CN-121330192-B - Construction method of mine three-dimensional model

CN121330192BCN 121330192 BCN121330192 BCN 121330192BCN-121330192-B

Abstract

The invention discloses a method for constructing a three-dimensional model of a mine, and relates to the technical field of geological exploration. The method comprises the steps of obtaining exploration comprehensive data of a mine to be modeled, scanning a roadway of the mine to be modeled to obtain a plurality of point cloud data, sequentially carrying out noise point detection and denoising on the point cloud data to obtain denoised point cloud data, dividing the denoised point cloud data into a plurality of grids in a three-dimensional space, marking each grid through grid characteristic vectors to obtain grid data, modeling through SKUA-GOCAD based on the grid data and the exploration comprehensive data to obtain a mine three-dimensional model of the mine to be modeled, and carrying out real-time updating on the mine three-dimensional model through underground small geological abnormal bodies and point cloud data after tunneling when a new drilling hole or roadway of the mine to be modeled is disclosed, wherein the underground small geological abnormal bodies are detected through a geophysical prospecting means comprising geological radars. The method improves the modeling precision of the mine three-dimensional model.

Inventors

XIAO BIN
LI SHENG
JIANG CHUNHAO
GUO DONGXU
FAN CHAOJUN
LUO MINGKUN
YANG ZHENHUA
DONG GUOJI
MEI KAIFENG
MO YU

Assignees

辽宁工程技术大学

Dates

Publication Date: 20260508
Application Date: 20251117

Claims (8)

1. The construction method of the mine three-dimensional model is characterized by comprising the following steps of: acquiring exploration comprehensive data of a mine to be modeled, wherein the exploration comprehensive data comprises horizon information, roadway information, geological abnormal body information, drilling information and drilling information; scanning a roadway of a mine to be modeled to obtain a plurality of point cloud data, wherein each point cloud data comprises three-dimensional coordinates and attribute information, and the attribute information comprises a top plate, a bottom plate and side walls; Sequentially carrying out noise detection and denoising on the point cloud data to obtain denoised point cloud data, wherein the noise comprises characters in a roadway, screws on the inner wall of the roadway and lamp tubes on the inner wall of the roadway; Dividing the denoised point cloud data into a plurality of grids in a three-dimensional space, and marking each grid through grid feature vectors to obtain grid data, wherein the grid feature vectors comprise three-dimensional coordinates, reflectivity and variance, and the variance is used for representing the density degree of points in one grid; Modeling is carried out through SKUA-GOCAD based on the grid data and the exploration comprehensive data, and a mine three-dimensional model of a mine to be modeled is obtained; When a newly added borehole or roadway of a mine to be molded is exposed, the mine three-dimensional model is updated in real time through underground small geological abnormal bodies and point cloud data after tunneling; the method comprises the steps of obtaining underground small geological abnormal bodies in a roadway tunneling process in a drilling geological radar mode, updating a mine three-dimensional model in real time through the small geological abnormal bodies, scanning a newly-generated roadway through a three-dimensional laser scanning technology after the tunneling process is finished to obtain a first-class las file, generating a first-class las file through a VoxelNet deep learning algorithm based on the first-class las file, converting the generated first-class las algorithm into a second-class las file, and guiding the second-class las file into SKUA-GOCAD to update the mine three-dimensional model in real time.
2. The method of claim 1, wherein the sequentially performing noise detection and denoising on the point cloud data to obtain denoised point cloud data, specifically comprises: Performing noise detection on the point cloud data through a deep learning model VoxelNet; And separating the detected noise points from the point cloud data to generate denoised point cloud data.
3. The method of claim 1, wherein the modeling by SKUA-GOCAD based on the grid data and the survey composite data results in a three-dimensional model of the mine to be modeled, comprising: constructing an attribute model based on the attribute information in the grid data; Constructing a horizon three-dimensional model based on the horizon information, constructing a roadway three-dimensional model based on the roadway information, constructing a geological anomaly three-dimensional model based on geological anomaly information, constructing a drilling three-dimensional model based on drilling information, and constructing a drilling three-dimensional model based on drilling information; Determining the horizon three-dimensional model, the roadway three-dimensional model, the geological abnormal body three-dimensional model, the drilling three-dimensional model and the drilling three-dimensional model as basic three-dimensional models; Synthesizing the relativity between the grid data and the exploration comprehensive data to generate stratum; And determining the attribute model, the basic three-dimensional model and the stratum body as the mine three-dimensional model.
4. A method according to claim 3, wherein said constructing an attribute model based on attribute information in said grid data, in particular comprises: correcting the grid data; converting the corrected grid data into labs format data; And importing the converted labs format data into SKUA-GOCAD, and establishing the attribute model.
5. The method of claim 4, wherein correcting the grid data comprises: In the modeling process of SKUA-GOCAD, discrete smooth difference values are adopted to establish stratum interfaces and fault interfaces of the roadway; Correcting the grid data through stratum sections and fault interfaces to simulate the spatial spreading form, position and contact relation of the structure, wherein the structure is different types of interfaces and comprises horizons and geological abnormal bodies.
6. The device for constructing the three-dimensional model of the mine is characterized by comprising the following components: The acquisition module is used for acquiring exploration comprehensive data of the mine to be modeled, wherein the exploration comprehensive data comprises horizon information, roadway information, geological abnormal body information, drilling information and drilling information; The system comprises a scanning module, a data processing module and a data processing module, wherein the scanning module is used for scanning a roadway of a mine to be modeled to obtain a plurality of point cloud data, and each point cloud data comprises three-dimensional coordinates and attribute information; The denoising module is used for sequentially carrying out noise point detection and denoising on the point cloud data to obtain denoised point cloud data, wherein the noise point comprises a person in a roadway, a screw on the inner wall of the roadway and a lamp tube on the inner wall of the roadway; The system comprises a dividing and marking module, a marking module and a processing module, wherein the dividing and marking module is used for dividing the denoised point cloud data into a plurality of grids in a three-dimensional space, and marking each grid through a grid characteristic vector to obtain grid data; the modeling module is used for modeling through SKUA-GOCAD based on the grid data and the exploration comprehensive data to obtain a mine three-dimensional model of a mine to be modeled; The updating module is used for updating the mine three-dimensional model in real time through underground small geological abnormal bodies and point cloud data after tunneling when newly added holes or tunnels of the mine to be molded are exposed; the method comprises the steps of obtaining underground small geological abnormal bodies in a roadway tunneling process in a drilling geological radar mode, updating a mine three-dimensional model in real time through the small geological abnormal bodies, scanning a newly-generated roadway through a three-dimensional laser scanning technology after the tunneling process is finished to obtain a first-class las file, generating a first-class las file through a VoxelNet deep learning algorithm based on the first-class las file, converting the generated first-class las algorithm into a second-class las file, and guiding the second-class las file into SKUA-GOCAD to update the mine three-dimensional model in real time.
7. A computer readable storage medium, characterized in that the storage medium stores a computer program which, when executed by a processor, implements the method according to any one of claims 1-5.
8. A computer device comprising a memory, a processor and a computer program stored on the memory and executable on the processor, the processor implementing the method of any one of claims 1 to 5 when executing the program.

Description

Construction method of mine three-dimensional model Technical Field The invention relates to the technical field of geological exploration, in particular to a method for constructing a three-dimensional model of a mine. Background The mine three-dimensional modeling is a core foundation for intelligent construction of the coal mine, multi-source data such as drilling, geophysical prospecting and the like are integrated, a high-precision mine three-dimensional model is constructed, the form, grade distribution and fault structure of a mine body can be visually presented, the accuracy of resource reserve estimation is remarkably improved, and the dynamic updating capability and modeling precision directly influence mine safety production and resource development decision. Traditional roadway measurement, such as Real-time dynamic carrier phase difference technology (Real-TIME KINEMATIC ‌,) measurement method, section method and total station measurement, relies on manual splicing section or interpolation in the modeling process, a model is a static result, when a newly added drill hole or roadway is exposed, the model needs to be modified again or even reconstructed again, model updating is difficult, and the modeling accuracy of a three-dimensional model of a mine is low. Disclosure of Invention Based on the above, it is necessary to provide a method for constructing a three-dimensional model of a mine. The method improves the modeling precision of the mine three-dimensional model. The invention adopts the following technical scheme: The invention provides a method for constructing a three-dimensional model of a mine, which comprises the following steps: Acquiring exploration comprehensive data of a mine to be modeled, wherein the exploration comprehensive data comprises horizon information, roadway information, geological abnormal body information, drilling information and drilling information; scanning a roadway of a mine to be modeled to obtain a plurality of point cloud data, wherein each point cloud data comprises three-dimensional coordinates and attribute information, and the attribute information comprises a top plate, a bottom plate and side walls; sequentially carrying out noise point detection and denoising on the point cloud data to obtain denoised point cloud data, wherein the noise point comprises characters in a roadway, screws on the inner wall of the roadway and lamp tubes on the inner wall of the roadway; dividing the denoised point cloud data into a plurality of grids in a three-dimensional space, and marking each grid through grid feature vectors to obtain grid data, wherein the grid feature vectors comprise three-dimensional coordinates, reflectivity and variance, and the variance is used for representing the density degree of points in one grid; modeling is carried out through SKUA-GOCAD based on the grid data and the exploration comprehensive data, and a mine three-dimensional model of a mine to be modeled is obtained; when a newly added borehole or roadway of the mine to be molded is exposed, the mine three-dimensional model is updated in real time through underground small geological abnormal bodies and point cloud data after tunneling, and the underground small geological abnormal bodies are detected through geophysical prospecting means comprising geological radars. Preferably, the method sequentially performs noise point detection and denoising on the point cloud data to obtain denoised point cloud data, and specifically includes: performing noise detection on the point cloud data through a deep learning model VoxelNet; and separating the detected noise points from the point cloud data to generate denoised point cloud data. Preferably, modeling is performed through SKUA-GOCAD based on grid data and exploration comprehensive data to obtain a mine three-dimensional model of a mine to be modeled, and the method specifically comprises the following steps: constructing an attribute model based on the attribute information in the grid data; constructing a horizon three-dimensional model based on horizon information, constructing a roadway three-dimensional model based on roadway information, constructing a geological anomaly three-dimensional model based on geological anomaly information, constructing a drilling three-dimensional model based on drilling information, and constructing a drilling three-dimensional model based on drilling information; determining a horizon three-dimensional model, a roadway three-dimensional model, a geological abnormal body three-dimensional model, a drilling three-dimensional model and a drilling three-dimensional model as basic three-dimensional models; Synthesizing the relativity between the grid data and the exploration comprehensive data to generate stratum; and determining the attribute model, the basic three-dimensional model and the stratum body as a mine three-dimensional model. Preferably, the attribute model is constructed based on attribute inf