Tunneling cutting path planning method based on laser scanning

Abstract

The invention relates to the technical field of coal mine production, in particular to a tunneling cutting path planning method based on laser scanning, which is characterized in that modeling is performed based on tunneling roadway point cloud data generated by laser scanning, and accurate modeling and rapid reconstruction of a roadway three-dimensional model are realized by optimizing a point cloud data processing algorithm and improving data processing efficiency; an advanced path planning algorithm is applied, the real-time performance of path planning is improved, and the tunneling path can be rapidly adjusted; more environmental perception and intelligent optimization algorithms are introduced, potential risk factors in a roadway environment are comprehensively considered, a path optimization strategy is perfected, and the tunneling safety is improved; by means of the laser scanning technology, roadway environment rapid modeling and environment data obtaining are achieved, the tunneling path can be automatically adjusted in the complex and changeable tunneling environment, and intelligent self-adaptive path planning is achieved.

Description

Technical Field

[0001] The invention relates to the technical field of coal mine production, and in particular to a tunneling and cutting path planning method based on laser scanning. Background Art

[0002] Coal mining is an important and dangerous work in the coal mining production process. The key technologies of intelligent control of tunnel boring machines are mainly automatic control of tunnel section forming and control of tunneling direction. In the process of automatic control of tunnel section forming, how to automatically control the cantilever of the tunnel boring machine to drive the cutting head to perform cutting operations in the tunnel section operation and cut out the specified cross-sectional shape according to the designed path is the key to the technology. However, traditional tunneling path planning mainly relies on manual experience and simple calculations, and there are the following problems: due to reliance on manual experience, the accuracy of tunneling path planning is difficult to guarantee, which easily causes the tunnel boring machine to deviate from the predetermined trajectory during the tunneling process, affecting work efficiency and safety; manual path planning is slow and cannot adjust the tunneling path in real time, resulting in low tunneling efficiency; in complex tunnel environments, manual path planning cannot fully consider all potential dangerous factors, which easily leads to safety accidents such as tunnel boring machine collision and collapse. In recent years, the application of laser scanning technology in various fields has gradually become popular. Laser scanning technology can quickly and accurately obtain three-dimensional point cloud data of the environment and generate high-precision three-dimensional models. Laser scanning technology has been widely used in the fields of construction, geological exploration, and unmanned driving for environmental modeling and path planning. However, in the field of coal mining, the application of laser scanning technology is still in its infancy.

[0003] The existing solution for coal mine excavation path planning using laser scanning technology mainly includes the following steps: 1. Deployment of laser scanning equipment: Deploy laser scanning equipment at the excavation site to obtain three-dimensional point cloud data of the tunnel through laser scanning; 2. Point cloud data processing: Process the acquired point cloud data, including data denoising, de-discrete point processing, etc.; 3. Tunnel morphology restoration: Linear fitting of the tunnel morphology to calculate the tunnel centerline position and direction; 4. Path planning: According to the spatial position and heading of the tunnel boring machine, combined with the calculated tunnel centerline position and direction, the optimized cutting path is automatically selected. The existing technology has the following shortcomings in coal mine excavation path planning: 1. For models without drawings, with numerous details and dynamic changes, such as tunneling tunnels, existing planning and cutting technologies usually use simplified processing to model tunnels, or obtain simplified tunnel models through laser scanning, which cannot meet the development requirements of digital twin technology for the digital transformation of the coal mining industry; 2. Poor real-time performance of path planning: The existing path planning algorithm has poor real-time performance in complex and dynamic tunneling environments, making it difficult to quickly adjust the tunneling path; 3. Imperfect path optimization: The existing path optimization strategy is not perfect enough to fully consider all potential dangerous factors in the tunnel environment, affecting the safety of excavation; 4. Lack of intelligence: Most existing technologies rely on pre-set rules and parameters, lack intelligent adaptive capabilities, and are difficult to cope with complex and changing tunneling environments. The amount of geological exploration data is small, and it is impossible to generate an accurate three-dimensional model of the tunnel. Summary of the invention

[0004] In response to the above-mentioned defects and problems, the present invention provides a tunneling and cutting path planning method based on laser scanning. By generating a three-dimensional terrain model of the tunnel based on laser scanning technology and automatically planning the cutting path of the tunneling machine, efficient and safe tunneling operations are achieved, solving the problems of inaccurate path planning, low efficiency and insufficient safety in the coal mine tunneling process.

[0005] The solution adopted by the present invention to solve the technical problem is: a tunneling cutting path planning method based on laser scanning, comprising the following steps: Step 1: Obtain and process laser scanning data: First, obtain the three-dimensional scene point cloud data of the excavation working face through the laser scanning equipment, then eliminate the noise of the cloud data and remove duplicate points, and then perform coordinate system registration and sampling processing; Step 2: Generating a 3D terrain model: Perform Poisson reconstruction based on the preprocessed point cloud data to generate a 3D terrain model; Step 3: Intelligently plan the path: Based on the generated three-dimensional terrain model of the tunnel, apply the RRT algorithm to plan the full traversal excavation path; Step 4: Path optimization and obstacle avoidance strategy: Optimize the initially generated excavation path, use environmental perception technology to detect obstacles and dangerous areas in the tunnel in real time, apply obstacle avoidance algorithms, and automatically adjust the excavation path to avoid collisions and dangers.

[0006] Furthermore, the specific operation steps of step 1 are: S11. Deploy laser scanning equipment to obtain the three-dimensional point cloud data of the tunnel in real time during the excavation process. The laser scanner starts from the starting position (0,0,0) and moves along the direction of the working face to obtain the three-dimensional scene point cloud data of the excavation working face; S12, cloud data denoising, filtering out abnormal noise caused by inherent limitations of sensors or fluctuations in external conditions; S13. Through the optimized point cloud data processing algorithm, duplicate points are removed, coordinate systems are merged, and sampling is performed on the acquired point cloud data to obtain point cloud data that meets the requirements of three-dimensional terrain reconstruction.

[0007] Furthermore, the specific steps of the three-dimensional terrain model in step 2 are: S21, calculating the original point cloud normal vector, voxelizing, constructing the weight matrix, solving the Poisson equation and generating a triangular mesh according to the preprocessed point cloud data, thereby generating a high-precision three-dimensional terrain model of the roadway; S22. In the process of generating the 3D model, an adaptive mesh refinement algorithm is applied to improve the detail expression and accuracy of the model to ensure that the model can accurately reflect the actual situation of the laneway; S23. Reduce redundant data of the three-dimensional model through model optimization technology and improve the calculation efficiency and storage efficiency of the model.

[0008] Furthermore, the method of planning the path using the RRT algorithm is: S31. First, read the generated three-dimensional model of the lane, and intelligently mark the obstacles in the lane model; S32, secondly, define the lane boundary and working space; S33, then, initialize the RRT tree and set the starting point and target point; S34, then, the RRT tree is continuously expanded through random sampling, nearest neighbor search, expansion and collision detection until it approaches the target; S35. Finally, the path is extracted and smoothed, and its feasibility is verified by simulation.

[0009] Furthermore, in the path planning process, it is necessary to comprehensively consider the terrain characteristics of the tunnel, the kinematic constraints and dynamic constraints of the tunnel boring machine to generate a reasonable tunneling path; use dynamic programming technology to adjust and optimize the tunneling path in real time based on the environmental data obtained in real time during the tunneling process to ensure the continuity and safety of the tunneling.

[0010] Furthermore, in step 4, environmental perception can be performed through a laser scanner to detect obstacles and dangerous areas in the lane in real time. When an obstacle is detected ahead, a local path planning algorithm is used to calculate a path to bypass the obstacle. If local adjustments cannot avoid the obstacle, or an impassable dangerous area is encountered, the entire path needs to be re-planned by executing the fast random tree algorithm again.

[0011] Beneficial effects of the present invention: The present invention performs modeling based on the point cloud data of the tunnel generated by laser scanning, improves the data processing efficiency by optimizing the point cloud data processing algorithm, and realizes the accurate modeling and rapid reconstruction of the tunnel three-dimensional model; applies advanced path planning algorithms to improve the real-time performance of path planning, and can quickly adjust the tunneling path; introduces more environmental perception and intelligent optimization algorithms, comprehensively considers the potential dangerous factors in the tunnel environment, improves the path optimization strategy, and improves the safety of tunneling; through laser scanning technology, realizes rapid modeling of the tunnel environment and obtains environmental data, can autonomously adjust the tunneling path in a complex and changeable tunneling environment, and realizes intelligent adaptive path planning. The present invention realizes efficient, accurate, and safe tunneling path planning by applying laser scanning technology and intelligent path planning algorithms, significantly improves the efficiency and safety of tunneling work, and at the same time has the characteristics of intelligence and adaptability, and can adapt to complex and changeable tunneling environments. BRIEF DESCRIPTION OF THE DRAWINGS

[0012] Figure 1 It is a flow chart of the present invention. DETAILED DESCRIPTION

[0013] The present invention is further described below in conjunction with the accompanying drawings and embodiments.

[0014] See also Figure 1 The present invention provides a technical solution for a tunneling and cutting path planning method based on laser scanning: Embodiment 1: A tunneling and cutting path planning method based on laser scanning comprises the following steps: Step 1: Acquire and process laser scanning data: First, obtain the three-dimensional scene point cloud data of the excavation working face through the laser scanning equipment, then eliminate the noise and remove duplicate points of the cloud data, and then perform coordinate system alignment and sampling processing.

[0015] S11. Deploy laser scanning equipment to obtain real-time 3D point cloud data of the tunnel during the excavation process. The laser scanner starts from the starting position (0,0,0) and moves along the direction of the working face to obtain 3D scene point cloud data of the excavation working face.

[0016] S12. Eliminate noise and remove duplicate points from the cloud data. Noise elimination means filtering out abnormal noise points caused by inherent limitations or fluctuations in external conditions, such as eliminating noise points in the cloud data caused by sensor hardware errors and environmental factors such as changes in illumination and temperature. Removing duplicate points means removing duplicate scanning points caused by multiple scans of the same position in the tunneling tunnel based on the denoised cloud data.

[0017] S13, coordinate system registration and sampling processing. Ensure that all points in the point cloud data are expressed in the same coordinate system, so as to merge the scanned data of each part of the tunnel into a complete tunnel working face scene model, and then perform sampling processing. Through the optimized point cloud data processing algorithm, the obtained point cloud data is subjected to duplicate point removal, coordinate system merging and sampling processing to obtain point cloud data that meets the requirements of three-dimensional terrain reconstruction, and ensure that the density of the tunnel point cloud data meets the requirements of subsequent processing.

[0018] Step 2: Generate a 3D terrain model: Perform Poisson reconstruction based on the preprocessed point cloud data to generate a 3D terrain model. The specific steps are as follows: S21. Based on the preprocessed point cloud data, the original point cloud normal vector is calculated, voxelized, the weight matrix is ​​constructed, the Poisson equation is solved, and a triangular mesh is generated, thereby generating a high-precision three-dimensional terrain model of the roadway.

[0019] Calculation of the original point cloud normal vector: Use the Principal Component Analysis (PCA) method to estimate the normal of each point. First, calculate the covariance matrix of the local neighborhood of the point cloud, and then find the eigenvalue and eigenvector of the matrix. The eigenvector corresponding to the smallest eigenvalue is the normal. Voxelization: The octree voxelization method is used to divide the three-dimensional space into cube grids of different resolutions. Each cube (voxel) contains a certain amount of point cloud data for subsequent surface reconstruction. Weight matrix construction: The weight matrix is ​​constructed using the Shepard weight method. The weight value depends on the distance from each point in the point cloud to the target point. The closer the distance, the greater the weight. Poisson equation solution: The Poisson equation is discretized using the finite difference method (FDM) and solved on the voxelized grid to reconstruct a smooth triangular mesh model.

[0020] S22. In the process of generating triangular meshes of the tunnel model, an adaptive mesh refinement algorithm is applied to improve the detail expression and accuracy of the model and ensure that the model can accurately reflect the actual situation of the tunnel. Adaptive mesh refinement algorithm: Based on the generated triangular mesh, the mesh density is dynamically adjusted according to the geometric features and curvature changes of the model. For areas with large curvature or obvious features, a denser mesh is used to improve the detail expression and accuracy of the model.

[0021] S23. Reduce redundant data of 3D models and improve the computational efficiency and storage efficiency of models through model optimization technology. Model optimization technology: Use simplification technology such as the Quadric Error Metric (QEM) algorithm to reduce the number of vertices and facets of triangular meshes while maintaining the geometric features of the model to improve the computational efficiency and storage efficiency of the model.

[0022] Step three, intelligently plan the path: Based on the generated three-dimensional terrain model of the tunnel, apply the RRT algorithm to perform full traversal excavation path planning.

[0023] The RRT (Rapid Random Tree) algorithm can easily handle scenes with obstacles and differential motion constraints. First, read the generated 3D model of the tunnel and intelligently identify obstacles in the tunnel model, such as rock protrusions; second, define the tunnel boundary and workspace; then, initialize the RRT tree and set the starting point and target point; then, continuously expand the RRT tree through random sampling, nearest neighbor search, expansion and collision detection until it approaches the target; finally, extract and smooth the path, and simulate to verify its feasibility.

[0024] In the process of path planning, it is necessary to comprehensively consider the terrain characteristics of the tunnel, the kinematic constraints and dynamic constraints of the tunnel boring machine to generate a reasonable tunneling path. For example, the tunnel boring machine has a minimum turning radius, and the curvature radius at the turning point must be greater than or equal to this value when planning the path. In addition, the starting and braking capabilities of the tunnel boring machine are limited, and path planning should avoid sharp acceleration or deceleration. Dynamic programming technology is used to adjust and optimize the tunneling path in real time based on the environmental data obtained in real time during the tunneling process to ensure the continuity and safety of the tunneling.

[0025] Step 4: Path optimization and obstacle avoidance strategy: Optimize the initially generated excavation path, use environmental perception technology to detect obstacles and dangerous areas in the tunnel in real time, apply obstacle avoidance algorithms, and automatically adjust the excavation path to avoid collisions and dangers.

[0026] The initially generated excavation path is optimized, including path smoothing, path length minimization, etc., to ensure the stability and economy of the path. In addition, the roadheader has angle and maximum speed restrictions when working on steep slopes or other different tunnel conditions, and these parameters need to be combined for optimization during path optimization.

[0027] The laser scanner can sense the environment and detect obstacles and dangerous areas in the tunnel in real time. Then, the obstacle avoidance algorithm is applied to automatically adjust the excavation path to avoid collisions and dangers. For example, when an obstacle is detected ahead, the local path planning algorithm is used to calculate the path to bypass the obstacle; if the local adjustment cannot avoid the obstacle, or an inaccessible dangerous area is encountered, the fast random tree algorithm needs to be executed again to replan the entire path.

[0028] Comprehensively consider possible emergencies that may occur during the excavation process, such as tunnel collapse, equipment failure, etc., dynamically adjust and optimize the excavation path based on the real-time environmental data obtained during the excavation process, formulate emergency response strategies, and ensure the safety and continuity of the excavation work.

[0029] The above description is only a preferred embodiment of the present invention and does not limit the present invention. Any modifications, equivalent substitutions and improvements made within the spirit and principle of the present invention should be included in the protection scope of the present invention.

Claims

1. A tunneling and cutting path planning method based on laser scanning, characterized in that: The following steps are involved: Step 1: Obtain and process laser scanning data: First, obtain the three-dimensional scene point cloud data of the excavation working face through the laser scanning equipment, then eliminate the noise of the cloud data and remove duplicate points, and then perform coordinate system registration and sampling processing; Step 2: Generating a 3D terrain model: Perform Poisson reconstruction based on the preprocessed point cloud data to generate a 3D terrain model; Step 3: Intelligently plan the path: Based on the generated three-dimensional terrain model of the tunnel, apply the RRT algorithm to plan the full traversal excavation path; Step 4: Path optimization and obstacle avoidance strategy: Optimize the initially generated excavation path, use environmental perception technology to detect obstacles and dangerous areas in the tunnel in real time, apply obstacle avoidance algorithms, and automatically adjust the excavation path to avoid collisions and dangers.

2. The excavation cutting path planning method based on laser scanning according to claim 1 is characterized in that: The specific operation steps of step 1 are: S11. Deploy laser scanning equipment to obtain the three-dimensional point cloud data of the tunnel in real time during the excavation process. The laser scanner starts from the starting position (0,0,0) and moves along the direction of the working face to obtain the three-dimensional scene point cloud data of the excavation working face; S12, cloud data denoising, filtering out abnormal noise caused by inherent limitations of sensors or fluctuations in external conditions; S13. Through the optimized point cloud data processing algorithm, duplicate points are removed, coordinate systems are merged, and sampling is performed on the acquired point cloud data to obtain point cloud data that meets the requirements of three-dimensional terrain reconstruction.

3. The excavation cutting path planning method based on laser scanning according to claim 1 is characterized in that: The specific steps of the three-dimensional terrain model in step 2 are: S21, calculating the original point cloud normal vector, voxelizing, constructing the weight matrix, solving the Poisson equation and generating a triangular mesh according to the preprocessed point cloud data, thereby generating a high-precision three-dimensional terrain model of the roadway; S22. In the process of generating the 3D model, an adaptive mesh refinement algorithm is applied to improve the detail expression and accuracy of the model to ensure that the model can accurately reflect the actual situation of the laneway; S23. Reduce redundant data of the three-dimensional model through model optimization technology and improve the calculation efficiency and storage efficiency of the model.

4. The excavation cutting path planning method based on laser scanning according to claim 1 is characterized in that: The method of planning a path using the RRT algorithm is: S31. First, read the generated three-dimensional model of the lane, and intelligently mark the obstacles in the lane model; S32, secondly, define the lane boundary and working space; S33, then, initialize the RRT tree and set the starting point and target point; S34, then, the RRT tree is continuously expanded through random sampling, nearest neighbor search, expansion and collision detection until it approaches the target; S35. Finally, the path is extracted and smoothed, and its feasibility is verified by simulation.

5. The excavation cutting path planning method based on laser scanning according to claim 1 or 4, characterized in that: During the path planning process, it is necessary to comprehensively consider the terrain characteristics of the tunnel, the kinematic constraints and dynamic constraints of the tunnel boring machine to generate a reasonable tunneling path; dynamic planning technology is used to adjust and optimize the tunneling path in real time based on the environmental data obtained in real time during the tunneling process to ensure the continuity and safety of the tunneling.

6. The excavation cutting path planning method based on laser scanning according to claim 1 is characterized in that: In step 4, the laser scanner can be used to sense the environment and detect obstacles and dangerous areas in the lane in real time. When an obstacle is detected ahead, a local path planning algorithm is used to calculate a path to bypass the obstacle. If the local adjustment cannot avoid obstacles or encounters an impassable dangerous area, the entire path needs to be replanned by executing the fast random tree algorithm again.

Images