A small opening underground cavity volume detection method and device

Abstract

The application discloses a small-opening underground cavity volume detection method and device, detects an underground cavity, and sends the laser radar to perform point cloud data collection; one set of acquired point cloud data is evenly grouped, the cavity volume is calculated respectively, and the average volume A and the standard deviation B are solved; if B / A is greater than or equal to a set precision threshold value, the number of point cloud data in each set of point cloud data is increased to be recalculated, and finally the best data scale in each set of point cloud data is obtained; the vertical position of the laser radar in the cavity is adjusted, and the newly acquired one set of point cloud data is grouped to calculate the cavity volume in the best data scale; the process is repeated until the average volume A obtained by solving is greater than the last time, the position of the cavity volume obtained by calculation is the cavity opening position, and the maximum cavity volume is taken as the final cavity volume; and the internal morphology of the whole cavity is generated by using any one set of point cloud data in one set of point cloud data at the position of the maximum cavity volume.

Description

Technical Field

[0001] This invention belongs to the field of road, especially airport road, detection technology, and specifically relates to a method and device for detecting the volume of small underground cavities. Background Technology

[0002] Airport runways and aprons bear immense loads from aircraft takeoffs and landings, ground traffic, and natural environmental factors. Simultaneously, the interplay of factors such as underground pipeline laying, groundwater flow, and natural soil settlement can alter the soil structure beneath the airport, potentially creating cavities. Current conventional detection methods include visual inspection, radar, acoustic methods, infrared detection, and laser scanning. While these technologies can locate underground cavities, their accuracy needs improvement. Furthermore, addressing these cavities requires not only location detection but also measurement of their shape and volume.

[0003] As a current mainstream method, ground-penetrating radar (GPR) technology is a highly effective non-destructive detection method. Its main components include a radar and an antenna. The detection method involves moving the GPR device along a predetermined path, transmitting and receiving radar waves, and analyzing the collected data to determine the location of the cavity. However, it has significant drawbacks, primarily manifested in:

[0004] (1) Although ground penetrating radar can detect the location of cavities, it cannot detect the shape and volume of cavities.

[0005] (2) For smaller cavities, radar waves may not generate enough reflected signals to detect potential cavities, which may lead to misjudgment.

[0006] (3) Ground penetrating radar equipment and data processing software are expensive, especially when high precision and high resolution are required. Summary of the Invention

[0007] In order to overcome the shortcomings of the prior art, the present invention aims to provide a method and device for detecting the volume of a small underground cavity, so as to solve the problem that the prior art can only detect the location of the cavity but cannot obtain the data inside the cavity, and reduce the equipment cost and data processing cost.

[0008] To achieve the above objectives, the technical solution adopted by the present invention is as follows:

[0009] A method for detecting the volume of a small-aperture underground cavity includes the following steps:

[0010] S1 detects underground cavities and sends a lidar sensor from the top of the cavity.

[0011] S2, turn on the lidar to collect point cloud data. The point cloud data collected within the effective collection time of the lidar is recorded as a point cloud data.

[0012] S3, divide the point cloud data into average groups, calculate the cavity volume for each of the several groups of point cloud data, and solve for the average volume A and standard deviation B.

[0013] S4. If B / A is greater than or equal to the set accuracy threshold, increase the number of point cloud data in each group of point cloud data, recalculate the average volume A and standard deviation B until B / A is less than the set accuracy threshold. At this time, the number of point cloud data in each group of point cloud data is the optimal data scale, and the average volume obtained from each group of point cloud data is the cavity volume corresponding to that point cloud data.

[0014] S5, adjust the vertical position of the lidar in the cavity, repeat S2, and group a point cloud data according to the data scale to calculate the cavity volume;

[0015] S6, repeat S5, until the average volume A obtained by the solution is greater than the previous solution. The position with the largest calculated cavity volume is the cavity opening position, and the largest cavity volume is taken as the final cavity volume.

[0016] S7 uses any set of point cloud data from the point cloud data at the location where the cavity volume is at its largest in S6 to generate the entire internal shape of the cavity.

[0017] In one embodiment, step S1 involves using ground-penetrating radar to detect underground cavities and predicting the top position of the cavity. A hole is then drilled at the predicted top position of the cavity to send a lidar sensor into the cavity.

[0018] In one embodiment, the lidar is an ultra-wide-angle 4D lidar with a field of view extended to 360° horizontally and 90° vertically, enabling the acquisition of three-dimensional spatial point clouds with a hemispherical field of view.

[0019] In one embodiment, in step S3, the point cloud data is divided into multiple point cloud data groups. The convex hull algorithm is used to solve for the N cavity volumes of the first N point cloud data groups, and the average volume A and standard deviation B of the N cavity volumes are calculated.

[0020] In one embodiment, in step S3, each group of point cloud data contains 2000 point cloud data points, and in step S4, 2000 point cloud data points are added each time, and the accuracy threshold is set to 5%.

[0021] In one embodiment, volume comparison is performed using radars at different locations to ensure that the radar is positioned at the cavity opening, thereby ensuring that the laser point cloud can reflect the entire internal boundary of the cavity and accurately measure the cavity volume.

[0022] In one embodiment, the measured point cloud is used to display the cavity outline in a three-dimensional coordinate system using drawing software.

[0023] In another aspect, the present invention also provides a small-aperture underground cavity volume detection device for implementing the small-aperture underground cavity volume detection method, the detection device comprising a ground support device, a protective pipe, a telescopic device and a lidar.

[0024] The top end of the telescopic device is connected to the ground support device, and the bottom end is equipped with the lidar. The inner diameter of the protective tube is larger than that of the telescopic device. It is used to pre-embed and penetrate the top of the cavity after detecting the underground cavity. The telescopic device carries the lidar through the protective tube to probe into the cavity.

[0025] In one embodiment, the ground support device is a tripod, and the telescopic device is an electric push rod equipped with a wireless controller, connected to the bottom of the tripod.

[0026] In one embodiment, the telescopic device is connected to the ground support device via a precision control device, which is a movable rod including an outer fixing ring welded to the inner side of the ground support device. A swing arm, a gear, and an internal screw are installed inside the outer fixing ring. The internal screw is connected to the top of the telescopic device, the swing arm is connected to the gear, and the gear meshes with the internal screw. By shaking the swing arm, the gear is driven to rotate, which in turn drives the internal screw, thereby achieving precise displacement control.

[0027] Compared with existing technologies, this invention can not only detect the location of the cavity, but also use lidar to scan and obtain data inside the cavity, calculate the cavity volume based on this data and generate a three-dimensional data map, thereby obtaining accurate data such as its shape and volume. Attached Figure Description

[0028] Figure 1 This is a flowchart of the detection method of the present invention.

[0029] Figure 2 This is a schematic diagram of the detection device of the present invention.

[0030] Figure 3 This is a schematic diagram showing how the B / A ratio and the true error change with the number of point clouds. Detailed Implementation

[0031] The embodiments of the present invention will now be described in detail with reference to the accompanying drawings and examples.

[0032] refer to Figure 1 As shown, the present invention provides a method for detecting the volume of a small-aperture underground cavity, comprising the following steps:

[0033] S1 detects underground cavities and sends a lidar sensor from the top of the cavity.

[0034] After using ground-penetrating radar (GPR) to locate potential cavities, the position of the cavity's top is predicted. A hole is drilled at the predicted top position, and a lifting device is used to send a lidar sensor into the cavity. Furthermore, in this step, if... Figure 2 The device shown can deliver a lidar into a cavity. The device mainly includes a ground support device 1, a protective tube 2, and a telescopic device 4. When combined with the lidar delivered by the device, it constitutes the main components of the small-aperture underground cavity volume detection device of the present invention.

[0035] In the diagram, the ground support device 1 is a tripod, providing ground support. The top of the telescopic device 4 is connected to the ground support device 1, specifically, the top can be connected directly below the apex of the tripod, in a vertical position, with the lidar installed at its bottom. The inner diameter of the protective pipe 2 is larger than that of the telescopic device 4. After detecting the underground cavity, the protective pipe 2 is pre-embedded, allowing it to penetrate the top of the cavity. The length of the protective pipe 2 should not be too long, just enough to penetrate the top of the cavity. The protective pipe 2 serves as the entry and exit path for the telescopic device 4, which carries the lidar through the protective pipe 2 to enter the cavity.

[0036] In this embodiment of the invention, the telescopic device 4 can be an electric push rod equipped with a wireless controller to quickly control the raising and lowering of the lidar for initial adjustment. Simultaneously, the ground support device 1 and the telescopic device 4 can be connected via a precision control device 3 for precise adjustment. For example, the precision control device 3 is a movable rod that can be manually controlled. It mainly includes a swing arm, a gear, an internal screw, and an external fixing ring. The swing arm, gear, and internal screw are all installed within the external fixing ring. The internal screw is connected to the top of the telescopic device 4, and the precision control device 3 is welded to the inner side of the ground support device 1 via the external fixing ring. The swing arm is connected to the gear, and the gear meshes with the internal screw. By shaking the swing arm, the gear rotates, thereby driving the internal screw and achieving precise displacement control. The precision control device 3 enables small-amplitude raising and lowering of the telescopic device 4, thus precisely adjusting the position of the lidar.

[0037] In this embodiment of the invention, the protective tube 2 can be a PVC pipe, which is first inserted into the cavity after the channel is made to provide a channel for the laser radar to be inserted.

[0038] Before inserting the lidar into a potential cavity, a hole is first drilled in the ground, and then excavation is carried out down to the vicinity of the cavity entrance indicated by the ground-penetrating radar. Next, the protective pipe 2 is pre-embedded along the pre-drilled hole. After pre-embedding, the ground support device 1 is connected to the telescopic device 4, and the lidar is inserted into the cavity along the protective pipe 2. This process provides some protection for the lidar.

[0039] S2, turn on the lidar to collect point cloud data.

[0040] Based on the ground drilling operation, the height of the cavity top from the ground is predicted. The lidar is then extended to the predicted height via operation S1 and activated to collect point cloud data inside the cavity. In this invention, the lidar is an ultra-wide-angle 4D lidar (Unitree 4D LiDar), which has ultra-wide-angle scanning capability, with a field of view (FOV) extended to 360° horizontally and 90° vertically, enabling the acquisition of three-dimensional spatial point clouds with a hemispherical field of view.

[0041] Point cloud data collected within the effective acquisition time of the lidar is recorded as a single point cloud data set. In this embodiment, the effective acquisition time is set to be more than 20 seconds.

[0042] Furthermore, approximately 20,000 data points are collected per second, and over 400,000 point cloud data points can be collected in more than 20 seconds. The point cloud data collected within the effective collection time is recorded as one point cloud data point.

[0043] S3. Divide a point cloud data set into average groups, calculate the cavity volume for each of the several groups of point cloud data, and solve for the average volume A and standard deviation B.

[0044] In this embodiment, a point cloud data is divided into multiple point cloud data groups. By default, 2000 point cloud data are taken as a group, and then the computer is used to solve for the 10 cavity volumes based on the first 10 point cloud data groups.

[0045] Furthermore, this step uses the convex hull algorithm to calculate the volume for each set of data. Based on the 10 cavity volumes obtained from the first 10 point cloud data sets, the average volume A and the standard deviation B of these 10 cavity volumes are calculated.

[0046] S4, determine the size of the data set.

[0047] Upon completion of step S3, the ratio of the standard deviation B to the average volume A of the volume calculation results for different groups of 2000 point cloud data is examined. If the ratio is less than the set accuracy threshold, such as 5%, the accuracy is considered acceptable. Otherwise, the accuracy is considered insufficient, and the number of point cloud data in each group needs to be increased. Step S3 is repeated, and the average volume A and standard deviation B of each group are calculated after increasing the number of point cloud data. In this embodiment, the number of point cloud data is increased by 2000 each time. This continues until the ratio of the standard deviation B to the average volume A is less than the set accuracy threshold, which is 5% in this embodiment. At this point, the number of point cloud data in each group is the optimal data size, and the average volume obtained from each group of point cloud data is the accurate cavity volume corresponding to that point cloud data.

[0048] Taking a rectangular cavity with dimensions of 0.53 × 0.46 × 0.35 mm as an example, the lidar acquired 432,000 point cloud data points after 20 seconds of operation. This 432,000 point cloud data points constitute one set. This set of data was then grouped according to the acquisition time, with 2,000 point cloud data points per group, resulting in 216 groups. Similarly, groups of 4,000, 6,000, 8,000, 10,000, and 20,000 point cloud data points were grouped into 108, 72, 54, 43, and 21 groups respectively. The first 10 groups were selected, and the volume average, variance, and other results were calculated, as shown in Table 1. The B / A ratio and the actual error are plotted on [the graph / chart / database]. Figure 3 .

[0049] Table 1

[0050] Point cloud quantity Standard deviation B Volume average value A Standard volume B / A True error 2000 0.0322 0.0653 0.08533 49.31087 23.47357 4000 0.0159 0.0714 0.08533 22.26891 16.32486 6000 0.0102 0.0803 0.08533 12.70237 5.894762 8000 0.00628 0.0832 0.08533 7.548077 2.496191 10000 0.00341 0.0848 0.08533 4.021226 0.621118 20000 0.00281 0.0845 0.08533 3.325444 0.972694

[0051] From Table 1 and Figure 3 It can be seen that for a true volume of 0.0853m 3 In the cavity, when the number of data points in each group of point cloud data was 2000, 4000, 6000, and 8000, the B / A ratio was greater than 5%, indicating a large error in the calculated results. Therefore, the number of data points in a single group was increased to 10000, at which point the B / A ratio was 4.02%, less than 5%, indicating a relatively small error. 10000 data points can be used as the number of data points in each group. To verify this, the number of data points in a single group was increased to 20000, and the B / A ratio was 3.33%, also less than 5%. However, the calculation time and cloud map output time for 20000 point cloud data points were much longer than for 10000. Therefore, for 0.0853m... 3 For the cavity, 10,000 is the optimal number of point clouds. Comparing the actual errors of 10,000 and 20,000 point cloud data, both are found to be less than 1%, therefore both have sufficient accuracy.

[0052] S5, volume comparison.

[0053] To ensure that the predicted position is the top of the cavity, after completing the test and calculation of one set of point cloud data, the vertical elevation of the lidar can be finely adjusted by the movable rod. S2 is repeated, and a set of point cloud data is grouped according to the data scale determined in S4 to calculate the cavity volume.

[0054] S6. Repeat S5 for multiple sets of experiments until the calculated average volume A is greater than the previous result. The location with the largest calculated cavity volume is the location of the lidar, i.e., the cavity opening location, and the largest cavity volume is used as the final cavity volume. That is, the volume measurement is most accurate when the lidar is basically located at the cavity opening. This step ensures that the lidar is located at the cavity opening by comparing the calculated volumes of the lidar at different positions, thus ensuring that the lidar point cloud can reflect the entire internal boundary of the cavity and measure the accurate cavity volume.

[0055] S7 calculates the volume and displays the shape. After completing S6, using any set of point cloud data from the point cloud data at the location with the largest cavity volume, the cavity shape is generated for visualization. For example, using the measured point cloud data, the point cloud is displayed in a three-dimensional coordinate system using drawing software to show the cavity outline.

[0056] Through the above-described scheme, this invention enables high-precision, automated detection of underground cavities, significantly improving the accuracy and efficiency of detection. It also allows for visual interactive control and can accurately, quickly, and easily detect the shape and volume of small underground cavities, providing precise cavity location information, which facilitates rapid localization and repair.

Claims

1. A method of detecting a small opening in an underground cavity volume, characterized by, It comprises the following steps: S1, detecting an underground cavity and sending a laser radar into the cavity from the top position of the cavity; S2, opening the laser radar to collect point cloud data, and the point cloud data collected within the effective collection time of the laser radar is recorded as a point cloud data; S3, grouping the point cloud data, calculating the cavity volume for each group of point cloud data, and solving the average volume A and the standard deviation B; S4, if B / A is greater than or equal to the set precision threshold, the number of point cloud data in each group of point cloud data is increased, the average volume A and the standard deviation B are recalculated until B / A is less than the set precision threshold, at which time the number of point cloud data in each group of point cloud data is the optimal data size, and the average volume obtained by each group of point cloud data is the cavity volume corresponding to the point cloud data; S5, adjusting the vertical position of the laser radar in the cavity, repeating S2, and grouping a point cloud data at the data size to calculate the cavity volume; S6, repeating S5 until the average volume A obtained is greater than the last time, calculating the position of the maximum cavity volume as the cavity opening position, and taking the maximum cavity volume as the final cavity volume; S7, using any one group of point cloud data in a point cloud data obtained at the position of the maximum cavity volume in S6 to generate the internal morphology of the entire cavity.

2. The method of claim 1, wherein, S1, detecting an underground cavity using a geological radar and predicting the top position of the cavity, drilling a hole at the predicted top position of the cavity, and sending a laser radar into the cavity.

3. The method of claim 1, wherein The laser radar is a super-wide-angle 4D laser radar, the field of view angle is expanded to 360° horizontally and 90° vertically, and three-dimensional space point cloud acquisition with a hemispherical field of view angle is realized.

4. The method of claim 1, wherein S3, grouping the point cloud data to obtain a plurality of point cloud data groups, and using a convex hull algorithm to solve N cavity volumes for the first N point cloud data groups, and solving the average volume A and the standard deviation B of the N cavity volumes.

5. The method of claim 1, wherein, In S3, each group of point cloud data has 2000 point cloud data, in S4, 2000 point cloud data are added each time, and the precision threshold is set to 5%.

6. The method of claim 1, wherein, Using different position radars to calculate the volume comparison, ensuring that the radar is at the cavity opening, so that the laser point cloud can feedback the entire cavity internal boundary, and the accurate cavity volume is measured.

7. The method of claim 1, wherein the small opening is a borehole. Using the measured point cloud, the point cloud is displayed in a three-dimensional coordinate system through drawing software, and the cavity contour is displayed.

8. A small opening underground cavity volume detection device, characterized in that, A small opening underground cavity volume detection method according to any one of claims 1 to 7, comprising a ground support device (1), a protection pipe (2), a telescopic device (4) and a laser radar; The top end of the telescopic device (4) is connected with the ground support device (1), the bottom end is provided with the laser radar, the inner diameter of the protection pipe (2) is greater than the telescopic device (4), and the protection pipe (2) is used to be pre-buried and penetrate the top of the cavity after detecting the underground cavity, and the telescopic device (4) carries the laser radar to penetrate into the cavity through the protection pipe (2).

9. The small opening underground cavity volume detection device according to claim 8, characterized in that, The ground support device (1) is a tripod, and the telescopic device (4) is an electric push rod equipped with a wireless controller and connected below the tripod.

10. The small opening underground cavity volume detection device according to claim 8, characterized in that, The telescopic device (4) is connected to the ground support device (1) through a precision control device (3). The precision control device (3) is a movable rod, including an outer fixing ring welded to the inner side of the ground support device (1). A swing arm, a gear, and an internal screw are installed inside the outer fixing ring. The internal screw is connected to the top of the telescopic device (4). The swing arm is connected to the gear, and the gear meshes with the internal screw. By shaking the swing arm, the gear is driven to rotate, which in turn drives the internal screw, thus achieving precise displacement control.

Images