A compaction degree unmanned automatic detection system and method based on three-dimensional laser scanning

Abstract

This invention relates to an unmanned automatic compaction degree detection system and method based on three-dimensional laser scanning, comprising a vehicle-mounted terminal, a database and application server, and a monitoring client. The vehicle-mounted terminal processes and calculates the collected data and displays relevant information. The database and application server includes a database module, an analysis and calculation module, and an information feedback module, which stores, analyzes, and transmits data. The monitoring client creates tasks, sets parameters, performs remote real-time monitoring, and compiles and outputs result reports after the measurement is completed. This invention uses three-dimensional laser scanning technology to replace the traditional sand-filling method for measuring test pit volume, improving measurement efficiency and facilitating subsequent processing. Combined with automated excavation, soil removal, and drying technology, it reduces human intervention, achieving unmanned intelligent acquisition and synchronous remote monitoring of road base compaction degree, ensuring the authenticity and controllability of test results.

Description

Technical Field

[0001] This invention belongs to the field of highway construction quality inspection technology, specifically relating to an unmanned automatic compaction inspection system based on three-dimensional laser scanning. Background Technology

[0002] The compaction quality of the road base course is a crucial indicator of road construction quality, directly impacting the road's service life and operational safety. According to the current "Specifications for Field Testing of Highway Subgrade and Pavement" (JTG 3450-2019), methods such as pit excavation and sand filling are commonly used to measure road compaction. The compaction degree is determined by measuring the pit volume and soil mass through pit sampling and calculating the dry density. However, since the entire process requires manual operation, human factors significantly influence the measurement, resulting in low work efficiency.

[0003] Currently, some automatic compaction measurement devices based on the sand cone method exist, enabling automated excavation of test pits and measurement of compaction degree with minimal manual intervention. However, how to automatically recycle and process the used standard sand remains a key issue. While 3D laser scanning technology is widely used in target volume measurement for numerous engineering projects, including measuring the volume of test pits for dam material compaction testing, this technology is rarely used for calculating the volume of test pits for road compaction testing, and currently, most applications require on-site personnel for operation.

[0004] Therefore, it is considered to use a portable 3D laser scanner to measure the volume of the test pit, and combine it with automated mechanical equipment to realize the entire process of compaction measurement, such as pit excavation, soil sampling, and weighing, without human intervention. At the same time, the on-site measurement status can be monitored remotely in real time, avoiding the subsequent cleanup problems caused by the sand filling method, reducing the complexity of the whole set of equipment, and reducing manpower input. Summary of the Invention

[0005] The purpose of this invention is to overcome the shortcomings of the prior art and provide an unmanned automatic compaction degree detection system and method based on three-dimensional laser scanning. It uses three-dimensional laser scanning technology to replace the traditional sand filling method to measure the volume of test pits, improves measurement efficiency, facilitates subsequent processing, and combines with automated excavation, soil removal and drying technology to reduce human intervention. It realizes unmanned intelligent acquisition and synchronous remote monitoring of road base compaction degree, ensuring the authenticity and controllability of test results.

[0006] The technical problem solved by this invention is achieved through the following technical solution:

[0007] An unmanned automatic compaction degree detection system based on three-dimensional laser scanning is characterized by comprising a vehicle-mounted terminal, a vehicle-mounted terminal, a database and application server, and a monitoring client.

[0008] The vehicle-mounted terminal includes a GNSS positioning module, a drilling device, a soil sampling and drying device, a three-dimensional scanning device, an integrated controller, and a wireless control device.

[0009] The GNSS positioning module acquires the planar position of the detection system in real time; the drilling device uses spiral blades to automatically excavate the selected area.

[0010] The soil sampling and drying device sucks out the residual soil residue inside the test pit and dries and weighs the collected soil.

[0011] The three-dimensional laser scanning device scans the test pit to obtain point cloud data inside the test pit;

[0012] The integrated controller controls the operation of the on-board mechanical components;

[0013] The wireless control device remotely controls the movement of the vehicle-mounted terminal;

[0014] The vehicle-mounted terminal processes and calculates the collected data and displays relevant information;

[0015] The database and application server include a database module, an analysis and calculation module, and an information feedback module, which store, analyze, and transmit data.

[0016] The monitoring client is responsible for creating tasks, setting parameters, remote real-time monitoring, and compiling and outputting result reports after the measurement is completed.

[0017] An unmanned automatic method for compaction degree detection based on three-dimensional laser scanning, characterized in that the method comprises the following steps:

[0018] 1) Log in to the monitoring client, create a road base flatness detection task, set the drilling depth, drying time, scanning range, valve opening time, excavation location, and issue a dispatch command for the measuring device;

[0019] 2) The information from 1) is stored in the database and application server's database module via the Internet. The location information of the vehicle is measured in real time using GNSS-RTK positioning technology and sent to the remote server.

[0020] 3) Based on the device's location information, the client observes the surrounding environment through the vehicle-mounted camera deployed on the device and uses the wireless control device to move the device to the location to be tested.

[0021] 4) The vehicle terminal accesses the database and application server via 5G communication, receives corresponding task information from the database module, and controls the device to run through the integrated controller;

[0022] 5) The drilling device lowers the sleeve to a tight fit with the ground, and uses spiral blades to excavate a test pit and transport the excavated material to the soil storage box of the soil extraction device;

[0023] 6) The soil sampling device uses a negative pressure pump to suck up the residual soil inside the test pit, and uses an electric valve to let a portion of the soil in the storage box fall into the sampling box. It then uses a heating coil to heat and dry the soil sample, and weighs the soil sample before and after heating using a weighing sensor.

[0024] 7) The three-dimensional laser scanning device scans the test pit after soil sampling, collects point cloud data, and the vehicle terminal processes and calculates the point cloud data to obtain the internal volume of the test pit. Combined with the weight of the soil sample before and after drying, the dry density of the soil at the corresponding location is calculated to obtain the compaction degree of the road base layer, and it is displayed together with other relevant information.

[0025] 8) The wireless control device sends the above data to the database and application server. The analysis and calculation module within the server analyzes and processes the information, storing the analysis results and data in the database module. The monitoring client monitors the road compaction measurement in real time based on the information from the server. An alarm is triggered if the compaction degree is unqualified. After the measurement is completed, the client can query and statistically analyze the results and output reports.

[0026] Simultaneously, the point cloud data of the test pit is processed. The analysis and calculation module uses methods such as density method and normal vector clustering segmentation to divide the point cloud data inside and outside the test pit. The specific steps are as follows:

[0027] ① After collecting point cloud data of the test pit and its vicinity, obtain the maximum height z in the point cloud coordinates. max And extract the heights z≤z from all point clouds. max -5 (base plate thickness is 5cm) point cloud data, this part of the data is the point cloud data of the ground surface and the inside of the test pit;

[0028] ② Based on the different point cloud densities on the ground surface and inside the test pit, select an appropriate neighborhood radius r and calculate the number of neighborhood point clouds N for each point. ri Statistical analysis was performed, and the detection parameters were calculated using the following formula.

[0029]

[0030] Where: N r For detection parameters;

[0031] n is the total number of point clouds;

[0032] k is a coefficient less than 1;

[0033] When point A i (x i ,y i ,z iThe number of neighborhood point clouds N) ri The number of points is greater than the detection parameter N. r When this happens, it is classified as point Q inside the test pit. i Conversely, it is classified as a surface point P. i This divides all point cloud data into two parts: the point cloud inside the test pit and the point cloud on the ground surface.

[0034] ③ The segmented point cloud data is further processed using the normal vector clustering segmentation method. For the point cloud data Q inside the test pit, the normal vector nQ of each point is calculated using the pcnormals function in MATLAB. i Then consider making corrections.

[0035] Let the coordinates of point O be...

[0036]

[0037] In the formula, x i y i z i These are the spatial coordinates of each point in the point cloud data;

[0038] n is the total number of point clouds;

[0039] If you click Q i normal vector nQ i With vector OQ i If the included angle α is greater than 90°, the direction of the corresponding point cloud normal vector is corrected; otherwise, the original normal vector is retained.

[0040] For surface point cloud data P, the normal vector of its fitted plane is used as the feature normal vector n; first, let the equation of the fitted plane be:

[0041] ax + by + cz + d = 0

[0042] The equation can be transformed into

[0043] z = a0x + a1y + a2

[0044] The following matrix is ​​obtained using the least squares method.

[0045]

[0046] The characteristic normal vector n(a,b,c) is obtained by solving the problem, and then the normal vector n of each point cloud in point Q inside the test pit is obtained. Qi The angle θ between the vector and the eigenvector n;

[0047] ④ Select the surface point cloud P as the seed point cloud set, and select a point P. i As a seed point, a search is conducted within point Q inside the test pit with a search radius r0 to form a point cloud set N.q Meanwhile, the angle θ between the normal vector and the feature vector n of each point is calculated.

[0048] If N q If there are no points in Q with θ greater than the threshold θ', then it proves that all points in Q belong to the ground surface. Therefore, delete N from Q. q Meanwhile, increase r0 (r0 = r0 + r0) and continue searching in point Q inside the test pit;

[0049] If N q If there are points in P with an included angle θ greater than the threshold θ', it proves that some of these points belong to the interior of the test pit. Therefore, delete N from P. p And select a new point P. i Repeat the above steps until P is empty, then the point cloud data of the test pit surface can be obtained.

[0050] Furthermore, the volume of the test pit is determined by calculating the segmented point cloud data using the slicing method. The specific steps are as follows:

[0051] ① Cut the point cloud data inside the test pit from top to bottom using n+1 equally spaced horizontal planes to obtain a series of horizontal point cloud slices S. i The following formula is used for calculation;

[0052]

[0053]

[0054] In the formula, h is the spacing value;

[0055] H represents the maximum height of the point cloud inside the test pit;

[0056] n is the number of slices;

[0057] x, y, z are the coordinates of the point cloud points;

[0058] Construct a square on the slice that can contain all points, with its diagonal and focus O. i Let Q be any point on the slice, and let Q be an endpoint. i Constructing ray O i Q i Using this as the initial scan line, a counter-clockwise scan is performed, connecting the scanned points sequentially to generate the boundary contour polygon M. i (i = 0, 1, ..., n);

[0059] ② Use the following formula to calculate polygon M i Area A i Calculation

[0060]

[0061] In the formula xi y i The outer contour polygon M of the point cloud in the slice plane i Vertex m of (i = 0, 1, ..., n) j The coordinates of (j = 0, 1, ..., k); j is the vertex number of the polygon of the outer contour of the point cloud slice; i is the number of the point cloud slice.

[0062] ③ The volume of each segment of the test pit is calculated using the following formula and then summed to obtain the total volume V of the test pit.

[0063]

[0064] In the formula A i The outer contour polygon M of the point cloud in the slice plane i The area; h is the spacing value;

[0065] ④ The mass of all solid particles inside the test pit is estimated by estimating the mass of solid particles in the soil sample obtained by drying. Combined with the calculated volume V, the dry density of the soil sample is calculated using the following formula.

[0066]

[0067] In the formula: ρ d This refers to the dry density of the soil sample.

[0068] M represents the mass of solid particles inside the test pit;

[0069] The vehicle-mounted display terminal will display calculation parameters (number of slices, spacing, etc.), calculation results, and test pit models.

[0070] The advantages and beneficial effects of this invention are as follows:

[0071] 1. The present invention is an unmanned automatic compaction detection system based on three-dimensional laser scanning, which can quickly excavate test pits at sampling locations, take soil samples, dry and weigh them, and calculate the dry density of the soil sample based on the volume of the test pit measured by three-dimensional laser scanning technology, thereby realizing the automatic measurement of compaction degree.

[0072] 2. The present invention is a three-dimensional laser scanning-based unmanned automatic compaction degree detection system, which can display the compaction degree measurement results and related information on the vehicle-mounted terminal and transmit the field data to the remote client. After the measurement is completed, the compaction degree distribution is automatically counted. Relevant personnel can remotely monitor the compaction degree measurement in real time and control the movement of the device through a wireless remote control device to achieve remote real-time control.

[0073] 3. The present invention is a three-dimensional laser scanning-based unmanned automatic compaction detection system. Combining three-dimensional laser scanning technology with automatic excavation, soil extraction and drying technology, it can not only realize the automatic measurement of the compaction degree of road base, but also realize the automatic data collection and analysis, as well as remote real-time monitoring.

[0074] 4. This invention relates to an unmanned automatic compaction degree detection system based on three-dimensional laser scanning. Targeting the characteristics of road base compaction degree measurement, it utilizes a portable three-dimensional laser scanner combined with automatic excavation, soil sampling, and drying technology to develop a highly automated road base compaction degree measurement device. This device can automatically excavate and sample soil from the road base, and quickly dry and weigh it; it can perform three-dimensional laser scanning inside the test pit to quickly measure the pit volume, thereby calculating the dry density of the soil sample and obtaining the compaction degree at the corresponding location; and it can feed back on-site measurement data to a vehicle-mounted terminal and a client terminal for remote control. Attached Figure Description

[0075] Figure 1 This is a schematic diagram of the detection system of the present invention;

[0076] Figure 2 This is a flowchart of the point cloud data processing of the present invention;

[0077] Figure 3 This is a schematic diagram of the drilling device of the present invention;

[0078] Figure 4 This is a schematic diagram of the soil drying device of the present invention;

[0079] Figure 5 This is a schematic diagram of the structure of the three-dimensional laser scanning device of the present invention. Detailed Implementation

[0080] The present invention will be further described in detail below through specific embodiments. The following embodiments are merely descriptive and not limiting, and should not be used to limit the scope of protection of the present invention.

[0081] like Figure 1 As shown, an unmanned automatic compaction detection system based on three-dimensional laser scanning is innovative in that it includes a vehicle-mounted terminal, a database and application server, and a monitoring client.

[0082] The vehicle-mounted terminal includes a GNSS positioning module, a drilling device, a soil sampling and drying device, a 3D scanning device, an integrated controller, and a wireless control device. The GNSS positioning module acquires the planar position of the detection system in real time. The drilling device uses a spiral blade to automatically excavate the selected area. The soil sampling and drying device sucks out the residual soil inside the test pit and dries and weighs the collected soil. The 3D laser scanning device scans the test pit to acquire point cloud data inside the pit. The integrated controller controls the operation of the mechanical components of the vehicle-mounted terminal. The wireless control device remotely controls the movement of the vehicle-mounted terminal.

[0083] The vehicle-mounted terminal processes and calculates the collected data and displays relevant information.

[0084] The database and application server include a database module, an analysis and calculation module, and an information feedback module, which store, analyze, and transmit data.

[0085] The monitoring client is responsible for creating tasks, setting parameters, remote real-time monitoring, and compiling and outputting result reports after the measurement is completed.

[0086] like Figure 2 As shown, an unmanned automatic compaction degree detection method based on three-dimensional laser scanning is innovative in that the method comprises the following steps:

[0087] (1) User login monitoring client

[0088] Log in to the developed monitoring client and verify user permissions, which include browsing, operation, and management. Browsing users can perform functions such as remote real-time monitoring, observing on-site conditions, and querying compaction degree; operation users, with browsing user permissions, can perform functions such as creating compaction degree measurement tasks, inputting measurement parameters, and remotely controlling device movement; management users, in addition to having the permissions of the above two types of users, are responsible for managing the permissions of other users.

[0089] (2) Create a new road compaction measurement task and input the device operating parameters.

[0090] Create a new compaction measurement task on the monitoring client, including the measurement location coordinates, compaction measurement standard, and measurement start time. Input the operating parameters such as drill hole depth, soil sample weight, and drying time according to the specific measurement task, and send the above information to the remote database and application server.

[0091] (3) The vehicle-mounted terminal obtains measurement tasks and operating parameters from the database and application server.

[0092] The vehicle-mounted terminal starts up, retrieves compaction measurement task information and device operating parameters from the database and application server, and feeds them back to the integrated controller as operating parameters for each device in the subsequent compaction measurement process.

[0093] (4) The vehicle-mounted terminal sends the location coordinates to the database and application server.

[0094] By establishing a differential reference station at the construction site and installing a GNSS positioning antenna on the automatic measuring device, the coordinate information of the device is collected in real time using GNSS-RTK positioning technology and sent to the database and application server.

[0095] (5) Using the device's location coordinates, move the compaction testing device to the measurement position via a wireless remote control device.

[0096] The database and application server converts the device location coordinates acquired by the GNSS device into device center coordinates and feeds them back to the monitoring client. The monitoring client compares the compaction degree detection location coordinates with the device center coordinates, and observes the surrounding environment of the device in real time through vehicle-mounted cameras installed in the center and on both sides of the device. It uses a wireless control device to move the device to the detection position, while fixing the device wheels to ensure that it does not deviate during operation.

[0097] (6) The drilling device is used to excavate and remove soil from the test pit at the measurement location.

[0098] Drilling device as attached Figure 3 As shown, stepper motor 31 controls screw 21 to rotate, causing base sleeve 1 to descend into close contact with the ground. Stepper motor 32 starts, controlling the rotating blade on screw 22 to rotate downwards in a spiral motion, drilling a hole to a specified depth. After the test pit is drilled, stepper motor 22 rotates in the opposite direction, causing the spiral blade to rotate upwards and carry out the excavated soil. At the same time, a scraper is used to clean the residual soil on the blade, and the excavated soil is transported to the storage box in the soil extraction device through a soil conveying hose.

[0099] (7) The soil sampling and drying device extracts the residue inside the test pit and dries and weighs a portion of the soil sample.

[0100] The structure of the soil drying device is shown in the attached figure. Figure 4 As shown, the negative pressure pump 4 extracts residual soil from the excavated test pit and sucks it into the soil storage box 5. The weighing sensor 61 monitors the weight of the soil in the storage box in real time until the value remains constant, at which point the data is sent to the server. The electric valve 8 is activated, and after a preset opening time, a portion of the soil is transported to the heating device. The actual mass of this portion of soil is measured by the weighing sensor 62. The soil is then dried using the heating coil 7. When the value of the weighing sensor 62 remains constant, drying is complete, and the sensor value at this moment is recorded. The vehicle-mounted terminal processes and calculates the above information and transmits it to the database and application server.

[0101] (8) The three-dimensional laser scanning device performs a three-dimensional scan of the test pit and collects point cloud data inside the test pit.

[0102] 3D laser scanning device as attached Figure 5 After the test pit drilling is completed, the sleeve and spiral blades rise, and the portable 3D laser scanner 9, positioned at the rear of the device, slides to the end of the track via the horizontal guide slider 10, positioning the portable 3D laser scanner directly above the test pit. The scanner, via a rear-connected rotary motor 11, scans the inner wall of the test pit at a fixed scanning rate f within a set angle, collecting point cloud data A of the test pit area. i (x i ,y i ,z iThe data is then transmitted to the airborne terminal for further analysis.

[0103] (9) Analyze and process point cloud data

[0104] The test pit point cloud data is processed. The analysis and calculation module uses methods such as density method and normal vector clustering to divide the point cloud data inside and outside the test pit. The specific steps are as follows:

[0105] ① After collecting point cloud data of the test pit and its vicinity, obtain the maximum height z in the point cloud coordinates. max And extract the heights z≤z from all point clouds. max The point cloud data is -5 (base plate thickness is 5cm). This part of the data is the point cloud data of the ground surface and the inside of the test pit.

[0106] ② Based on the different point cloud densities on the ground surface and inside the test pit, select an appropriate neighborhood radius r and calculate the number of neighborhood point clouds N for each point. ri Statistical analysis was performed, and the detection parameters were calculated using the following formula.

[0107]

[0108] Where: N r For detection parameters;

[0109] n is the total number of point clouds;

[0110] k is a coefficient less than 1

[0111] When point A i (x i ,y i ,z i The number of neighborhood point clouds N) ri The number of points is greater than the detection parameter N. r When this happens, it is classified as point Q inside the test pit. i Conversely, it is classified as a surface point P. i This divides all point cloud data into two parts: the point cloud inside the test pit and the point cloud on the ground surface.

[0112] ③ The segmented point cloud data is further processed using normal vector clustering. For the point cloud data Q inside the test pit, the normal vector nQ of each point is calculated using the pcnormals function in MATLAB. i Then consider revising it.

[0113] Let the coordinates of point O be...

[0114]

[0115] In the formula, x i y i zi These are the spatial coordinates of each point in the point cloud data;

[0116] n represents the total number of point clouds.

[0117] If you click Q i normal vector nQ i With vector OQ i If the included angle α is greater than 90°, the direction of the corresponding point cloud normal vector is corrected; otherwise, the original normal vector is retained.

[0118] For surface point cloud data P, the normal vector of its fitted plane is used as the feature normal vector n; first, let the equation of the fitted plane be...

[0119] ax + by + cz + d = 0

[0120] The equation can be transformed into

[0121] z = a0x + a1y + a2

[0122] The following matrix is ​​obtained using the least squares method.

[0123]

[0124] The characteristic normal vector n(a,b,c) is obtained by solving the problem, and then the normal vector n of each point cloud in point Q inside the test pit is obtained. Qi The angle θ between the eigenvector and the eigenvector n

[0125] ④ Select the surface point cloud P as the seed point cloud set, and select a point P. i As a seed point, a search is conducted within point Q inside the test pit with a search radius r0 to form a point cloud set N. q Meanwhile, the angle θ between the normal vector and the feature vector n of each point is calculated.

[0126] If N q If there are no points in Q with θ greater than the threshold θ', then it proves that all points in Q belong to the ground surface. Therefore, delete N from Q. q Meanwhile, increase r0 (r0 = r0 + r0) and continue searching in point Q inside the test pit;

[0127] If N q If there are points in P with an included angle θ greater than the threshold θ', it proves that some of these points belong to the interior of the test pit. Therefore, delete N from P. p And select a new point P. i Repeat the above steps until P is empty, then the point cloud data of the test pit surface can be obtained.

[0128] (10) The vehicle-mounted terminal processes and calculates the data to determine the volume of the test pit, and obtains the compaction degree by combining the weight of the dried soil sample. At the same time, it displays the relevant information.

[0129] The volume of the test pit is calculated from the segmented point cloud data using the slicing method. The specific steps are as follows:

[0130] ① Cut the point cloud data inside the test pit from top to bottom using n+1 equally spaced horizontal planes to obtain a series of horizontal point cloud slices S. i The following formula is used for calculation;

[0131]

[0132]

[0133] In the formula, h is the spacing value;

[0134] H represents the maximum height of the point cloud inside the test pit;

[0135] n is the number of slices;

[0136] x, y, z are the coordinates of the point cloud points.

[0137] Construct a square on the slice that can contain all points, with its diagonal and focus O. i Let Q be any point on the slice, and let Q be an endpoint. i Constructing ray O i Q i Using this as the initial scan line, a counter-clockwise scan is performed, connecting the scanned points sequentially to generate the boundary contour polygon M. i (i = 0, 1, ..., n).

[0138] ② Use the following formula to calculate polygon M i Area A i Calculation

[0139]

[0140] In the formula x i y i The outer contour polygon M of the point cloud in the slice plane i Vertex m of (i = 0, 1, ..., n) j The coordinates of (j = 0, 1, ..., k); j is the vertex number of the polygon of the outer contour of the point cloud slice; i is the number of the point cloud slice.

[0141] ③ The volume of each segment of the test pit is calculated using the following formula and then summed to obtain the total volume V of the test pit.

[0142]

[0143] In the formula A i The outer contour polygon M of the point cloud in the slice plane i The area; h is the spacing value.

[0144] ④ The mass of all solid particles inside the test pit is estimated by calculating the mass of solid particles in the dried soil sample. Combined with the calculated volume V, the dry density of the soil sample is calculated using the following formula.

[0145]

[0146] In the formula: ρ d This refers to the dry density of the soil sample.

[0147] M represents the mass of solid particles inside the test pit;

[0148] The vehicle-mounted display terminal will display calculation parameters (number of slices, spacing, etc.), calculation results, and test pit models.

[0149] (11) The wireless transmission device sends the measurement information back to the database and application server for analysis, processing and storage.

[0150] The vehicle-mounted terminal sends measurement data to a remote database and application server, mainly including test pit point cloud data, test pit volume, test pit volume calculation parameters, test pit model, excavated soil weight, soil sample weight before and after drying, and soil dry density. The analysis and calculation module of the database and application server matches the road base compaction data with the location coordinates emitted by the GNSS device; the database module of the database and application server stores the above data and analysis results for subsequent query and statistics; the information feedback module of the database and application server sends the analyzed data information to the monitoring client in real time.

[0151] (12) Monitor the client to obtain data from the database and application server to achieve real-time monitoring.

[0152] The monitoring client receives data from the database and application server information feedback module in real time, displays the compaction measurement values ​​at various locations on the road BIM map, and issues an alarm when the compaction of the road base layer is unqualified, thus realizing real-time monitoring of the measurement site.

[0153] (13) Statistical analysis of test results and generation of reports.

[0154] After the compaction measurement is completed, the monitoring client can query and statistically analyze the compaction measurement results from the database and the database module of the application server, automatically generate a compaction distribution map, and output it in the form of charts.

[0155] Although embodiments and drawings of the present invention have been disclosed for illustrative purposes, those skilled in the art will understand that various substitutions, variations and modifications are possible without departing from the spirit and scope of the present invention and the appended claims. Therefore, the scope of the present invention is not limited to the contents disclosed in the embodiments and drawings.

Claims

1. A method for unmanned automatic detection of compaction degree based on three-dimensional laser scanning, characterized in that: The detection system employed in this method includes a vehicle-mounted terminal, a vehicle-mounted system, a database and application server, and a monitoring client. The vehicle-mounted terminal includes a GNSS positioning module, a drilling device, a soil sampling and drying device, a 3D laser scanning device, an integrated controller, and a wireless control device. The GNSS positioning module acquires the planar position of the detection system in real time. The drilling device automatically excavates the selected area using a spiral blade. The soil sampling and drying device removes residual soil from the test pit and dries and weighs the collected soil. The 3D laser scanning device scans the test pit to acquire point cloud data inside. The integrated controller controls the operation of the mechanical components of the vehicle-mounted terminal. The wireless control device remotely controls the movement of the vehicle-mounted terminal. The vehicle-mounted terminal processes and calculates the collected data and displays relevant information. The database and application server includes a database module, an analysis and calculation module, and an information feedback module, which stores, analyzes, and transmits data. The monitoring client is responsible for creating tasks, setting parameters, remote real-time monitoring, and compiling and outputting result reports after the measurement is completed. The steps of the method are as follows: 1) Log in to the monitoring client, create a road base flatness detection task, set the drilling depth, drying time, scanning range, valve opening time, excavation location, and issue a dispatch command for the measuring device; 2) The information from 1) is stored in the database and application server's database module via the Internet. The location information of the vehicle is measured in real time using GNSS-RTK positioning technology and sent to the remote server. 3) Based on the device's location information, the client observes the surrounding environment through the vehicle-mounted camera deployed on the device and uses a wireless control device to move the device to the location to be tested. 4) The vehicle terminal accesses the database and application server via 5G communication, receives corresponding task information from the database module, and controls the operation of the device through the integrated controller; 5) The drilling device lowers the sleeve to a tight fit with the ground, and uses spiral blades to excavate a test pit and transport the excavated material to the soil storage box of the soil extraction device; 6) The soil sampling device uses a negative pressure pump to suck up the residual soil inside the test pit, and uses an electric valve to let a portion of the soil in the storage box fall into the sampling box. It then uses a heating coil to heat and dry the soil sample, and weighs the soil sample before and after heating using a weighing sensor. 7) The three-dimensional laser scanning device scans the test pit after soil sampling, collects point cloud data, and the vehicle terminal processes and calculates the point cloud data to obtain the internal volume of the test pit. Combined with the weight of the soil sample before and after drying, the dry density of the soil at the corresponding location is calculated to obtain the compaction degree of the road base layer, and it is displayed together with other relevant information. 8) The wireless control device sends the above data to the database and application server. The analysis and calculation module analyzes and processes the above information and stores the analysis results and data information in the database module. The monitoring client monitors the road compaction measurement in real time based on the information in the server. An alarm will be triggered when the compaction is not up to standard. After the measurement is completed, the results can be queried and statistically analyzed, and reports can be output. The point cloud data of the test pit is processed. The analysis and calculation module uses density method and normal vector clustering segmentation method to divide the point cloud data inside and outside the test pit. The specific steps are as follows: ① After collecting point cloud data of the test pit and its vicinity, obtain the maximum height z in the point cloud coordinates. max And extract the heights z≤z from all point clouds. max -5 point cloud data, this part of the data is the point cloud data of the ground surface and the inside of the test pit; ② Based on the different point cloud densities on the ground surface and inside the test pit, select an appropriate neighborhood radius r and calculate the number of neighborhood point clouds N for each point. ri Statistical analysis was performed, and the detection parameters were calculated using the following formula. ； Where: N r For detection parameters; n is the total number of point clouds; k is a coefficient less than 1; When point A i (x i ,y i ,z i The number of neighborhood point clouds N) ri The number of points is greater than the detection parameter N. r When this happens, it is classified as point Q inside the test pit. i Conversely, it is classified as a surface point P. i This divides all point cloud data into two parts: the point cloud inside the test pit and the point cloud on the ground surface. ③ The segmented point cloud data is further processed using the normal vector clustering segmentation method. For the point cloud data Q inside the test pit, the normal vector nQ of each point is calculated using the pcnormals function in MATLAB. i Then consider making corrections to it; Let the coordinates of point O be... ； In the formula, x i y i z i These are the spatial coordinates of each point in the point cloud data; n is the total number of point clouds; If you click Q i normal vector nQ i With vector OQ i If the included angle α is greater than 90°, the direction of the corresponding point cloud normal vector is corrected; otherwise, the original normal vector is retained. For surface point cloud data P, the normal vector of its fitted plane is used as the feature normal vector n; first, let the equation of the fitted plane be: ； The equation can be transformed into ； The following matrix is ​​obtained using the least squares method. ； The characteristic normal vector n(a,b,c) is obtained by solving the problem, and then the normal vector n of each point cloud in point Q inside the test pit is obtained. Qi The angle θ between the vector and the eigenvector n; ④ Select the surface point cloud P as the seed point cloud set, and select a point P. i As a seed point, a search is conducted within point Q inside the test pit with a search radius r0 to form a point cloud set N. q Meanwhile, the angle θ between the normal vector and the feature vector n of each point is calculated. If N q If there are no points in Q with θ greater than the threshold θ', then it proves that all points in Q belong to the ground surface. Therefore, delete N from Q. q Meanwhile, increase r0 and continue searching in point Q inside the test pit; If N q If there are points in P with an included angle θ greater than the threshold θ', it proves that some of these points belong to the interior of the test pit. Therefore, delete N from P. p And select a new point P. i Repeat the above steps until P is empty, then the point cloud data of the test pit surface can be obtained.

2. The unmanned automatic compaction degree detection method based on three-dimensional laser scanning according to claim 1, characterized in that: The volume of the test pit is determined by calculating the segmented point cloud data using the slicing method. The specific steps are as follows: ① Cut the point cloud data inside the test pit from top to bottom using n+1 equally spaced horizontal planes to obtain a series of horizontal point cloud slices S i The following formula is used for calculation; ； ； In the formula, h is the spacing value; H represents the maximum height of the point cloud inside the test pit; n is the number of slices; x, y, z are the coordinates of the point cloud points; Construct a square on the slice that can contain all points, with its diagonal and focus O. i Let Q be any point on the slice, and let Q be an endpoint. i Constructing ray O i Q i Using this as the initial scan line, a counter-clockwise scan is performed, connecting the scanned points sequentially to generate the boundary contour polygon M. i (i=0,1,…,n); ② Use the following formula to calculate polygon M i Area A i Calculation ； In the formula x i y i M is the outer contour polygon of the point cloud in the slice plane. i Vertex m of (i=0,1,…,n) j The coordinates of (j=0,1,…,k); j is the vertex number of the polygon of the outer contour of the point cloud slice; i is the number of the point cloud slice. ③ The volume of each segment of the test pit is calculated using the following formula and then summed to obtain the total volume V of the test pit. ； In the formula A i The outer contour polygon M of the point cloud in the slice plane i The area; h is the spacing value; ④ The mass of all solid particles inside the test pit is estimated by estimating the mass of solid particles in the soil sample obtained by drying. Combined with the calculated volume V, the dry density of the soil sample is calculated using the following formula. ； In the formula: ρ d This refers to the dry density of the soil sample. M represents the mass of solid particles inside the test pit; The vehicle-mounted display terminal will display the calculation parameters, calculation results, and test pit model.

Images