Mine working face roof deformation monitoring method based on laser radar

Abstract

The invention is suitable for the technical field of nondestructive testing of non-coal mine internal investigation and safety detection, and provides a mine working face roof deformation monitoring method based on a laser radar, which is specially used for a complex environment of a mine working face. The method comprises the following steps: S1, collecting laser radar data; s2, preprocessing the point cloud data; s3, deformation monitoring of a working face top plate; and S4, deformation analysis of the working face roof. The method comprises the following steps: after specific data preprocessing, constructing a triangular mesh model for scanned two-stage point clouds by utilizing triangular mesh division, performing central point extraction on each triangular surface of a second-stage triangular mesh model to form a second-stage central point cloud, performing distance calculation on the second-stage central point cloud and the first-stage triangular mesh model to form a second-stage central point cloud, and calculating the distance between the second-stage central point cloud and the first-stage triangular mesh model; the method is used as a basis for deformation quantity and deformation risk assessment. According to the method, the nondestructive testing technology is utilized, the rock wall in the mine does not need to be damaged, and the detection safety is improved under the condition that the labor cost and the workload are reduced.

Description

Technical Field

[0001] The present invention belongs to the technical field of non-destructive testing for internal exploration and safety detection of non-coal mines, and in particular relates to a method for monitoring roof deformation of a mine working face based on laser radar. Background Art

[0002] The working face of a mine is an essential mining route for mining mineral resources and other mining resources, especially the roof above it, which is often a safety issue that needs to be considered. The working face of a mine often has dangers such as collapse, rock shedding, rock wall cracks, and rock deformation. Among them, the rock shedding and deformation of the roof account for the largest proportion of all kinds of diseases. At present, non-destructive testing technology has been increasingly used in the monitoring of dangerous rock masses in the working face of a mine and the normal maintenance of the operating tunnels to ensure the personal safety of the staff.

[0003] Nondestructive testing will not damage the internal structure of the mine. Based on the principles of light and electricity, it is possible to detect whether the roof of the working face has fallen off, deformed or other diseases. The patent with the authorization announcement number CN203412630U, "A coal mine roadway roof detection system based on fiber grating", discloses a fiber grating method that can be used to detect dangerous rock masses on the roof of coal mine roadways. The solution is to fix the fiber grating on the load-bearing block, and then fix the load-bearing block on the side wall or roof of the roadway.

[0004] There are two obvious problems with this solution: first, as time goes by and the harsh environment inside the mine changes, the fiber grating fixed on the force-bearing block will become unstable, resulting in inaccurate detection data; second, the technology requires multiple force-bearing blocks to be fixed on the side walls or roof of the tunnel, which will cause changes in the internal structure of the mine and be somewhat destructive.

[0005] The use of laser radar scanning detection will not cause any damage to the interior of the mine, and it is also more adaptable to the environment. The monitoring of the mine working face roof is more accurate and safer than the fiber grating method. Therefore, laser radar scanning detection is a feasible and effective solution. Summary of the invention

[0006] In view of the above problems, the purpose of the present invention is to provide a mine working face roof deformation monitoring method based on laser radar, aiming to solve the above technical problems.

[0007] The present invention adopts the following technical solution:

[0008] A method for monitoring roof deformation of a mine working face based on laser radar comprises the following steps:

[0009] Step S1: LiDAR data collection

[0010] Deformation monitoring devices including laser radar and inclination sensors are arranged inside the mine, and then time-difference laser radar scanning and data collection are carried out to obtain the original point cloud data and corresponding inclination information of the mine working face;

[0011] Step S2: Point cloud data preprocessing

[0012] After the point cloud data is format converted and frame stacked, the entire point cloud data coordinates are calibrated using the inclination information to achieve the point cloud data coordinate angle transformation, and then pre-processed through ICP registration and interference point filtering, and finally point cloud segmentation is performed;

[0013] Step S3: Working surface roof deformation monitoring

[0014] The Delaunay triangulation is used to construct a triangular mesh model for the scanned point cloud data of the two phases, and the center point of each triangular face of the second phase triangular mesh model is extracted to form the second phase center point cloud. The distance between the second phase center point cloud and the first phase triangular mesh model is calculated to obtain the point-surface distance.

[0015] Step S4: deformation analysis of the top plate of the working surface

[0016] The deformation point cloud is extracted according to the point-surface distance to form a deformation point cloud cluster. The DBSCAN density clustering algorithm is then used to perform regional density clustering on the deformation point cloud cluster to identify and mark the area where the roof of the mine working face has deformed. Finally, the deformation risk level is evaluated.

[0017] Furthermore, the specific process of step S2 is as follows:

[0018] S21. Use the Linux ROS operating system to convert the original point cloud data in LVX format into point cloud data in PCD format, and use the frame stacking algorithm to overlay each frame to obtain a denser and more realistic point cloud data.

[0019] S22, using the PCL algorithm to transform the coordinate angle of the point cloud after the frame overlap according to the inclination information;

[0020] S23, preprocessing by ICP registration and interference point filtering;

[0021] S24. Using the point cloud after coordinate angle transformation, the point cloud data of the entire tunnel roof is segmented into XYZ coordinate point cloud. The process is as follows: first calculate the length, width, and height of the tunnel, then segment the XY coordinates of the point cloud according to the length and width in the point cloud data, and then segment the Z coordinates of the point cloud according to the height of the roof, and finally obtain the point cloud data containing only the tunnel roof.

[0022] Furthermore, the specific process of step S3 is as follows:

[0023] S31. Constructing triangular mesh model

[0024] The normals of the two-phase roof point cloud data after segmentation are calculated, and the triangular mesh model is constructed using the Delaunay triangular mesh surface reconstruction algorithm;

[0025] S32, extract the center point of each triangular face of the second phase triangular mesh model to form the second phase center point cloud

[0026]

[0027] Among them, x1, y1, z1; x2, y2, z2; x3, y3, z3 are the three-axis coordinate values ​​of the three vertices of the triangle surface, and x, y, z are the three-axis coordinates of the center point of the triangle surface;

[0028] S33, determine the nearest triangle

[0029] For the center point P of a triangle, if the difference between the sum of the areas of the three triangles formed by the three sides of a triangle of the first phase triangular mesh model and the area of ​​this triangle is less than a sufficiently small threshold, then the current triangle is determined to be the nearest triangle to the center point P.

[0030] S33, calculate the distance d between the center point P and the nearest triangle

[0031]

[0032] Where x0, y0, z0 are the coordinates of the current center point P, A, B, C, E are the coefficients of the fitting equation Ax+By+Cz+E=0 of the nearest triangular surface;

[0033] Finally, the distance between the center point of each triangle face of the second-phase triangular mesh model and the corresponding nearest triangle face of the first-phase mesh model is obtained, that is, the point-face distance.

[0034] Furthermore, the specific process of step S4 is as follows:

[0035] S41, the point-surface distance corresponds to the deformation of the center point, and different deformation ranges are marked with different colors;

[0036] S42, forming a deformed point cloud cluster for the extracted deformed point cloud, and then using the DBSCAN density clustering algorithm to perform regional density clustering on the deformed point cloud cluster to divide a plurality of deformed point cloud clusters in different regions;

[0037] S43, performing density filtering on the deformation point cloud clusters, filtering out interference point clouds with a small number of cloud clusters and sparse density, and identifying the remaining deformation point cloud clusters as point clouds in the deformation area of ​​the mine working face roof;

[0038] S44. Calculate the average deformation displacement and point cloud ratio in each deformation displacement interval as risk assessment grade standards for risk assessment.

[0039] Furthermore, in step S44, for the deformation displacement interval in which the deformation point cloud accounts for more than 25%, if the relative average deformation displacement is within 0 to 1 mm, it is set as no risk, the relative average deformation displacement is within 1 to 2 mm, it is set as low risk, the relative average deformation displacement is within 2 to 4 mm, it is set as medium risk, and the relative average deformation displacement is above 4 mm, it is set as high risk; for the deformation displacement interval in which the deformation point cloud accounts for less than or equal to 25%, it is set as no risk.

[0040] The beneficial effects of the present invention are as follows: the present invention collects point cloud data and inclination information of the roof through laser radar and inclination sensor, performs point cloud segmentation after format conversion, frame overlap, coordinate angle transformation, ICP registration and interference point filtering preprocessing, obtains the deformation area and deformation amount of the roof through deformation monitoring and deformation analysis of the working face roof, and finally performs cluster analysis on the deformation point cloud based on the DBSCAN density clustering algorithm, eliminates interference point clouds, and obtains the final deformation point cloud. The scanned multi-period point clouds are processed as above, and compared and calculated, and the calculated relative average deformation displacement and the deformation point cloud proportion of point clouds with different displacement amounts are used as the standard for deformation risk level assessment, and the deformation safety risk level of the mine working face roof is assessed, which provides effective technical guarantee for mining. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] Figure 1 is a flow chart of a mine working face roof deformation monitoring method based on laser radar provided in an embodiment of the present invention;

[0042] Figure 2 Schematic diagram of deformation monitoring device installation;

[0043] Figure 3 This is a reference diagram for judging the nearest triangle.

[0044] Figure 4 It is a schematic diagram of the deformation point marking of the top plate of the working face;

[0045] Figure 5 This is a schematic diagram of deformation density clustering point cloud cluster based on DBSCAN;

[0046] Figure 6 is a schematic diagram of the deformed point cloud cluster after removing the interference point cloud cluster;

[0047] Figure 7 This is a reference diagram of point cloud proportions with different deformation displacements. DETAILED DESCRIPTION

[0048] In order to make the purpose, technical solution and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0049] The technical solution of the present invention is to perform point cloud preprocessing, triangular mesh division, triangle center point generation, point-surface clustering calculation, deformation point cloud marking, deformation area visualization, and finally cluster analysis of the deformation point cloud based on the DBSCAN density clustering algorithm, eliminate interference point clouds, and obtain the final deformation point cloud. The scanned multi-period point clouds are processed as above, and compared and calculated. The calculated relative average displacement and the proportion of point clouds with different displacements are used as the standard for deformation risk level assessment, and the deformation safety risk level of the mine working face roof is assessed, which provides effective technical guarantee for mining. In order to illustrate the technical solution described in the present invention, it is described below through specific embodiments.

[0050] like Figure 1 As shown, the mine working face roof deformation monitoring method based on laser radar provided in this embodiment includes the following steps:

[0051] Step S1, laser radar data collection: a deformation monitoring device including a laser radar and an inclination sensor is arranged inside the mine, and then time-difference laser radar scanning and data collection are performed to obtain the original point cloud data and corresponding inclination information of the mine working face.

[0052] The method of this embodiment is used to carry out roof quality inspection at the working face of a mine. The deformation monitoring device used includes a laser radar and an inclination sensor. In addition, a visible light camera, a floodlight, etc. can be set. The deformation monitoring device is placed in the mine as follows: Figure 2 As shown in the figure, the tripod and the gimbal are used as supports to stably place the laser radar, visible light camera, and floodlight on the complex tunnel surface. The floodlight provides a bright field of view for the visible light camera.

[0053] After the device is installed, time-difference laser radar scanning and data collection are carried out. Time-difference scanning is scanning at regular intervals, and one period of data is obtained after each scan.

[0054] Step S2, point cloud data preprocessing: After the point cloud data is format converted and frame stacked, the entire point cloud data coordinates are calibrated using the inclination information to achieve the point cloud data coordinate angle transformation, and then preprocessed through ICP alignment and interference point filtering, and finally point cloud segmentation is performed.

[0055] In this step, the format of the collected point cloud data is first converted using the ROS operating system under Linux, and then the converted point cloud data is further processed by stacking, segmentation, registration and other dangerous rock area detection algorithms. The process of this step is as follows:

[0056] S21. Use the Linux ROS operating system to convert the original point cloud data in LVX format into point cloud data in PCD format. Use the frame stacking algorithm to overlay each frame to obtain a denser point cloud data that is closer to the actual point cloud data.

[0057] The original point cloud data collected by the lidar is in LVX format, which is converted into a point cloud in PCD format on the Linux ROS operating system using a virtual machine. Then, each frame is superimposed using a frame stacking algorithm to present a denser point cloud that is closer to the actual point cloud.

[0058] S22, using the PCL algorithm to transform the coordinate angle of the point cloud after the frame overlap according to the inclination information.

[0059] Since the point cloud data collected by the radar is saved in the form of coordinates, and the ground of the tunnel in the mine is very uneven, the laser radar needs to be supported by a tripod for scanning, and the upward angle of the laser radar needs to be adjusted to better scan the entire roof. All coordinates of the scanned point cloud will have a certain angle of coordinate offset, so it is necessary to use the acquired inclination information to calibrate the entire point cloud coordinates. After calibration, find the precise coordinate values ​​of the working face roof and the tunnel wall. This step uses the PCL algorithm to transform the coordinate angle of the point cloud after the frame is overlapped.

[0060] S23, preprocessing through ICP registration and interference point filtering.

[0061] S24, using the point cloud after the coordinate angle transformation, the point cloud data of the entire tunnel roof is segmented into XYZ coordinate point cloud, the process is as follows: first, by calculating the length, width and height of the tunnel, and then segmenting the XY coordinates of the point cloud according to the length and width in the point cloud data, and then segmenting the Z coordinates of the point cloud according to the height of the roof, and finally obtaining the point cloud data containing only the tunnel roof. This step realizes the accurate segmentation of the roof according to the set coordinate range.

[0062] Step S3, working face roof deformation monitoring: Use Delaunay triangulation to construct a triangular mesh model for the scanned two-phase point cloud data, and extract the center point of each triangular face of the second-phase triangular mesh model to form a second-phase center point point cloud. Calculate the distance between the second-phase center point point cloud and the first-phase triangular mesh model to obtain the point-to-surface distance.

[0063] This step uses the nearest triangle face determination algorithm to calculate the point-face distance, which is used as the basis for deformation and deformation risk assessment. The specific process of this step is as follows:

[0064] S31. Constructing a triangular mesh model:

[0065] The normals of the two-phase roof point cloud data after segmentation are calculated, and the triangular mesh model is constructed using the Delaunay triangular mesh surface reconstruction algorithm.

[0066] S32, extract the center point of each triangular face of the second-phase triangular mesh model to form a second-phase center point cloud:

[0067] Based on the second phase point cloud, in order to increase the point cloud density and improve the accuracy of the subsequent point cloud-triangular mesh model calculation, it is necessary to extract and calculate the center points of the triangular faces.

[0068]

[0069] Among them, x1, y1, z1; x2, y2, z2; x3, y3, z3 are the three-axis coordinate values ​​of the three vertices of the triangle surface, and x, y, z are the three-axis coordinates of the center point of the triangle surface.

[0070] S33, determine the nearest triangle:

[0071] For the center point P of a triangle, if the sum of the areas of the three triangles formed by the three sides of a triangle of the first phase triangular mesh model and the area difference of this triangle are less than a sufficiently small threshold set, then the current triangle is determined to be the nearest triangle to the center point P.

[0072] like Figure 3 As shown, for the two triangles ABD and BCD in the first triangular mesh model, first calculate the area S1 of triangle ABD, the area S2 of triangle BCD, the area S3 of triangle ABP, the area S4 of triangle ADP, and the area S5 of triangle BDP in the first tetrahedron; the area S6 of triangle BCP and the area S8 of triangle DCP in the second tetrahedron. By setting a sufficiently small threshold ξ, if the calculated result is that the sum of S3, S4, and S5 is very close to S1, that is, |(S3+S4+S5)-S1|<ξ, it can be determined that the triangle face closest to point P is ABD. If the calculated result is that the sum of S6, S5, and S8 is very close to S2, that is, |(S6+S5+S8)-S2|<ξ, it can be determined that the triangle face closest to point P is triangle BCD.

[0073] S33. Calculate the distance d between the center point P and the nearest triangle:

[0074] After determining the nearest corresponding triangle, it is necessary to calculate the distance between the center point and the triangle. The triangle fitting equation is as follows:

[0075]

[0076] Where A, B, C, and E are the coefficients of the nearest triangular surface fitting equation; for the coordinates of the current center point P (x0, y0, z0), the distance formula from point to surface is used for calculation, and we can get:

[0077]

[0078] Finally, the distance between the center point of each triangle face of the second-phase triangular mesh model and the corresponding nearest triangle face of the first-phase mesh model is obtained, that is, the point-face distance.

[0079] Step S4, deformation analysis of the working face roof: extract the deformation point cloud according to the point-to-surface distance to form a deformation point cloud cluster, and then use the DBSCAN density clustering algorithm to perform regional density clustering on the deformation point cloud cluster, identify and mark the area where the mine working face roof is deformed, and finally perform deformation risk level assessment.

[0080] The process of this step is as follows:

[0081] S41. The point-surface distance corresponds to the deformation variable of the center point, and different deformation variable ranges are marked with different colors.

[0082] If the roof of the working face is deformed or rock falls off, the two phases of point clouds can be processed by the above method, and finally the point-surface distance can be calculated to monitor the roof deformation. The larger the deformation, the larger the calculated result value. Figure 4 As shown in the figure, different colors are used to mark different deformation ranges. Green points are within 5mm, blue points are within 5-10mm, yellow points are within 10-15mm, and red points are above 15mm.

[0083] S42, forming a deformed point cloud cluster for the extracted deformed point cloud, and then using the DBSCAN density clustering algorithm to perform regional density clustering on the deformed point cloud cluster to divide the deformed point cloud clusters in multiple different regions.

[0084] By calculating the distance between the center point and the nearest triangle, the deformed point cloud is extracted to form a deformed point cloud cluster. Then, the DBSCAN density clustering algorithm is used to perform regional density clustering on the deformed point cloud cluster, thereby dividing the deformed point cloud clusters in different regions. The effect is as follows: Figure 5 As shown (here only the point cloud group with a deformation of 10 mm is considered).

[0085] S43, density filtering is performed on the deformation point cloud clusters to filter out interference point clouds with a small number of cloud clusters and sparse density, and the remaining deformation point cloud clusters are identified as point clouds in the deformation area of ​​the mine working face roof.

[0086] When a certain area of ​​the roof of the working face is deformed, its stress will change accordingly, and the surrounding area will also be affected. Therefore, the deformation of the roof is basically concentrated in a certain area and basically does not present a discrete distribution. Therefore, it is necessary to process these deformation point cloud clusters through density filtering. The point clouds with a small number of deformation point cloud clusters and sparse density can be considered as interference point clouds and can be excluded. The deformation point cloud clusters that are finally retained are considered to be the point clouds of the deformation area of ​​the roof of the working face of the mine. The effect is as follows: Figure 6 shown.

[0087] S44. Calculate the average deformation displacement and point cloud ratio in each deformation displacement interval as risk assessment grade standards for risk assessment.

[0088] According to the processing of the two-phase point cloud scans, the relative average displacement of the two-phase point cloud after filtering and the point cloud proportion are calculated as the standard for risk assessment level. In this step, the deformation displacement interval is defined as four intervals, namely 0-5mm, 5-10mm, 10-15mm, and above 15mm, and marked with different colors (green, blue, yellow, and red). As an example, the proportion of points in each deformation displacement interval to the entire point cloud is as follows: Figure 7 As shown, the average deformation displacements in each deformation displacement interval are calculated to be 2.3 mm, 6.9 mm, 11.6 mm, and 27 mm, respectively.

[0089] Referring to the above method, the point clouds of the roof of the multi-period working face are monitored and analyzed, and the relative average deformation displacement is calculated after each comparison. The deformation risk assessment is divided according to the relative average deformation displacement in the deformation displacement interval. For example, in the four deformation displacement intervals, for the deformation point cloud accounting for more than 25%, if the relative average deformation displacement is within 0 to 1mm, it is set as no risk, the relative average deformation displacement is within 1 to 2mm, it is set as low risk, the relative average deformation displacement is within 2 to 4mm, it is set as medium risk, and the relative average deformation displacement is above 4mm, it is set as high risk; for the deformation point cloud accounting for less than or equal to 25%, it is set as no risk. The specific results are shown in Table 1 below.

[0090] Table 1 Results of deformation analysis of roof monitoring at the mine working face

[0091]

[0092] This scientific assessment provides solid data support for the mine's safe production management and decision-making.

[0093] In summary, the present invention aims at the demand for accurate monitoring of the roof deformation of the mine working face, and designs a set of mine working face roof deformation monitoring methods based on laser radar, which is specially used for the complex environment of the mine working face. The method of the present invention adopts time difference scanning to obtain two-phase point cloud data of the mine working face. After pre-processing such as angle transformation and ICP registration, the laser radar posture is calibrated in real time by combining a high-precision three-axis tilt sensor to accurately extract the roof point cloud morphology, and then the Delaunay triangulation method is used to process the two-phase point cloud, and the point-surface monitoring algorithm is innovatively integrated to calculate the distance between the center point of the second phase and the triangular surface of the first phase, so as to improve the deformation monitoring accuracy and quickly and accurately quantify the deformation data. Finally, a deformation point cloud density clustering filtering algorithm is proposed based on the DBSCAN clustering algorithm to identify and mark the deformation area, and on this basis, a risk level assessment is made to realize the intelligent monitoring of roof deformation. The results of the field study show that the monitoring method of the present invention has a monitoring accuracy range of 1mm to 5mm, which is significantly improved compared with the traditional monitoring method. In addition, the system can realize the visualization of deformation areas, providing strong technical support for mine safety production and further enhancing the safety and reliability of mine operations.

[0094] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions and improvements made within the spirit and principles of the present invention should be included in the protection scope of the present invention.

Claims

1. A method for monitoring roof deformation of a mine working face based on laser radar, characterized in that: The monitoring method comprises the following steps: Step S1: LiDAR data collection Deformation monitoring devices including laser radar and inclination sensors are arranged inside the mine, and then time-difference laser radar scanning and data collection are carried out to obtain the original point cloud data and corresponding inclination information of the mine working face; Step S2: Point cloud data preprocessing After the point cloud data is format converted and frame stacked, the entire point cloud data coordinates are calibrated using the inclination information to achieve the point cloud data coordinate angle transformation, and then pre-processed through ICP registration and interference point filtering, and finally point cloud segmentation is performed; Step S3: Working surface roof deformation monitoring The Delaunay triangulation is used to construct a triangular mesh model for the scanned point cloud data of the two phases, and the center point of each triangular face of the second phase triangular mesh model is extracted to form the second phase center point cloud. The distance between the second phase center point cloud and the first phase triangular mesh model is calculated to obtain the point-surface distance. Step S4: deformation analysis of the top plate of the working surface The deformation point cloud is extracted according to the point-surface distance to form a deformation point cloud cluster. The DBSCAN density clustering algorithm is then used to perform regional density clustering on the deformation point cloud cluster to identify and mark the area where the roof of the mine working face has deformed. Finally, the deformation risk level is evaluated.

2. The method for monitoring roof deformation of a mine working face based on laser radar as claimed in claim 1, characterized in that: The specific process of step S2 is as follows: S21. Use the Linux ROS operating system to convert the original point cloud data in LVX format into point cloud data in PCD format, and use the frame stacking algorithm to overlay each frame to obtain a denser and more realistic point cloud data. S22, using the PCL algorithm to transform the coordinate angle of the point cloud after the frame overlap according to the inclination information; S23, preprocessing by ICP registration and interference point filtering; S24. Using the point cloud after coordinate angle transformation, the point cloud data of the entire tunnel roof is segmented into XYZ coordinate point cloud. The process is as follows: first calculate the length, width, and height of the tunnel, then segment the XY coordinates of the point cloud according to the length and width in the point cloud data, and then segment the Z coordinates of the point cloud according to the height of the roof, and finally obtain the point cloud data containing only the tunnel roof.

3. The method for monitoring roof deformation of a mine working face based on laser radar as claimed in claim 2, characterized in that: The specific process of step S3 is as follows: S31. Constructing triangular mesh model The normals of the two-phase roof point cloud data after segmentation are calculated, and the triangular mesh model is constructed using the Delaunay triangular mesh surface reconstruction algorithm; S32, extract the center point of each triangular face of the second phase triangular mesh model to form the second phase center point cloud Among them, x1, y1, z1; x2, y2, z2; x3, y3, z3 are the three-axis coordinate values ​​of the three vertices of the triangle surface, and x, y, z are the three-axis coordinates of the center point of the triangle surface; S33, determine the nearest triangle For the center point P of a triangle, if the difference between the sum of the areas of the three triangles formed by the three sides of a triangle of the first phase triangular mesh model and the area of ​​this triangle is less than a sufficiently small threshold, then the current triangle is determined to be the nearest triangle to the center point P. S33, calculate the distance d between the center point P and the nearest triangle Among them, x0, y0, z0 are the coordinates of the current center point P, and A, B, C, E are the fitting equations of the nearest triangle. The coefficients of ; Finally, the distance between the center point of each triangle face of the second-phase triangular mesh model and the corresponding nearest triangle face of the first-phase mesh model is obtained, that is, the point-face distance.

4. The method for monitoring roof deformation of a mine working face based on laser radar as claimed in claim 3, characterized in that: The specific process of step S4 is as follows: S41, the point-surface distance corresponds to the deformation of the center point, and different deformation ranges are marked with different colors; S42, forming a deformed point cloud cluster for the extracted deformed point cloud, and then using the DBSCAN density clustering algorithm to perform regional density clustering on the deformed point cloud cluster to divide a plurality of deformed point cloud clusters in different regions; S43, performing density filtering on the deformation point cloud clusters, filtering out interference point clouds with a small number of cloud clusters and sparse density, and identifying the remaining deformation point cloud clusters as point clouds in the deformation area of ​​the mine working face roof; S44. Calculate the average deformation displacement and point cloud ratio in each deformation displacement interval as risk assessment grade standards for risk assessment.

5. The method for monitoring roof deformation of a mine working face based on laser radar as claimed in claim 4, characterized in that: In step S44, for the deformation displacement interval in which the deformation point cloud accounts for more than 25%, if the relative average deformation displacement is within 0 to 1 mm, it is set as no risk, if the relative average deformation displacement is within 1 to 2 mm, it is set as low risk, if the relative average deformation displacement is within 2 to 4 mm, it is set as medium risk, and if the relative average deformation displacement is above 4 mm, it is set as high risk; for the deformation displacement interval in which the deformation point cloud accounts for less than or equal to 25%, it is set as no risk.

Images