Unsupervised learning method for feature extraction of tunnel point cloud and overbreakage detection in drill and blast method

Abstract

The present application belongs to the technical field of tunnel engineering digitized construction, and particularly relates to a method for extracting point cloud features and detecting overbreak and underbreak of a tunnel by drilling and blasting method without supervised learning. Firstly, three-dimensional laser scanning is used to obtain initial point cloud data of tunnel excavation, and point cloud thinning and registration are used for pretreatment. Then, a convex hull algorithm is used to calculate the actual excavation axis, and an excavation section with a thickness of 0.1m is intercepted at a fixed interval of 10m. The K-means clustering algorithm is used for noise reduction processing of the point cloud slices, and the convex hull vertices are screened to accurately extract the actual excavation contour line. Finally, the overbreak and underbreak of each convex hull is calculated to realize real-time quantitative analysis of the overbreak and underbreak of the tunnel. By fusing three-dimensional laser scanning, unsupervised learning and the convex hull algorithm, the present application realizes the leap of overbreak and underbreak detection of a tunnel from "discrete, lagging and manual" to "continuous, real-time and intelligent", and provides efficient, accurate and reliable digital technical support for the drilling and blasting method tunnel construction.

Description

TECHNICAL FIELD

[0001] The present application belongs to the technical field of tunnel engineering digital construction, and particularly relates to a point cloud feature extraction and overbreak and underbreak detection method for drill-and-blast tunneling based on unsupervised learning, which is suitable for real-time analysis of overbreak and underbreak in drill-and-blast construction and optimization of blasting parameters. BACKGROUND

[0002] With the continuous advancement of infrastructure construction in China, tunnel engineering plays an increasingly important role in the fields of transportation, water conservancy, etc. The quality of tunnel construction is directly related to the safety, service life and economic benefits of the project. Traditional tunnel measurement methods have low precision, low efficiency, and incomplete data, which cannot meet the requirements of modern tunnel construction quality control. Therefore, accurate and efficient real-time detection and repair of tunnel overbreak and underbreak are particularly important.

[0003] The traditional overbreak and underbreak detection method mainly has the following technical bottlenecks and discrete sampling defects: total station and section instrument need to be collected point by point manually, the data density is insufficient, usually < 50 points / section (Zhang Xiangdong, Zhang Xuedong. Application of laser section instrument in tunnel section detection [J]. Transportation technology, 2014, (05): 94-96.), it is difficult to construct a continuous excavation contour, and there is a 20-30cm detection blind area. Lack of timeliness: the existing method needs to be measured on site and processed offline, the data lag is up to 4-6 hours, which cannot meet the real-time demand of dynamic adjustment of blasting parameters (tunnel drill-and-blast cycle is usually < 24 hours). Low degree of automation: manual measurement error can reach ±3cm, and is significantly affected by the tunnel environment, such as dust, lighting, etc. According to statistics, the efficiency of manual detection is only 15-20 sections / work shift.

[0004] A tunnel overbreak and underbreak automatic calculation method based on Monte Carlo method (patent number 202411871094.X) is disclosed, which judges the positional relationship of random points with two convex hulls to judge the three-dimensional shape of the tunnel excavation surface, especially in rock tunnels or complex geological conditions, the value of random points is different, the calculated data will have slight changes, thereby affecting the calculation accuracy.

[0005] A tunnel overbreak and underbreak identification and volume calculation method based on laser point cloud (patent number 202411871097.3) is disclosed, for each region, the design contour point cloud and the actual excavation region point cloud constituting the region are constructed into polyhedrons, the polyhedrons are divided into quadrilaterals, and the volume of each quadrilateral is calculated, thereby calculating the volume of each overbreak and underbreak region. Using quadrilaterals to calculate the volume will make it difficult to calculate irregular regions, and it is difficult to avoid a certain loss of accuracy.

[0006] Disclosed is a tunnel over-under excavation detection method and device, electronic equipment and storage medium. The original three-dimensional point cloud of a tunnel to be detected is obtained. The original three-dimensional point cloud is preprocessed to obtain a preprocessed three-dimensional point cloud. The preprocessed three-dimensional point cloud is subjected to ground plane removal processing to obtain a target three-dimensional point cloud. A tunnel center axis in the target three-dimensional point cloud is extracted. The target three-dimensional point cloud is rotated and rectified based on related parameters of the tunnel center axis to obtain a rectified three-dimensional point cloud. The rectified three-dimensional point cloud is intercepted to obtain a tunnel cross section. The cross section is used to fit an ideal tunnel contour curve. Ideal design tunnel excavation surface information is obtained according to the ideal contour curve. Irrelevant point clouds are removed from the ideal design tunnel excavation surface information to obtain a measured tunnel three-dimensional point cloud. The ideal design tunnel excavation surface information and the measured tunnel three-dimensional point cloud are used to calculate the over-under excavation of each point in the point cloud by using the polar coordinate method. Although the "rectified point cloud" is mentioned, the specific method for extracting the actual axis and calculating the space transformation matrix is not disclosed.

[0007] Disclosed is a tunnel cross section feature point accurate extraction method based on laser point cloud measurement. The disadvantages of traditional threshold denoising are that it relies on manual experience and is easy to lose details. The present scheme uses dynamic threshold + generative model to avoid noise misjudgment and retain contour details, especially suitable for scenes with uneven tunnel point cloud density.

[0008] The above-mentioned patent only calculates the volume or local deviation, and does not establish a closed loop correlation with the construction parameter optimization. The traditional method directly assumes that the point cloud is aligned with the design axis, ignoring the systematic deviation existing in the construction, resulting in distorted over-under excavation calculation. The unsupervised clustering algorithm in point cloud denoising is not disclosed in the comparative documents 1-4; the DCGAN generative model distinguishes between noise and real features; the dynamic threshold formula solves the general denoising problem of sparse / dense point clouds. SUMMARY

[0009] The present application is based on three-dimensional scanning technology and unsupervised learning, and proposes a tunnel point cloud feature extraction and over-under excavation real-time detection method. First, three-dimensional laser scanning is used to obtain tunnel excavation initial point cloud data, and point cloud thinning and registration are used for preprocessing. Then, the convex hull algorithm is used to calculate the actual excavation axis, the excavation section with a thickness of 0.1m is intercepted at a fixed interval of 10m, the K-means clustering algorithm is used for denoising processing of the point cloud slice, and the convex hull vertices are screened to accurately extract the actual excavation contour line. Finally, the over-under excavation amount of each convex hull is calculated to realize real-time quantitative analysis of the tunnel over-under excavation situation.

[0010] The technical scheme of the present application is as follows:

[0011] The method comprises the following steps:

[0012] S1, obtaining tunnel initial point cloud data by three-dimensional laser scanning, and sequentially performing thinning and registration preprocessing;

[0013] S2, axis correction: extracting an actual excavation axis from the registration point cloud, calculating a spatial transformation matrix of the actual excavation axis and a design axis, and correcting systematic deviation;

[0014] S3, cross section slice extraction: intercepting a slice with a thickness of δ along the design axis direction at a fixed interval Δs, Δs=10m, δ=0.1m;

[0015] S4, point cloud noise reduction: using a K-means clustering algorithm on the slice point cloud, determining a cluster number N according to a contour coefficient, identifying and removing noise clusters through a density threshold: 3-5 for sparse point clouds and 10-20 for dense point clouds, and using a deep convolutional generative adversarial network (DCGAN) as an auxiliary noise reduction means to distinguish noise and real features through a generative model and avoid detail loss caused by threshold dependence;

[0016] S5, convex hull generation: using a Graham algorithm to screen convex hull vertices from the noise-reduced point cloud, and sequentially connecting the convex hull vertices to form an actual excavation contour line;

[0017] S6, overbreak and underbreak calculation: calculating the shortest distance di between the convex hull vertices and the design contour line, di>0 indicating overbreak and di<0 indicating underbreak;

[0018] S7, trend analysis: statistically analyzing overbreak and underbreak amounts of each cross section along the tunnel axis direction, evaluating overall construction quality, and outputting adjustment suggestions.

[0019] In the step S1, a three-dimensional laser scanner calculates a ranging observation value S of a target point from the scanner by emitting a laser pulse signal and determining a reflection time, and simultaneously records lateral and longitudinal scanning angle observation values α and θ of each laser pulse by controlling an encoder; in an internal coordinate system Pi of the instrument, an X axis is located in a lateral scanning plane, a Y axis is perpendicular to the X axis and is located in the lateral scanning plane, and a Z axis is orthogonal to the scanning plane; based on the coordinate system and the measurement data, three-dimensional coordinates (XS, YS, ZS) of a scanning point P can be obtained, and a calculation formula is as follows:

[0020] The original data is generally saved in the form of point clouds, and can be preprocessed by special point cloud data software, and the specific steps are as follows:

[0021] (1) Decimation: Due to the large amount of original point cloud data, the cost of calculation and storage increases, and the density of the point cloud can be reduced by different sampling methods while retaining the original point cloud features. Common methods include decimation algorithm based on uniform distribution random number and voxel filtering method;

[0022] (2) Registration: Merge multiple sets of point cloud data obtained by scanning and convert them to a unified coordinate system. Common methods include ICP and FPFH.

[0023] In step S2, the actual tunnel excavation axis is extracted from the registered point cloud, and the spatial transformation matrix between the design axis and the actual axis is calculated by the least squares method to correct the systematic deviation of the point cloud data.

[0024] In step S4, K-means is a clustering algorithm, and its main purpose is to divide the data set into N non-intersecting clusters. The sum of squared errors in the cluster is minimized as shown in equation (2), i.e. by iteratively optimizing the cluster center, the data points in the cluster are as close as possible and the difference between the clusters is as large as possible;

[0025] Where: J is the sum of squared errors in the cluster, Ci is the point set of the ith cluster, p is a data point in Ci, represents the squared Euclidean distance between the data point and the cluster center, q i is the ith cluster center; K is the number of clusters, and X is any data point in the data set;

[0026] The algorithm first randomly selects N data points as initial cluster centers, assigns each data point to the nearest initial cluster center according to the Euclidean distance as shown in equation (3), and then calculates the mean of all points in each cluster to update the position of the cluster center as shown in equation (4). The iteration is stopped when the change of the cluster center is less than a certain threshold or the preset number of iterations is reached.

[0027] Where: d is the Euclidean distance of each point to the cluster center, (x1, y1, z1) is the three-dimensional coordinates of each point, and (x0, y0, z0) is the three-dimensional coordinates of q.

[0028] After setting a reasonable threshold for the data points in the clusters divided by the K-means algorithm, it can be applied to point cloud denoising. The specific operation is as follows:

[0029] a) For each point cloud slice, use the contour coefficient method to determine its appropriate cluster number N, and determine the optimal clustering result of each cluster through the update iteration process of the K-means algorithm.

[0030] b) After clustering is completed, a density threshold is set to analyze the size and density of each cluster to identify noise clusters, and generally smaller clusters or clusters with lower density are more likely to be noise. In order to more accurately identify noise points, a reasonable density threshold can be set according to the actual situation of the initial point cloud: for sparse point clouds, the threshold is set to (3, 5); for dense point clouds, it is set to (10, 20), and a deep convolutional generative adversarial network (DCGAN) is used as an auxiliary noise reduction means to distinguish noise from real features through a generative model, avoiding detail loss caused by threshold dependence; when the number of points in a cluster is less than the set threshold, the points in the cluster are considered to be noise points;

[0031] c) The points in the cluster identified as noise are removed from the point cloud to obtain the denoised point cloud data; this process effectively removes noise points in the point cloud while preserving the main features of the point cloud, thereby achieving accurate feature extraction of the point cloud.

[0032] In step S6, overbreak and underbreak refer to the spatial relationship between the actual excavation profile and the design section: when the actual convex hull profile line exceeds the design section profile line, it is called overbreak; when the design section profile line envelopes the actual convex hull profile line, it is called underbreak; the core of overbreak and underbreak calculation is to compare the spatial position relationship between the actual convex hull and the design section, and to quantify the overbreak and underbreak amount; a virtual line is drawn from the design axis to the convex hull vertex, and intersects the tunnel design section profile line at a point, and the distance between the convex hull vertex and the point is the overbreak and underbreak amount, and the calculation formula is as follows:

[0033] where di is the overbreak and underbreak amount of a convex hull vertex i, xi is the horizontal coordinate of i, yi is the vertical coordinate of i, xh is the horizontal coordinate of the point on the design section closest to i, and yh is the vertical coordinate of the point.

[0034] The unsupervised learning drill-and-blast tunnel point cloud feature extraction and overbreak and underbreak detection method has the following significant technical effects:

[0035] High detection accuracy, significantly reduced error:

[0036] By combining the K-means clustering algorithm with the DCGAN generative noise reduction means, noise points are effectively identified and removed, overcoming the detail loss problem caused by threshold dependence in traditional methods; the convex hull algorithm is combined to accurately extract the actual excavation profile, and the systematic deviation is corrected through the spatial transformation matrix, greatly improving the calculation accuracy of the overbreak and underbreak amount. The embodiments show that, compared with the traditional method, the error of the present method is reduced by at least 11.89%, and the maximum reduction is 40.40%.

[0037] Strong real-time, supports dynamic optimization of construction:

[0038] The method can realize real-time quantitative analysis of overbreak and underbreak, greatly shorten the data processing lag time, meet the demand of dynamic adjustment of blasting parameters within the drilling and blasting cycle (usually < 24 hours), and effectively improve the construction efficiency and timeliness of quality control.

[0039] High degree of automation, reducing manual intervention:

[0040] Using unsupervised learning algorithm for point cloud denoising and feature extraction, avoiding the subjectivity and instability of setting threshold value by artificial experience in traditional methods, reducing the influence of environmental factors (such as dust, lighting) on measurement results, improving the objectivity and consistency of detection results.

[0041] Strong adaptability, suitable for complex geological conditions:

[0042] Through dynamic density threshold (3~5 for sparse point cloud, 10~20 for dense point cloud) and DCGAN auxiliary denoising, it can effectively handle the scene of uneven tunnel point cloud density and complex noise distribution, especially suitable for rock tunnels or engineering environments with variable geological conditions.

[0043] Comprehensive quantitative analysis, supporting construction decision:

[0044] Not only realizes accurate calculation of single-section overbreak and underbreak, but also can perform multi-section trend analysis along the tunnel axis to evaluate the overall construction quality, output specific suggestions such as drilling and blasting parameter optimization and support timing adjustment, forming a closed-loop quality control system of "measurement-analysis-feedback".

[0045] Good engineering application effect, improving economic benefits:

[0046] Through accurate identification of overbreak and underbreak distribution rules (such as arch top and side wall as key control areas), targeted repair and support are guided, reducing concrete overconsumption and rework risk, reducing construction cost, and improving the safety and durability of tunnel engineering.

[0047] In summary, the present application realizes the leap from "discrete, lagging, manual" to "continuous, real-time, intelligent" in tunnel overbreak and underbreak detection by integrating three-dimensional laser scanning, unsupervised learning and convex hull algorithm, providing efficient, accurate and reliable digital technical support for drilling and blasting method tunnel construction. BRIEF DESCRIPTION OF DRAWINGS

[0048] Figure 1 Point cloud processing and real-time overbreak and underbreak detection flowchart;

[0049] Figure 2 Three-dimensional laser scanner space coordinate schematic diagram;

[0050] Figure 3 Tunnel three-dimensional laser scanning point cloud model;

[0051] Figure 4 Tunnel cross-section point cloud slices;

[0052] Figure 5 Tunnel cross-section overbreak and underbreak histogram. DETAILED DESCRIPTION

[0053] The technical solutions of the present application will be described below in conjunction with the drawings and specific embodiments. Obviously, the described embodiments are part of the embodiments of the present application, rather than all the embodiments. EMBODIMENT

[0054] 1. Instrumentation

[0055] This project uses a three-dimensional laser scanner to collect tunnel point cloud data. According to the size of the tunnel cross-section and the performance index of the scanner, the stations are arranged in the longitudinal direction of the tunnel according to the equal-interval layout principle. The station spacing is determined to be 15 meters through theoretical calculation and field verification, ensuring that there is enough overlapping area between adjacent stations to facilitate subsequent point cloud registration.

[0056] 2. Point cloud preprocessing

[0057] Twelve scans of the tunnel were performed using a three-dimensional laser scanner to obtain 12 sets of initial point cloud data sets and spatial coordinates as shown in Figure 2 . First, the point cloud data was thinned by a downsampling algorithm based on uniformly distributed random numbers to reduce data redundancy. Then, the ICP algorithm was used to accurately register the multi-station scanning data to achieve spatial unification of the point cloud data. Due to the influence of construction equipment shielding, the point cloud data in the entrance and exit sections of the tunnel is significantly missing, so this study excludes this section and selects the 20-110m range with good point cloud integrity as the analysis object. Finally, a tunnel point cloud reference model is constructed based on the actual axis as shown in Figure 3 , providing a geometric reference for subsequent overbreak analysis. This preprocessing procedure effectively improves the geometric accuracy and structural integrity of the point cloud data, meeting the precision requirements of engineering analysis.

[0058] 3. Cross-section feature extraction and convex hull calculation

[0059] Based on preprocessing, a spatial transformation matrix is used for coordinate system conversion to make the point cloud perpendicular to the design axis direction. An equidistant cross-section analysis method is used, with an analysis cross-section set every 10m along the longitudinal direction of the tunnel, and the thickness of each cross-section slice set to 0.1m to ensure comprehensive coverage of the tunnel structure while ensuring uniformity and systematicness of the analysis. To further improve the quality of the point cloud data, a K-means noise reduction algorithm is used to denoise the point cloud slices, with a threshold of (5, 20) to eliminate measurement noise and outliers while ensuring that the effective information in the point cloud is retained to the maximum extent, completing the feature extraction of the cross-section point cloud slices and providing accurate data basis for convex hull calculation.

[0060] For each denoised section slice, the Graham convex hull algorithm is used to select the convex hull vertex. The number of convex hull vertices of each section fluctuates between 60 and 110. The main reason for this fluctuation is the spatial difference in overbreak and underbreak of different sections. The extracted convex hull vertices are connected in a clockwise direction to form a closed curve, which constitutes the actual excavation contour line at the tunnel section, as shown in the figure. Figure 4 The green line part is the tunnel design section contour line, and the red line part is the actual section convex hull contour line. The points on the red line are the convex hull vertices. By spatially comparing the two, the overbreak and underbreak distribution characteristics can be intuitively presented.

[0061] 4. Section overbreak and underbreak analysis

[0062] The allowable overbreak and underbreak of tunnels in Q / CR 9604-2015 "High-speed Railway Tunnel Engineering Construction Technical Specification" are shown in Table 1.

[0063] Table 1. Tunnel allowable overbreak and underbreak specifications

[0064]

[0065] Based on the convex hull vertices, the overbreak and underbreak of each section are calculated. If the value is greater than 0, it is overbreak, and if it is less than 0, it is underbreak. The overbreak and underbreak distribution of the ten sections is visualized, as shown in the figure. Figure 5 The results show that: 1) all sections show significant overbreak characteristics; 2) underbreak only occurs locally, and the values are generally small; 3) the overbreak serious area has a clear spatial distribution, mainly concentrated in the sidewall and vault.

[0066] Statistical analysis of the overbreak and underbreak of all points in each section is shown in Table 2. Analysis shows that there is a serious overbreak and underbreak in this tunnel section: from the extreme value distribution, the extreme points appear at the 20+0.1m and 110+0.1m sections, reaching the maximum overbreak of 0.473m and the maximum underbreak of 0.334m; from the statistical analysis of the proportion, the overbreak rate of each section exceeds 80%; the underbreak rate shows a fluctuating distribution of 0-20%; from the mean value distribution, the average overbreak and underbreak exceeds 0.11m. Compared with the traditional measurement method, the method proposed in this study greatly improves the detection speed and greatly reduces the detection error, with an error reduction of at least 11.89% and a maximum error reduction of 40.40%, indicating that the method proposed in this study is effective.

[0067] Table 2. Tunnel single-section overbreak and underbreak statistical indicators

[0068]

[0069] Since the tunnel arch and abutment are critical stress-bearing parts, over- or under-excavation can easily lead to lining deformation and other defects, and are difficult to repair. Therefore, the over-excavation value of these protruding points is analyzed separately. From a compliance perspective, the average over-excavation value of the ten sections does not exceed the allowable average linear over-excavation limit (15cm) for surrounding rock specified in the "Technical Specification for Construction of High-Speed ​​Railway Tunnels" (Q / CR 9604-2015). However, the maximum over-excavation value exceeds the standard more seriously (see Table 3 for details). The exceedance rate at the 70+0.1m section is as high as 17%, indicating a significant risk of rework at the arch and abutment.

[0070] The above analysis indicates that there are systemic deviations in the current construction quality control. Optimization and adjustments are needed for drilling and blasting parameters and the timing of temporary support.

[0071] Table 3 Statistical Indicators of Over-excavation in the Arch Section of Each Tunnel Section

[0072]

[0073] 5. Analysis of Over-excavation and Under-excavation Trends

[0074] A systematic analysis was conducted on the histogram distribution characteristics and statistical parameters of over- and under-excavation at various tunnel sections to derive the overall over- and under-excavation trend. The analysis is as follows: 1) The spatial distribution of the histogram shows a clear decreasing trend in over- and under-excavation along the tunnel excavation direction; 2) The average over- and under-excavation values ​​at each section generally decrease with tunnel depth; 3) The standard deviation of over- and under-excavation also shows a decreasing trend with increasing excavation distance, reflecting a gradual reduction in construction dispersion. In summary, as the tunnel project progresses, the construction team effectively improved the control accuracy of the excavation profile by dynamically optimizing drilling and blasting parameters and adjusting the support scheme in a timely manner. This significantly improved the similarity between the actual excavated section and the designed section, meaning that the overall over- and under-excavation situation gradually decreased, effectively improving the stability and reliability of construction quality.

[0075] 6. Construction Recommendations

[0076] Based on the analysis of the current over-excavation and under-excavation status of this tunnel project, the following improvement suggestions are proposed for subsequent construction:

[0077] (1) Optimize the design of drilling and blasting parameters. For the sidewalls and arch tops where over-excavation is concentrated, it is recommended to appropriately reduce the spacing between blast holes and reduce the charge accordingly. For under-excavated sections, a reverse adjustment strategy should be adopted.

[0078] (2) Strengthen zoning control and promote the precise blasting mode of "section zoning and charge segmentation", namely the dynamic adjustment mechanism of "one blast, one design" to adapt to the monitoring data of different sections of surrounding rock.

[0079] (3) Perfect the measurement monitoring system, measure the overbreak and underbreak during construction, and solve the problem early to reduce the subsequent rework.

[0080] (4) Develop a hierarchical repair scheme: for overbreak areas, the principle of "support first, then process" must be followed. If the overbreak is small, directly spray concrete to smooth out to the design contour line. If it is larger, set up a reinforced steel mesh or a steel arch support before spraying concrete in layers. For local underbreak areas, if the underbreak is small, use small machinery for precise trimming and combine anchor spray support for fine adjustment of the section. If the underbreak is large, use secondary blasting repair technology.

[0081] (5) Implement the quality responsibility system, link the section deviation rate with the performance of the construction team, and form a virtuous cycle of "measurement-construction-evaluation" to promote fine construction management.

Claims

1. A method for feature extraction of point cloud and overbreak detection in drill and blast tunneling with unsupervised learning, characterized in that, The method comprises the following steps: S1, obtaining tunnel initial point cloud data by three-dimensional laser scanning, and sequentially performing thinning and registration preprocessing; S2, axis correction: extracting an actual excavation axis from the registered point cloud, calculating a spatial transformation matrix of the actual excavation axis and a design axis, and correcting systematic deviation; S3, cross section slice extraction: intercepting a slice with a thickness of δ along the design axis direction at a fixed interval Δs, Δs=10m, δ=0.1m; S4, point cloud noise reduction: using a K-means clustering algorithm on the slice point cloud, determining a clustering number N according to a contour coefficient, identifying and removing noise clusters through a density threshold: 3-5 for sparse point clouds and 10-20 for dense point clouds, and using a deep convolutional generative adversarial network (DCGAN) as an auxiliary noise reduction means to distinguish noise and real features through a generative model and avoid detail loss caused by threshold dependence; S5, convex hull generation: using a Graham algorithm to screen convex hull vertices from the noise-reduced point cloud, sequentially connecting the convex hull vertices to form an actual excavation contour line; S6, overbreak and underbreak calculation: calculating the shortest distance di of the convex hull vertices and the design contour line, di>0 indicating overbreak and di<0 indicating underbreak; S7, trend analysis: statistically analyzing overbreak and underbreak amounts of each cross section along the tunnel axis direction, evaluating overall construction quality, and outputting adjustment suggestions. 2.The method of claim 1, wherein, In the step S1, a three-dimensional laser scanner calculates a ranging observation value S of a target point and the scanner by emitting a laser pulse signal and determining a reflection time, and simultaneously records lateral and longitudinal scanning angle observation values α and θ of each laser pulse by controlling an encoder; in an internal coordinate system Pi of the instrument, an X axis is located in a lateral scanning plane, a Y axis is perpendicular to the X axis and is located in the lateral scanning plane, and a Z axis is orthogonal to the scanning plane; based on the coordinate system and the measurement data, three-dimensional coordinates (XS, YS, ZS) of a scanning point P can be obtained, and a calculation formula is as follows: The general form of saving the original data is point cloud, which can be preprocessed by special point cloud data software. The specific steps are as follows: (1) Thinning: due to a large amount of original point cloud data, calculation and storage costs increase, and different sampling methods can be used to reduce the density of the point cloud while retaining original point cloud features, and the sampling methods include a thinning algorithm based on a uniform distribution random number and a voxel filtering method; (2) Registration: merging multiple groups of point cloud data obtained by scanning and converting the point cloud data into a unified coordinate system, and the method includes ICP and FPFH. 3.The method of claim 1, wherein, In the step S2, an actual excavation axis of a tunnel is extracted from the registered point cloud, a spatial transformation matrix between a design axis and the actual axis is calculated by using a least square method, and systematic deviation of the point cloud data is corrected. 4.The method of claim 1, wherein, In the step S4, K-means is a clustering algorithm, and the main purpose is to divide a data set into N mutually disjoint clusters, and a cluster within square error sum is minimized as shown in formula (2), that is, cluster centers are iteratively optimized, so that data points within the cluster are as close as possible and differences between clusters are as large as possible; where: J is the within-cluster sum of squares, Ci is the set of points of the ith cluster, p is a data point in Ci, denotes the squared Euclidean distance between a data point and a cluster center, q i is the ith cluster center; K is the number of clusters, X is any one data point in the data set; The algorithm first randomly selects N data points as initial cluster centers, allocates each data point to a cluster with the nearest initial cluster center according to the Euclidean distance as shown in formula (3), then calculates the mean of all points in each cluster to update the position of the cluster center as shown in formula (4), and the iteration is performed until the change of the cluster center is less than a certain set threshold or a preset iteration number is reached. where: d is the Euclidean distance of each point to the cluster center, (xl, yl, zl) is the three-dimensional coordinates of each point, and (x0, y0, z0) is the three-dimensional coordinates of q; The data points in the clusters divided by the K-means algorithm can be applied to point cloud denoising after setting a reasonable threshold, and the specific operation is as follows: For each point cloud slice, the appropriate cluster number N is determined using the silhouette coefficient method, and the optimal clustering result of each cluster is determined through the update iteration process of the K-means algorithm; After clustering, a density threshold is set to analyze the size and density of each cluster to identify noise clusters. Generally, smaller clusters or clusters with lower density are more likely to be noise. To more accurately identify noise points, a reasonable density threshold can be set according to the actual situation of the initial point cloud: for sparse point clouds, the threshold is set to (3, 5); for dense point clouds, it is set to (10, 20), and a deep convolutional generative adversarial network (DCGAN) is used as an auxiliary noise reduction method to distinguish noise from real features through a generative model, avoiding detail loss caused by threshold dependence; When the number of points in a cluster is less than the set threshold, the points in the cluster are considered to be noise points. The points in the clusters identified as noise are removed from the point cloud to obtain the denoised point cloud data. 5.The method of claim 1, wherein, In step S6, overbreak and underbreak refer to the spatial relationship between the actual excavation profile and the design section: when the actual convex hull profile line exceeds the design section profile line, it is called overbreak; When the design section profile line envelopes the actual convex hull profile line, it is called underbreak; The core of overbreak and underbreak calculation is to compare the spatial position relationship between the actual convex hull and the design section, and to quantify the overbreak and underbreak amount. A virtual line is drawn from the design axis to the convex hull vertex, and intersects the tunnel design section profile line at a point. The distance between the convex hull vertex and the point is the overbreak and underbreak amount, and the calculation formula is as follows: where di is the overbreak or underbreak of the convex hull vertex i, xi is the horizontal coordinate of i, yi is the vertical coordinate of i, xh is the horizontal coordinate of the point on the nearest design section to i, and yh is the vertical coordinate of the point on the nearest design section to i.

Images