions of "front, back, left, and right", that is, each area should have at least 4 degrees of overlap; when setting the flight altitude, it is determined according to the performance of different laser radar equipment and terrain to ensure that the point cloud data is uniform, and the terrain-simulating flight mode is adopted, that is, the relative height to the ground is consistent;

[0011] The flight altitude calculation formula is:

[0012]

[0013] Where:

[0014] H represents the maximum flight altitude, and L represents the maximum measurement range of the designed laser emission frequency under the actual target reflectivity; in the design of the laser radar data acquisition parameters, the parameter emission frequency is set to 600kHz and the scanning rate is set to 200 revolutions. The UAV speed is 8m / s, the flight zone overlap is set to 50%, and the relative altitude is 200M. Laser data acquisition is carried out according to these parameters.

[0015] A method for establishing a rock mass information model based on laser point cloud, wherein the specific steps of establishing the tilt model are:

[0016] Site survey: Choose a flat, clean, hard site with an area of ​​no less than 5 meters by 5 meters, with no obstructions such as pipe corridors and wires above, and no communication interference during the remote control ground test as the drone take-off and landing site;

[0017] Data collection: Close-up photogrammetry, 80% heading overlap, 70% heading, circumferential flight for key areas, and the coordinate system of the image control points is consistent with the point cloud;

[0018] Tilt model establishment: Establish a tilt model to assist in modifying the rock morphology model.

[0019] A method for establishing a rock mass information model based on laser point cloud, wherein the specific steps of re-matching the laser point cloud flight path are:

[0020] Strip adjustment correction: single route adjustment method is used, P and Q are defined as the heading angle, pitch angle, and roll angle parameters of adjacent laser point cloud strips, and the correction equations for heading angle, pitch angle, and roll angle are established in the overlapping area between routes:

[0021]

[0022] Among them, d(X, Y, Z) represents the correction of the point cloud data with the same name in different strips in the X, Y, and Z directions in the geographic space coordinates, Tx, Ty, and Tz represent the difference of the point cloud data in the X, Y, and Z directions in the geographic space coordinates, and ah, ar, and ap represent the heading angle, pitch angle, and roll angle correction of the point cloud strip corresponding to the track line;

[0023] Manual correction and matching: The point cloud is displayed in the point cloud software with the flight strip as the mark through manual inspection and correction. The matching plane and elevation errors between the flight strips are analyzed block by block in the form of cross-section, and the matching problem areas are checked for further alignment; based on one of the through flight strips, the remaining flight strips are translated to the reference flight strip for strict alignment. If there are uneven errors between the flight strips, manual flight strip matching is required in different areas. The laser point cloud has temporal continuity. When its characteristics are difficult to distinguish due to spatial factors, the laser time and other information are referenced. After the relative position matching of the flight strip is completed, the image control points of the inclined model are used to align the point cloud data to the inclined model position; the data after manual flight strip alignment is repeatedly used to identify the problem area. If there is a problem, manual flight strip alignment is continued. If the data is qualified, the next step of point cloud classification is carried out.

[0024] A method for establishing a rock mass information model based on laser point cloud, wherein the specific steps of the refined point cloud classification are:

[0025] Automatic point cloud classification by software: starting from the lowest point of the point cloud as the ground point, the angle and jump height of the adjacent laser point and its connecting line are calculated to separate the ground point data;

[0026] Manual inspection and classification: Build a DEM model based on ground points, remove clutter, repair uncalculated ground points, and subdivide ground points.

[0027] A method for establishing a rock mass information model based on laser point cloud, wherein the specific steps of establishing the triangulated network model are:

[0028] Rock mass point cloud preprocessing: further classify the ground points according to different areas of the rock mass. For the smoother rock mass top surface, appropriate thinning can be done. For the complex structure, crack area and some areas with fewer laser points, appropriate point cloud data can be added according to the tilt model.

[0029] Constructing triangulated network model: constructing triangulated network for ground laser points, maintaining uniform change of adjacent triangulated network surface normals, pointing to the outside of the rock mass. The main steps involve obtaining point cloud normals, separating rock blocks according to normal changes, filtering and simplifying point clouds, constructing networks inside different rock blocks, constructing networks between rock blocks, and simplifying meshes;

[0030] Auxiliary network construction method: According to the position information and time information of the point cloud, the line segments of the adjacent time laser points of the single scan line are connected;

[0031] The triangulated network model was established: the final triangulated network model was constructed by basically matching the rock morphology in nature with the tilt model, with a continuous and smooth rock surface, continuous cracks, no adhesion between the rock walls on both sides of the cracks, and no vegetation or clutter features.

[0032] The specific steps of constructing the triangulated network model are:

[0033] Point cloud normal calculation: The calculation process is given a point p and its adjacent points {pi|i=1,…,Np}, the normal at p is np, and the surface is fitted on the adjacent points and the normal is calculated from the fitted surface. The n-jet surface model is used. The surface is represented by a polynomial function Jn: that is, R2→R, mapping the coordinates of point P (x, y) to their height z in the tangent space:

[0034]

[0035] To simplify notation, β is represented as a vector:

[0036]

[0037] At the same time, define

[0038]

[0039] At this time, the normal vector of the fitted surface is expressed as:

[0040]

[0041] Where (x, y) is the coordinate value of point P, the adjacent point of point P is Pi, z represents the coordinate (x, y) mapped to their height in the tangent space, and the Jn surface uses a polynomial function. is the vector coefficient, np is the normal vector of the fitted surface,

[0042] When constructing the mesh of the laser point cloud, the normal vector np of the fitting surface should be ensured to change gradually and continuously, and the points should be prioritized for meshing. The point cloud should be grouped and meshed according to the four cases where np is equal to 0 (360), 90, 180, and 270.

[0043] Connecting laser points to form a network: It mainly constructs the point cloud into a triangulated network. The principle of network construction is to take a triangle ABC from the triangle set, calculate the circumscribed circle of the triangle, then insert a point P1, find all triangles whose circumscribed circles contain this point, delete these triangles to form a cavity, and finally connect the laser points into a triangulated network;

[0044]

[0045] In the formula: where xyz corresponds to the coordinate values ​​of the interpolation point P and the triangle vertex ABC respectively, r represents the radius of the circumscribed circle of point P in the triangle. The specific vertices are P(x0,y0)A(x_{0},y_{0}), A(x1,y1)A(x_{1},y_{1}), B(x2,y2)B(x_{2},y_{2}), C(x3,y3)C(x_{3},y_{3}).

[0046] A method for establishing a rock mass information model based on laser point cloud, wherein the specific steps of establishing the rock morphology model are:

[0047] Basic model processing: correct the triangles with wrong directions, weld the nodes of triangulated mesh, and delete isolated points, lines and faces;

[0048] Structural surface repair: refer to the tilt model, analyze the rock structure, delete the structural surface that is not accurately matched with the tilt model and point cloud data, and reconnect and draw the correct structural surface according to the point cloud situation;

[0049] Point cloud fit check: The classified ground point cloud and tilt model are fit checked with the established triangulated network model. If an erroneous surface is found, the surface needs to be re-trimmed.

[0050] Generalized model: According to different rock forms, the model nodes are generalized within a certain tolerance range to reduce the number of structural faces. The simplified model is rechecked for point cloud fitting and the faces with large errors are modified.

[0051] Basic morphological measurement: Analyze and cut the rock according to the model's morphology, crack orientation and other characteristics to form small pieces of dangerous rock mass. Measure the occurrence of each small piece of rock mass, simplify the model, and perform morphological measurements of the center of gravity, inclination, and occurrence information.

[0052] A method for establishing a rock mass information model based on laser point cloud, wherein the specific steps of establishing the rock mass information model are:

[0053] Collection of survey attribute information: First, the basic structure and type of the rock are determined by the internal staff, and then verified by on-site survey; rock sampling is carried out to measure the volume, weight and other information, and drilling surveys are carried out on the surrounding terrain to collect geological condition information;

[0054] Information entry: Establish a model database, set up an independent ID number for each rock, establish an attribute information table based on the rock ID, enter the survey attribute information, rock added structure, rock type, density, and geological conditions' own attribute information into the corresponding rock model information table to form a rock information model.

[0055] A method for establishing a rock mass information model based on laser point cloud, wherein the specific steps of establishing the rock mass analysis system are:

[0056] System information input: Temperature, precipitation, wind direction, forest-related environmental information, groundwater, crustal movement, and fault zone-related geological information are input into the system as influencing factors. The corresponding influencing parameters of the influencing factors are given through experimental data. The parameters mainly include the magnitude, direction, duration, triggering conditions of the force exerted by the influencing factors on the rock, and the damage caused by the influencing factors to the rock body, including rainwater immersion corrosion, wind erosion of the rock, and vegetation weathering of the rock.

[0057] Motion simulation: The system uses time as a variable, accumulates the effects of various influencing factors on rocks over time until the rocks collapse, records the collapse process, and analyzes the impact of collapse and landslides; System information input: The system inputs temperature, precipitation, wind direction, forest-related environmental information, and groundwater, crustal movement, and fault zone-related geological information as influencing factors, and gives the corresponding influencing parameters of the influencing factors through experimental data. The parameters mainly include the magnitude, direction, duration, triggering conditions of the force exerted by the influencing factors on the rocks, and the damage caused by the influencing factors to the rock body, including rainwater immersion corrosion, wind erosion of rocks, and vegetation weathering of rocks.

[0058] Motion simulation: The system uses time as a variable, and accumulates the effects of various influencing factors on the rock over time until the rock collapses, records the collapse process, and analyzes the impact of the collapse landslide.

[0059] A method for establishing a rock mass information model based on laser point cloud, wherein the specific steps of the collapse simulation are:

[0060] Rock trajectory analysis: Rockfall is considered a rigid body motion. The rock trajectory is governed by the interaction between the rock and the ground. The model contains six main state variables: three translational velocities and three rotational velocities of the rockfall, from which the kinetic energy is obtained, jump distance and jump height, rock orientation and rotation speed are included in the rock / ground interaction. The model takes into account the real rock shape and calculates the motion trajectory of each rock body on the model, including jump height, sliding speed, rotation speed, rotation and total kinetic energy. All possible motion modes of the rockfall are deterministically simulated. The dynamics of the trajectory of a single rock body are examined separately, and statistical analysis is performed on multiple trajectory sets;

[0061] Simulation features: The system simulates the true nature of rock fall motion behavior, consistent with rock shape and release kinematics of site-specific rock fall hazards;

[0062] Report analysis: Based on the analysis of rock collapse and the damage caused to the surrounding areas, disaster prevention and mitigation measures are taken in advance to form a geological disaster analysis report.

[0063] It can be seen from this that:

[0064] The method in the embodiment of the present invention can use the unmanned aerial vehicle equipped with a laser radar to perform close-up terrain simulation flight, and use the good vegetation penetration ability of the laser point cloud to directly measure the collapse, cracks, falling rocks and other structures of the mountain under the dense forest. Compared with the original field survey, photogrammetry and other methods, the information obtained by the laser point cloud is more accurate and comprehensive, and can directly confirm the geometric size, cliff section shape, structural surface characteristics, lithology and boundary conditions of the dangerous rock mass. The laser radar collects high-density laser point cloud data to construct a refined rock mass information model, which can simplify complex field investigation work to indoor model analysis. Compared with manual experience judgment, the simulation model motion simulation method can effectively analyze and predict the collapse and landslide of the dangerous rock mass, avoid disasters, and reduce the impact of geological disasters. BRIEF DESCRIPTION OF THE DRAWINGS

[0065] Figure 1 A schematic flow chart of a method for establishing a rock mass information model based on laser point cloud provided in an embodiment of the present invention;

[0066] Figure 2 It is a schematic diagram of the process flow of designing the laser acquisition parameters of the UAV in an embodiment of the present invention;

[0067] Figure 3 A schematic diagram of the steps of establishing a tilt model in an embodiment of the present invention;

[0068] Figure 4 Schematic diagram of the laser point cloud flight strip re-matching process in an embodiment of the present invention;

[0069] Figure 5 Schematic diagram of the process of fine-grained point cloud classification in an embodiment of the present invention;

[0070] Figure 6 A schematic diagram of the steps for establishing a triangulated network model in an embodiment of the present invention;

[0071] Figure 7 This is a schematic diagram of the steps of establishing a rock morphology model in an embodiment of the present invention;

[0072] Figure 8 It is a schematic diagram of the process of establishing a rock mass information model in an embodiment of the present invention;

[0073] Fig. 9 It is a schematic diagram of the steps of establishing a rock mass analysis system in an embodiment of the present invention;

[0074] Fig.10 is a schematic flow chart of collapse simulation steps in an embodiment of the present invention;

[0075] Fig.11A schematic diagram of the rock mass morphology model construction in an embodiment of the present invention;

[0076] Fig.12 This is a diagram of rock morphology results in an embodiment of the present invention.

[0077] Fig.13 This is a schematic diagram of rock morphology stress analysis in an embodiment of the present invention;

[0078] Fig.14 This is a diagram showing the effect of rock collapse in an embodiment of the present invention. DETAILED DESCRIPTION

[0079] In order to enable those skilled in the art to better understand the present invention, the present invention will be described in detail below with reference to the accompanying drawings and specific embodiments. The exemplary embodiments and descriptions of the present invention are used to explain the present invention but are not intended to limit the present invention.

[0080] Embodiment 1:

[0081] Figure 1 A schematic diagram of the process of establishing a rock mass information model based on laser point cloud provided in this embodiment is shown in FIG. Figure 1As shown, a method for establishing a rock mass information model based on laser point cloud is provided, wherein the method comprises the following steps: UAV laser data acquisition: the UAV is equipped with a high-frequency, long-range laser radar, and performs a terrain-simulating flight in the rock mass area, and performs a "well"-shaped flight acquisition technology to collect original point clouds, and uses the laser's good dense forest penetration ability to separate the ground points under the forest; tilt model establishment: the UAV is equipped with a high-definition camera to collect images of the rock mass and establish an oblique photography model by using a close-up photogrammetry method; laser point cloud flight strip re-matching: for different sorties, the matching of cross-flight flight strip data is relatively poor, and the surface point cloud situation around the dangerous rock mass is checked through the relatively smooth surface of the mountain, and the point cloud is re-accurately matched by the laser emission angle and emission time information; refined point cloud classification: refined point cloud classification is performed to distinguish different types of point clouds such as ground points, tree points, low-lying vegetation points, rock wall points, and crack points. ; Establishment of triangulated network model: Generate rock triangulated network model through classified laser point cloud; Establishment of rock morphological model: Combined with the tilt model, further repair the triangulated network model to establish the rock morphological model; Establishment of rock information model: On the basis of the rock morphological model, add the rock structure, rock type, density, geological conditions and other related property information to form a rock information model; Establishment of rock analysis system: Add environmental and geological information to the system to form a set of rock information models that are almost consistent with the outside world, and use simulation methods to study rock changes and predict collapse time and disaster severity; Collapse simulation: After setting the rock collapse, further simulate the rock's movement trajectory and fragmentation degree, analyze the rock trajectory, simulation characteristics and the impact of landslides, analyze rock collapse and the hazards caused to the surrounding areas, make disaster prevention and mitigation measures in advance, and form a geological disaster analysis report.

[0082] like Figure 2 As shown, a method for establishing a rock mass information model based on laser point cloud, wherein the specific steps of the drone laser data collection are:

[0083] On-site survey: Use a small drone to observe the mountain, analyze the on-site situation, determine the location of suspected dangerous rock mass, cracks and other unfavorable geological bodies, and select a flat, clean and hard site with an area of ​​no less than 5 meters by 5 meters as the drone landing site without any obstructions such as pipe galleries and wires above, and without communication interference from the remote control ground test;

[0084] Route design: Route is laid out along the main rock surface and crack direction, and routes are laid out crosswise in important rock mass areas, and the routes are laid out in a "well" shape;

[0085] Laser equipment parameter setting: To ensure the quality of collected data, point cloud density, overlap, flight altitude, and laser radar are set respectively. The appropriate laser emission frequency, scanning frequency, scanning angle and other parameters are set to ensure that the original point cloud density is greater than 1200 points / ㎡; the flight zone overlap is greater than 50%, and the routes are crossed to ensure that each area has point cloud coverage in the four directions of "front, back, left, and right", that is, each area should have at least 4 degrees of overlap; when setting the flight altitude, it is determined according to the performance of different laser radar equipment and terrain to ensure that the point cloud data is uniform, and the terrain-simulating flight mode is adopted, that is, the relative height to the ground is consistent;

[0086] The flight altitude calculation formula is:

[0087]

[0088] Where:

[0089] H represents the maximum flight altitude, and L represents the maximum measurement range of the designed laser emission frequency under the actual target reflectivity; in the design of the laser radar data acquisition parameters, the parameter emission frequency is set to 600kHz and the scanning rate is set to 200 revolutions. The UAV speed is 8m / s, the flight zone overlap is set to 50%, and the relative altitude is 200M. Laser data acquisition is carried out according to these parameters.

[0090] like Figure 3 As shown, a method for establishing a rock mass information model based on laser point cloud, wherein the specific steps of establishing the tilt model are:

[0091] Site survey: Choose a flat, clean, hard site with an area of ​​no less than 5 meters by 5 meters, with no obstructions such as pipe corridors and wires above, and no communication interference during the remote control ground test as the drone take-off and landing site;

[0092] Data collection: Close-up photogrammetry, 80% heading overlap, 70% heading, circumferential flight for key areas, and the coordinate system of the image control points is consistent with the point cloud;

[0093] Tilt model establishment: Establish a tilt model to assist in modifying the rock morphology model.

[0094] like Figure 4 As shown, a method for establishing a rock mass information model based on laser point cloud, wherein the specific steps of re-matching the laser point cloud flight path are:

[0095] Strip adjustment correction: single route adjustment method is used, P and Q are defined as the heading angle, pitch angle, and roll angle parameters of adjacent laser point cloud strips, and the correction equations for heading angle, pitch angle, and roll angle are established in the overlapping area between routes:

[0096]

[0097] Among them, d(X, Y, Z) represents the correction of the point cloud data with the same name in different strips in the X, Y, and Z directions in the geographic space coordinates, Tx, Ty, and Tz represent the difference of the point cloud data in the X, Y, and Z directions in the geographic space coordinates, and ah, ar, and ap represent the heading angle, pitch angle, and roll angle correction of the point cloud strip corresponding to the track line;

[0098] Manual correction and matching: The point cloud is displayed in the point cloud software with the flight strip as the mark through manual inspection and correction. The matching plane and elevation errors between the flight strips are analyzed block by block in the form of cross-section, and the matching problem areas are checked for further alignment; based on one of the through flight strips, the remaining flight strips are translated to the reference flight strip for strict alignment. If there are uneven errors between the flight strips, manual flight strip matching is required in different areas. The laser point cloud has temporal continuity. When its characteristics are difficult to distinguish due to spatial factors, the laser time and other information are referenced. After the relative position matching of the flight strip is completed, the image control points of the inclined model are used to align the point cloud data to the inclined model position; the data after manual flight strip alignment is repeatedly used to identify the problem area. If there is a problem, manual flight strip alignment is continued. If the data is qualified, the next step of point cloud classification is carried out.

[0099] like Figure 5 As shown, a method for establishing a rock mass information model based on laser point cloud, wherein the specific steps of the refined point cloud classification are:

[0100] Automatic point cloud classification by software: starting from the lowest point of the point cloud as the ground point, the angle and jump height of the adjacent laser point and its connecting line are calculated to separate the ground point data;

[0101] Manual inspection and classification: Build a DEM model based on ground points, remove clutter, repair uncalculated ground points, and subdivide ground points.

[0102] like Figure 6 As shown, a method for establishing a rock mass information model based on laser point cloud, wherein the specific steps of establishing the triangulated network model are:

[0103] Rock mass point cloud preprocessing: further classify the ground points according to different areas of the rock mass. For the smoother rock mass top surface, appropriate thinning can be done. For the complex structure, crack area and some areas with fewer laser points, appropriate point cloud data can be added according to the tilt model.

[0104] Constructing triangulated network model: constructing triangulated network for ground laser points, maintaining uniform change of adjacent triangulated network surface normals, pointing to the outside of the rock mass. The main steps involve obtaining point cloud normals, separating rock blocks according to normal changes, filtering and simplifying point clouds, constructing networks inside different rock blocks, constructing networks between rock blocks, and simplifying meshes;

[0105] Auxiliary network construction method: According to the position information and time information of the point cloud, the line segments of the adjacent time laser points of the single scan line are connected;

[0106] The triangulated network model was established: the final triangulated network model was constructed by basically matching the rock morphology in nature with the tilt model, with a continuous and smooth rock surface, continuous cracks, no adhesion between the rock walls on both sides of the cracks, and no vegetation or clutter features.

[0107] The specific steps of constructing the triangulated network model are:

[0108] Point cloud normal calculation: The calculation process is given a point p and its adjacent points {pi|i=1,…,Np}, the normal at p is np, and the surface is fitted on the adjacent points and the normal is calculated from the fitted surface. The n-jet surface model is used. The surface is represented by a polynomial function Jn: that is, R2→R, mapping the coordinates of point P (x, y) to their height z in the tangent space:

[0109]

[0110] To simplify notation, β is represented as a vector:

[0111]

[0112] At the same time, define

[0113]

[0114] At this time, the normal vector of the fitted surface is expressed as:

[0115]

[0116] Where (x, y) is the coordinate value of point P, the adjacent point of point P is Pi, z represents the coordinate (x, y) mapped to their height in the tangent space, and the Jn surface uses a polynomial function. is the vector coefficient, np is the normal vector of the fitted surface,

[0117] When constructing the mesh of the laser point cloud, the normal vector np of the fitting surface should be ensured to change gradually and continuously, and the points should be prioritized for meshing. The point cloud should be grouped and meshed according to the four cases where np is equal to 0 (360), 90, 180, and 270.

[0118] Connecting laser points to form a network: It mainly constructs the point cloud into a triangulated network. The principle of network construction is to take a triangle ABC from the triangle set, calculate the circumscribed circle of the triangle, then insert a point P1, find all triangles whose circumscribed circles contain this point, delete these triangles to form a cavity, and finally connect the laser points into a triangulated network;

[0119]

[0120] In the formula: where xyz corresponds to the coordinate values ​​of the interpolation point P and the triangle vertex ABC respectively, r represents the radius of the circumscribed circle of point P in the triangle. The specific vertices are P(x0,y0)A(x_{0},y_{0}), A(x1,y1)A(x_{1},y_{1}), B(x2,y2)B(x_{2},y_{2}), C(x3,y3)C(x_{3},y_{3}).

[0121] like Figure 7 As shown, a method for establishing a rock mass information model based on laser point cloud, wherein the specific steps of establishing the rock morphology model are:

[0122] Basic model processing: correct the triangles with wrong directions, weld the nodes of triangulated mesh, and delete isolated points, lines and faces;

[0123] Structural surface repair: refer to the tilt model, analyze the rock structure, delete the structural surface that is not accurately matched with the tilt model and point cloud data, and reconnect and draw the correct structural surface according to the point cloud situation;

[0124] Point cloud fit check: The classified ground point cloud and tilt model are fit checked with the established triangulated network model. If an erroneous surface is found, the surface needs to be re-trimmed.

[0125] Generalized model: According to different rock forms, the model nodes are generalized within a certain tolerance range to reduce the number of structural faces. The simplified model is rechecked for point cloud fitting and the faces with large errors are modified.

[0126] Basic morphological measurement: Analyze and cut the rock according to the model's morphology, crack orientation and other characteristics to form small pieces of dangerous rock mass. Measure the occurrence of each small piece of rock mass, simplify the model, and perform morphological measurements of the center of gravity, inclination, and occurrence information.

[0127] like Figure 8 As shown, a method for establishing a rock mass information model based on laser point cloud, wherein the specific steps of establishing the rock mass information model are:

[0128] Collection of survey attribute information: First, the basic structure and type of the rock are determined by the internal staff, and then verified by on-site survey; rock sampling is carried out to measure the volume, weight and other information, and drilling surveys are carried out on the surrounding terrain to collect geological condition information;

[0129] Information entry: Establish a model database, set up an independent ID number for each rock, establish an attribute information table based on the rock ID, enter the survey attribute information, rock added structure, rock type, density, and geological conditions' own attribute information into the corresponding rock model information table to form a rock information model.

[0130] like Fig. 9 As shown, a method for establishing a rock mass information model based on laser point cloud, wherein the specific steps of establishing the rock mass analysis system are:

[0131] System information input: Temperature, precipitation, wind direction, forest-related environmental information, groundwater, crustal movement, and fault zone-related geological information are input into the system as influencing factors. The corresponding influencing parameters of the influencing factors are given through experimental data. The parameters mainly include the magnitude, direction, duration, triggering conditions of the force exerted by the influencing factors on the rock, and the damage caused by the influencing factors to the rock body, including rainwater immersion corrosion, wind erosion of the rock, and vegetation weathering of the rock.

[0132] Motion simulation: The system uses time as a variable, accumulates the effects of various influencing factors on rocks over time until the rocks collapse, records the collapse process, and analyzes the impact of collapse and landslides; System information input: The system inputs temperature, precipitation, wind direction, forest-related environmental information, and groundwater, crustal movement, and fault zone-related geological information as influencing factors, and gives the corresponding influencing parameters of the influencing factors through experimental data. The parameters mainly include the magnitude, direction, duration, triggering conditions of the force exerted by the influencing factors on the rocks, and the damage caused by the influencing factors to the rock body, including rainwater immersion corrosion, wind erosion of rocks, and vegetation weathering of rocks.

[0133] like Fig.10 As shown, a method for establishing a rock mass information model based on laser point cloud, wherein the specific steps of the collapse simulation are:

[0134] Rock trajectory analysis: Rockfall is considered a rigid body motion. The rock trajectory is governed by the interaction between the rock and the ground. The model contains six main state variables: three translational velocities and three rotational velocities of the rockfall, from which the kinetic energy is obtained, jump distance and jump height, rock orientation and rotation speed are included in the rock / ground interaction. The model takes into account the real rock shape and calculates the motion trajectory of each rock body on the model, including jump height, sliding speed, rotation speed, rotation and total kinetic energy. All possible motion modes of the rockfall are deterministically simulated. The dynamics of the trajectory of a single rock body are examined separately, and statistical analysis is performed on multiple trajectory sets;

[0135] Simulation features: The system simulates the true nature of rock fall motion behavior, consistent with rock shape and release kinematics of site-specific rock fall hazards;

[0136] Report analysis: Based on the analysis of rock collapse and the damage caused to the surrounding areas, disaster prevention and mitigation measures are taken in advance to form a geological disaster analysis report.

[0137] The following is an example of a specific implementation:

[0138] In order to enable those skilled in the art to better understand the present invention, the present invention will be described in detail below with reference to the accompanying drawings and specific embodiments. The exemplary embodiments and descriptions of the present invention are used to explain the present invention but are not intended to limit the present invention.

[0139] Embodiment 2:

[0140] The project is located in a mountainous area in northern China. There is a village with about 50 households at the foot of the mountain. Landslides and mud-rock flows are relatively frequent, which have a great impact on the lives of the villagers. Figure 1 A schematic diagram of the process of establishing a rock mass information model based on laser point cloud provided in this embodiment is shown in FIG. Figure 1 As shown, a method for establishing a rock mass information model based on laser point cloud is provided, wherein the method comprises the following steps: UAV laser data acquisition: the UAV is equipped with a high-frequency, long-range laser radar, and performs a terrain-simulating flight in the rock mass area, and performs a "well"-shaped flight acquisition technology to collect original point clouds, and uses the laser's good dense forest penetration ability to separate the ground points under the forest; tilt model establishment: the UAV is equipped with a high-definition camera to collect images of the rock mass and establish an oblique photography model by using a close-up photogrammetry method; laser point cloud flight strip re-matching: for different sorties, the matching of cross-flight flight strip data is relatively poor, and the surface point cloud situation around the dangerous rock mass is checked through the relatively smooth surface of the mountain, and the point cloud is re-accurately matched by the laser emission angle and emission time information; refined point cloud classification: refined point cloud classification is performed to distinguish different types of point clouds such as ground points, tree points, low-lying vegetation points, rock wall points, and crack points. ; Establishment of triangulated network model: Generate rock triangulated network model through classified laser point cloud; Establishment of rock morphological model: Combined with the tilt model, further repair the triangulated network model to establish the rock morphological model; Establishment of rock information model: On the basis of the rock morphological model, add the rock structure, rock type, density, geological conditions and other related property information to form a rock information model; Establishment of rock analysis system: Add environmental and geological information to the system to form a set of rock information models that are almost consistent with the outside world, and use simulation methods to study rock changes and predict collapse time and disaster severity; Collapse simulation: After setting the rock collapse, further simulate the rock's movement trajectory and fragmentation degree, analyze the rock trajectory, simulation characteristics and the impact of landslides, analyze rock collapse and the hazards caused to the surrounding areas, make disaster prevention and mitigation measures in advance, and form a geological disaster analysis report.

[0141] like Figure 2 As shown, a method for establishing a rock mass information model based on laser point cloud, wherein the specific steps of the drone laser data collection are:

[0142] On-site survey: Use a small drone to observe the mountain, combine the on-site situation analysis, determine the location of suspected dangerous rock bodies, cracks and other unfavorable geological bodies, and select a flat, clean and hard site with an area of ​​not less than 5 meters by 5 meters as the drone landing site without any obstructions such as pipe corridors and wires above, and without communication interference in the remote control ground test. The project landing site was selected as a school basketball court. The survey found more than 50 suspected dangerous rock bodies and more than 10 large cracks;

[0143] Route design: Route is laid out along the main rock surface and crack direction, and routes are laid out crosswise in important rock mass areas, and the routes are laid out in a "well" shape;

[0144] Laser equipment parameter setting: To ensure the quality of collected data, point cloud density, overlap, flight altitude, and laser radar are set respectively. The appropriate laser emission frequency, scanning frequency, scanning angle and other parameters are set to ensure that the original point cloud density is greater than 1200 points / ㎡; the flight zone overlap is greater than 50%, and the routes are crossed to ensure that each area has point cloud coverage in the four directions of "front, back, left, and right", that is, each area should have at least 4 degrees of overlap; when setting the flight altitude, it is determined according to the performance of different laser radar equipment and terrain to ensure that the point cloud data is uniform, and the terrain-simulating flight mode is adopted, that is, the relative height to the ground is consistent;

[0145] The flight altitude calculation formula is:

[0146]

[0147] Where:

[0148] H represents the maximum flight altitude, and L represents the maximum measurement range of the designed laser emission frequency under the actual target reflectivity; in the design of the laser radar data acquisition parameters, the parameter emission frequency is set to 600kHz and the scanning rate is set to 200 revolutions. The UAV speed is 8m / s, the flight zone overlap is set to 50%, and the relative flight altitude is 200M. According to these parameters, laser data collection is carried out, and the total survey area is 0.85 square kilometers. The UAV laser flight is 4 sorties and 16 routes.

[0149] like Figure 3 As shown, a method for establishing a rock mass information model based on laser point cloud, wherein the specific steps of establishing the tilt model are:

[0150] Site survey: Choose a flat, clean, hard site with an area of ​​no less than 5 meters by 5 meters, with no obstructions such as pipe corridors and wires above, and no communication interference during the remote control ground test as the drone take-off and landing site;

[0151] Data collection: Close-up photogrammetry, 80% heading overlap, 70% heading, circumferential flight for key areas, and the coordinate system of the image control points is consistent with the point cloud;

[0152] Tilt model establishment: Establish a tilt model to assist in modifying the rock morphology model.

[0153] like Figure 4 As shown, a method for establishing a rock mass information model based on laser point cloud, wherein the specific steps of re-matching the laser point cloud flight path are:

[0154] Strip adjustment correction: single route adjustment method is used, P and Q are defined as the heading angle, pitch angle, and roll angle parameters of adjacent laser point cloud strips, and the correction equations for heading angle, pitch angle, and roll angle are established in the overlapping area between routes:

[0155]

[0156] Among them, d(X, Y, Z) represents the correction of the point cloud data with the same name in different strips in the X, Y, and Z directions in the geographic space coordinates, Tx, Ty, and Tz represent the difference of the point cloud data in the X, Y, and Z directions in the geographic space coordinates, and ah, ar, and ap represent the heading angle, pitch angle, and roll angle correction of the point cloud strip corresponding to the track line;

[0157] Manual correction and matching: The point cloud is displayed in the point cloud software with the flight strip as the mark through manual inspection and correction. The matching plane and elevation errors between the flight strips are analyzed block by block in the form of cross-section, and the matching problem areas are checked for further alignment; based on one of the through flight strips, the remaining flight strips are translated to the reference flight strip for strict alignment. If there are uneven errors between the flight strips, manual flight strip matching is required in different areas. The laser point cloud has temporal continuity. When its characteristics are not easy to distinguish due to spatial factors, the laser time and other information are referred to. After the relative position matching of the flight strip is completed, the image control points of the inclined model are used to align the point cloud data to the inclined model position; the data after manual flight strip alignment is repeatedly used to identify the problem area. If there is a problem, manual flight strip alignment is continued. Finally, the matching error between flight strips is controlled within 2 cm, and the data is qualified for the next step of point cloud classification.

[0158] like Figure 5 As shown, a method for establishing a rock mass information model based on laser point cloud, wherein the specific steps of the refined point cloud classification are:

[0159] Automatic point cloud classification by software: starting from the lowest point of the point cloud as the ground point, the angle and jump height of the adjacent laser point and its connecting line are calculated to separate the ground point data;

[0160] Manual inspection and classification: Build a DEM model based on ground points, remove clutter, repair uncalculated ground points, and subdivide ground points.

[0161] like Figure 6 As shown, a method for establishing a rock mass information model based on laser point cloud, wherein the specific steps of establishing the triangulated network model are:

[0162] Rock mass point cloud preprocessing: further classify the ground points according to different areas of the rock mass. For the smoother rock mass top surface, appropriate thinning can be done. For the complex structure, crack area and some areas with fewer laser points, appropriate point cloud data can be added according to the tilt model.

[0163] Constructing triangulated network model: constructing triangulated network for ground laser points, maintaining uniform change of adjacent triangulated network surface normals, pointing to the outside of the rock mass. The main steps involve obtaining point cloud normals, separating rock blocks according to normal changes, filtering and simplifying point clouds, constructing networks inside different rock blocks, constructing networks between rock blocks, and simplifying meshes;

[0164] Auxiliary network construction method: According to the position information and time information of the point cloud, the line segments of the adjacent time laser points of the single scan line are connected;

[0165] The triangulated network model was established: the final triangulated network model was constructed by basically matching the rock morphology in nature with the tilt model, with a continuous and smooth rock surface, continuous cracks, no adhesion between the rock walls on both sides of the cracks, and no vegetation or clutter features.

[0166] The specific steps of constructing the triangulated network model are:

[0167] Point cloud normal calculation: The calculation process is given a point p and its adjacent points {pi|i=1,…,Np}, the normal at p is np, and the surface is fitted on the adjacent points and the normal is calculated from the fitted surface. The n-jet surface model is used. The surface is represented by a polynomial function Jn: that is, R2→R, mapping the coordinates of point P (x, y) to their height z in the tangent space:

[0168]

[0169] To simplify notation, β is represented as a vector:

[0170]

[0171] At the same time, define

[0172]

[0173] At this time, the normal vector of the fitted surface is expressed as:

[0174]

[0175] Where (x, y) is the coordinate value of point P, the adjacent point of point P is Pi, z represents the coordinate (x, y) mapped to their height in the tangent space, and the Jn surface uses a polynomial function. is the vector coefficient, np is the normal vector of the fitted surface,

[0176] When constructing the mesh of the laser point cloud, the normal vector np of the fitting surface should be ensured to change gradually and continuously, and the points should be prioritized for meshing. The point cloud should be grouped and meshed according to the four cases where np is equal to 0 (360), 90, 180, and 270.

[0177] Connecting laser points to form a network: It mainly constructs the point cloud into a triangulated network. The principle of network construction is to take a triangle ABC from the triangle set, calculate the circumscribed circle of the triangle, then insert a point P1, find all triangles whose circumscribed circles contain this point, delete these triangles to form a cavity, and finally connect the laser points into a triangulated network;

[0178]

[0179] In the formula: where xyz corresponds to the coordinate values ​​of the interpolation point P and the triangle vertex ABC respectively, r represents the radius of the circumscribed circle of point P in the triangle. The specific vertices are P(x0,y0)A(x_{0},y_{0}), A(x1,y1)A(x_{1},y_{1}), B(x2,y2)B(x_{2},y_{2}), C(x3,y3)C(x_{3},y_{3}).

[0180] like Figure 7 As shown, a method for establishing a rock mass information model based on laser point cloud, wherein the specific steps of establishing the rock morphology model are:

[0181] Basic model processing: correct the triangles with wrong directions, weld the nodes of triangulated mesh, and delete isolated points, lines and faces;

[0182] Structural surface repair: refer to the tilt model, analyze the rock structure, delete the structural surface that is not accurately matched with the tilt model and point cloud data, and reconnect and draw the correct structural surface according to the point cloud situation;

[0183] Point cloud fit check: The classified ground point cloud and tilt model are fit checked with the established triangulated network model. If an erroneous surface is found, the surface needs to be re-trimmed.

[0184] Generalized model: According to different rock forms, the model nodes are generalized within a certain tolerance range to reduce the number of structural faces. The simplified model is rechecked for point cloud fitting and the faces with large errors are modified.

[0185] Basic morphological measurement: According to the model's morphology, crack orientation and other characteristics, the rock was analyzed and cut to form small pieces of dangerous rock mass. The occurrence of each small piece of rock mass was measured, the model was simplified, and the center of gravity, inclination, and occurrence information were measured. A total of 12 dangerous rock masses were found in the project, and 3 of them had a greater impact on the roads and villages below, as shown in the following table:

[0186]

[0187] like Figure 8 As shown, a method for establishing a rock mass information model based on laser point cloud, wherein the specific steps of establishing the rock mass information model are:

[0188] Collection of survey attribute information: First, the basic structure and type of the rock are determined by the internal staff, and then verified by on-site survey; rock sampling is carried out to measure the volume, weight and other information, and drilling surveys are carried out on the surrounding terrain to collect geological condition information;

[0189] Information entry: Establish a model database, set up an independent ID number for each rock, establish an attribute information table based on the rock ID, enter the survey attribute information, rock added structure, rock type, density, and geological conditions' own attribute information into the corresponding rock model information table to form a rock information model.

[0190] like Fig. 9 As shown, a method for establishing a rock mass information model based on laser point cloud, wherein the specific steps of establishing the rock mass analysis system are:

[0191] System information input: Temperature, precipitation, wind direction, forest-related environmental information, groundwater, crustal movement, and fault zone-related geological information are input into the system as influencing factors. The corresponding influencing parameters of the influencing factors are given through experimental data. The parameters mainly include the magnitude, direction, duration, triggering conditions of the force exerted by the influencing factors on the rock, and the damage caused by the influencing factors to the rock body, including rainwater immersion corrosion, wind erosion of the rock, and vegetation weathering of the rock.

[0192] Motion simulation: The system uses time as a variable, accumulates the effects of various influencing factors on rocks over time until the rocks collapse, records the collapse process, and analyzes the impact of collapse and landslides; System information input: The system inputs temperature, precipitation, wind direction, forest-related environmental information, and groundwater, crustal movement, and fault zone-related geological information as influencing factors, and gives the corresponding influencing parameters of the influencing factors through experimental data. The parameters mainly include the magnitude, direction, duration, triggering conditions of the force exerted by the influencing factors on the rocks, and the damage caused by the influencing factors to the rock body, including rainwater immersion corrosion, wind erosion of rocks, and vegetation weathering of rocks.

[0193] like Fig.10As shown, a method for establishing a rock mass information model based on laser point cloud, wherein the specific steps of the collapse simulation are:

[0194] Rock trajectory analysis: Rockfall is considered a rigid body motion. The rock trajectory is governed by the interaction between the rock and the ground. The model contains six main state variables: three translational velocities and three rotational velocities of the rockfall, from which the kinetic energy is obtained, jump distance and jump height, rock orientation and rotation speed are included in the rock / ground interaction. The model takes into account the real rock shape and calculates the motion trajectory of each rock body on the model, including jump height, sliding speed, rotation speed, rotation and total kinetic energy. All possible motion modes of the rockfall are deterministically simulated. The dynamics of the trajectory of a single rock body are examined separately, and statistical analysis is performed on multiple trajectory sets;

[0195] Simulation features: The system simulates the true nature of rock fall motion behavior, consistent with rock shape and release kinematics of site-specific rock fall hazards;

[0196] Report analysis: Based on the analysis of rock collapse and the damage caused to the surrounding areas, disaster prevention and mitigation measures were taken in advance and a geological disaster analysis report was formed. There are 3 dangerous rock bodies that pose a great threat to the villages and roads below and are expected to collapse within 5 years. The main external force source is rain erosion. Relevant departments have been advised to take disaster reduction measures in advance.

[0197] The foregoing method descriptions and flow charts are provided only as illustrative examples and are not intended to require or imply that the above operations or steps of various aspects must be performed in the order given. As will be appreciated by those skilled in the art, the order of the boxes in the foregoing aspects may be performed in any order. Words such as "thereafter," "then," "next," and the like are not intended to limit the order of operations or steps; these words are only used to guide the reader through the description of the method. In addition, any singular reference to a claim element, for example, the use of the article "a," "an," or "the" is not to be construed as limiting the element to the singular.

[0198] The various illustrative logic blocks, modules, circuits, and algorithm steps described in conjunction with the aspects disclosed herein can all be implemented as electronic hardware, computer software, or a combination thereof. In order to clearly represent the interchangeability between hardware and software, the various illustrative components, blocks, modules, circuits, and steps are generally described above around their functions. Whether such functions are implemented as hardware or software depends on the specific application and the design constraints imposed on the entire method. A skilled person can implement the described functions in an alternative manner for each specific application, but such implementation decisions should not be interpreted as causing a departure from the scope of protection of the present invention.

[0199] The method in the embodiment of the present invention utilizes a drone equipped with a laser radar to perform close-up terrain-simulating flight, and utilizes the good vegetation penetration ability of laser point clouds to directly measure structures such as collapse, cracks, and falling rocks in the mountain under the dense forest. Compared with the original field survey, photogrammetry and other methods, the information obtained by the laser point cloud is more accurate and comprehensive, and can directly confirm the geometric dimensions, cliff section shape, structural surface characteristics, lithology, and boundary conditions of the dangerous rock mass. The laser radar collects high-density laser point cloud data to construct a refined rock mass information model, which can simplify complex field investigation work to indoor model analysis. Compared with manual experience judgment, the simulation model motion simulation method can effectively analyze and predict the collapse and landslide of the dangerous rock mass, avoid disasters, and reduce the impact of geological disasters.

[0200] The foregoing description of the disclosed aspects is provided to enable any person skilled in the art to implement or use the present invention. Various modifications to these aspects will be apparent to those skilled in the art, and the general principles defined herein may also be applied to other embodiments without departing from the spirit and scope of the present invention. Therefore, the present invention is not intended to be limited to the aspects provided herein, but rather to the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A method for establishing a rock mass information model based on laser point cloud, characterized in that The method includes the following steps: UAV laser data acquisition: The UAV is equipped with a high-frequency, long-range laser radar to perform ground-flight simulation in the rock area, and uses the "well"-shaped flight acquisition technology to collect original point clouds, and uses the laser's good dense forest penetration ability to separate the ground points under the forest; Inclined model establishment: The UAV is equipped with a high-definition camera to use close-up photogrammetry to collect images of the rock mass and establish an inclined photography model; Laser point cloud flight belt re-matching: For different sorties, the matching of cross-flight flight belt data is relatively poor, and the surface point cloud situation of the relatively smooth mountain around the dangerous rock mass is checked, and the point cloud is re-accurately matched by the laser emission angle and emission time information; Refined point cloud classification: Refined point cloud classification is performed on different types of point clouds such as ground points, tree points, low-lying vegetation points, rock wall points, and crack points; Triangulated network model establishment: By classifying well Generate a rock mass triangulation model from the laser point cloud; establish a rock morphology model: combine the tilt model to further repair the triangulation model and establish a rock morphology model; establish a rock information model: on the basis of the rock morphology model, add the rock structure, rock type, density, geological conditions and other related property information to form a rock information model; establish a rock analysis system: add environmental and geological information to the system to form a set of rock information models that are almost consistent with the outside world, and use simulation methods to study rock changes and predict the collapse time and disaster extent; collapse simulation: after setting the rock collapse, further simulate the rock's movement trajectory and fragmentation degree, analyze the rock trajectory, simulation characteristics and the impact of landslides, analyze rock collapse and the hazards caused to the surrounding area, make disaster prevention and mitigation measures in advance, and form a geological disaster analysis report.

2. The method for establishing a rock mass information model based on laser point cloud according to claim 1, characterized in that: The specific steps of the UAV laser data collection are: On-site survey: Use a small drone to observe the mountain, analyze the on-site situation, and determine the location of suspected dangerous rock masses, cracks and other unfavorable geological bodies. At the same time, select a flat, clean and hard site with an area of ​​not less than 5 meters by 5 meters as the drone landing site without any obstructions such as pipe corridors and wires above, and without communication interference from the remote control ground test; Route design: Route is laid out along the main rock surface and crack direction, and routes are laid out crosswise in important rock mass areas, and the routes are laid out in a "well" shape; Laser equipment parameter setting: To ensure the quality of collected data, point cloud density, overlap, flight altitude, and laser radar are set respectively. The appropriate laser emission frequency, scanning frequency, scanning angle and other parameters are set to ensure that the original point cloud density is greater than 1200 points / ㎡; the flight zone overlap is greater than 50%, and the routes are crossed to ensure that each area has point cloud coverage in the four directions of "front, back, left, and right", that is, each area should have at least 4 degrees of overlap; when setting the flight altitude, it is determined according to the performance of different laser radar equipment and terrain to ensure that the point cloud data is uniform, and the terrain-simulating flight mode is adopted, that is, the relative height to the ground is consistent; The flight altitude calculation formula is: Where: H represents the maximum flight altitude, and L represents the maximum measurement range of the designed laser emission frequency under the actual target reflectivity; in the design of the laser radar data acquisition parameters, the parameter emission frequency is set to 600kHz and the scanning rate is set to 200 revolutions. The UAV speed is 8m / s, the flight zone overlap is set to 50%, and the relative altitude is 200M. Laser data acquisition is carried out according to these parameters.

3. The method for establishing a rock mass information model based on laser point cloud according to claim 1, characterized in that: The specific steps of establishing the tilt model are: Site survey: Choose a flat, clean, hard site with an area of ​​no less than 5 meters by 5 meters, with no obstructions such as pipe corridors and wires above, and no communication interference during the remote control ground test as the drone take-off and landing site; Data collection: Close-up photogrammetry, 80% heading overlap, 70% heading, circumferential flight for key areas, and the coordinate system of the image control points is consistent with the point cloud; Tilt model establishment: Establish a tilt model to assist in modifying the rock morphology model.

4. The method for establishing a rock mass information model based on laser point cloud according to claim 1, characterized in that: The specific steps of re-matching the laser point cloud flight path are as follows: Strip adjustment correction: single route adjustment method is used, P and Q are defined as the heading angle, pitch angle, and roll angle parameters of adjacent laser point cloud strips, and the correction equations for heading angle, pitch angle, and roll angle are established in the overlapping area between routes: Among them, d(X, Y, Z) represents the correction of the point cloud data with the same name in different strips in the X, Y, and Z directions in the geographic space coordinates, Tx, Ty, and Tz represent the difference of the point cloud data in the X, Y, and Z directions in the geographic space coordinates, and ah, ar, and ap represent the heading angle, pitch angle, and roll angle correction of the point cloud strip corresponding to the track line; Manual correction and matching: The point cloud is displayed in the point cloud software with the flight strip as the mark through manual inspection and correction. The matching plane and elevation errors between the flight strips are analyzed block by block in the form of cross-section, and the matching problem areas are checked for further alignment; based on one of the through flight strips, the remaining flight strips are translated to the reference flight strip for strict alignment. If there are uneven errors between the flight strips, manual flight strip matching is required in different areas. The laser point cloud has temporal continuity. When its characteristics are difficult to distinguish due to spatial factors, the laser time and other information are referenced. After the relative position matching of the flight strip is completed, the image control points of the inclined model are used to align the point cloud data to the inclined model position; the data after manual flight strip alignment is repeatedly used to identify the problem area. If there is a problem, manual flight strip alignment is continued. If the data is qualified, the next step of point cloud classification is carried out.

5. The method for establishing a rock mass information model based on laser point cloud according to claim 1, characterized in that: The specific steps of the refined point cloud classification are: Automatic point cloud classification by software: starting from the lowest point of the point cloud as the ground point, the angle and jump height of the adjacent laser point and its connecting line are calculated to separate the ground point data; Manual inspection and classification: Build a DEM model based on ground points, remove clutter, repair uncalculated ground points, and subdivide ground points.

6. The method for establishing a rock mass information model based on laser point cloud according to claim 1, characterized in that: The specific steps of establishing the triangulated network model are: Rock mass point cloud preprocessing: further classify the ground points according to different areas of the rock mass. For the smoother rock mass top surface, appropriate thinning can be done. For the complex structure, crack area and some areas with fewer laser points, appropriate point cloud data can be added according to the tilt model. Constructing triangulated network model: constructing triangulated network for ground laser points, maintaining uniform change of adjacent triangulated network surface normals, pointing to the outside of the rock mass. The main steps involve obtaining point cloud normals, separating rock blocks according to normal changes, filtering and simplifying point clouds, constructing networks inside different rock blocks, constructing networks between rock blocks, and simplifying meshes; Auxiliary network construction method: According to the position information and time information of the point cloud, the line segments of the adjacent time laser points of the single scan line are connected; The triangulated network model was established: the final triangulated network model was constructed by basically matching the rock morphology in nature with the tilt model, with a continuous and smooth rock surface, continuous cracks, no adhesion between the rock walls on both sides of the cracks, and no vegetation or clutter features. The specific steps of constructing the triangulated network model are: Point cloud normal calculation: The calculation process is given a point p and its adjacent points {pi|i=1,…,Np}, the normal at p is np, and the surface is fitted on the adjacent points and the normal is calculated from the fitted surface. The n-jet surface model is used. The surface is represented by a polynomial function Jn: that is, R2→R, mapping the coordinates of point P (x, y) to their height z in the tangent space: To simplify notation, β is represented as a vector: At the same time, define At this time, the normal vector of the fitted surface is expressed as: Where (x, y) is the coordinate value of point P, the adjacent point of point P is Pi, z represents the coordinate (x, y) mapped to their height in the tangent space, and the Jn surface uses a polynomial function. is the vector coefficient, np is the normal vector of the fitted surface, When constructing the mesh of the laser point cloud, the normal vector np of the fitting surface should be ensured to change gradually and continuously, and the points should be prioritized for meshing. The point cloud should be grouped and meshed according to the four cases where np is equal to 0 (360), 90, 180, and 270. Connecting laser points to form a network: It mainly constructs the point cloud into a triangulated network. The principle of network construction is to take a triangle ABC from the triangle set, calculate the circumscribed circle of the triangle, then insert a point P1, find all triangles whose circumscribed circles contain this point, delete these triangles to form a cavity, and finally connect the laser points into a triangulated network; In the formula: xyz corresponds to the coordinates of the interpolation point P and the triangle vertex ABC, r represents the radius of the circumscribed circle of point P in the triangle, and the specific vertex is P(x0,y0)A(x_{0},y_{0}), A(x1,y1)A(x_{1},y_{1}), B(x2,y2)B(x_{2},y_{2}), C(x3,y3)C(x_{3},y_{3}).

7. The method for establishing a rock mass information model based on laser point cloud according to claim 1, characterized in that: The specific steps of establishing the rock morphology model are: Basic model processing: correct the triangles with wrong directions, weld the nodes of triangulated mesh, and delete isolated points, lines and faces; Structural surface repair: refer to the tilt model, analyze the rock structure, delete the structural surface that is not accurately matched with the tilt model and point cloud data, and reconnect and draw the correct structural surface according to the point cloud situation; Point cloud fit check: The classified ground point cloud and tilt model are fit checked with the established triangulated network model. If an erroneous surface is found, the surface needs to be re-trimmed. Generalized model: According to different rock forms, the model nodes are generalized within a certain tolerance range to reduce the number of structural faces. The simplified model is rechecked for point cloud fitting and the faces with large errors are modified. Basic morphological measurement: Analyze and cut the rock according to the model's morphology, crack orientation and other characteristics to form small pieces of dangerous rock mass. Measure the occurrence of each small piece of rock mass, simplify the model, and perform morphological measurements of the center of gravity, inclination, and occurrence information.

8. The method for establishing a rock mass information model based on laser point cloud according to claim 1, characterized in that: The specific steps of establishing the rock mass information model are: Collection of survey attribute information: First, the basic structure and type of the rock are determined by the internal staff, and then verified by on-site survey; rock sampling is carried out to measure the volume, weight and other information, and drilling surveys are carried out on the surrounding terrain to collect geological condition information; Information entry: Establish a model database, set up an independent ID number for each rock, establish an attribute information table based on the rock ID, enter the survey attribute information, rock added structure, rock type, density, and geological conditions' own attribute information into the corresponding rock model information table to form a rock information model.

9. The method for establishing a rock mass information model based on laser point cloud according to claim 1, characterized in that: The specific steps of establishing the rock mass analysis system are: System information input: Temperature, precipitation, wind direction, forest-related environmental information, groundwater, crustal movement, and fault zone-related geological information are input into the system as influencing factors. The corresponding influencing parameters of the influencing factors are given through experimental data. The parameters mainly include the magnitude, direction, duration, triggering conditions of the force exerted by the influencing factors on the rock, and the damage caused by the influencing factors to the rock body, including rainwater immersion corrosion, wind erosion of the rock, and vegetation weathering of the rock. Motion simulation: The system uses time as a variable, accumulates the effects of various influencing factors on rocks over time until the rocks collapse, records the collapse process, and analyzes the impact of collapse and landslides; System information input: The system inputs temperature, precipitation, wind direction, forest-related environmental information, and groundwater, crustal movement, and fault zone-related geological information as influencing factors, and gives the corresponding influencing parameters of the influencing factors through experimental data. The parameters mainly include the magnitude, direction, duration, triggering conditions of the force exerted by the influencing factors on the rocks, and the damage caused by the influencing factors to the rock body, including rainwater immersion corrosion, wind erosion of rocks, and vegetation weathering of rocks. Motion simulation: The system uses time as a variable, and accumulates the effects of various influencing factors on the rock over time until the rock collapses, records the collapse process, and analyzes the impact of the collapse landslide.

10. The method for establishing a rock mass information model based on laser point cloud according to claim 1, characterized in that: The specific steps of the collapse simulation are: Rock trajectory analysis: Rockfall is considered a rigid body motion. The rock trajectory is governed by the interaction between the rock and the ground. The model contains six main state variables: three translational velocities and three rotational velocities of the rockfall, from which the kinetic energy is obtained, jump distance and jump height, rock orientation and rotation speed are included in the rock / ground interaction. The model takes into account the real rock shape and calculates the motion trajectory of each rock body on the model, including jump height, sliding speed, rotation speed, rotation and total kinetic energy. All possible motion modes of the rockfall are deterministically simulated. The dynamics of the trajectory of a single rock body are examined separately, and statistical analysis is performed on multiple trajectory sets; Simulation features: The system simulates the true nature of rock fall motion behavior, consistent with rock shape and release kinematics of site-specific rock fall hazards; Report analysis: Based on the analysis of rock collapse and the damage caused to the surrounding areas, disaster prevention and mitigation measures are taken in advance to form a geological disaster analysis report.

Images