Foundation pit support surface feature detection method and system

Abstract

The present application provides a method and system for detecting surface features of foundation pit support. The method for detecting surface features of foundation pit support first installs a laser radar and an inertial sensor on a tripod platform, and adjusts the tripod to a horizontal level according to the output information of the inertial sensor; then performs preprocessing to eliminate noise interference and extract the point cloud of the wall to be measured; then performs fine measurement of the wall point cloud; then performs plane fitting on the point cloud data of each interval to obtain a plane model; finally, obtains the wall flatness information, calculates the wall inclination angle, and calculates the distance between the laser radar origin and the plane model. The surface feature detection system for foundation pit support includes a detection unit, a control unit, a human-computer interaction unit, and a tripod, which are respectively used to collect point cloud data of the wall to be measured, process the original point cloud data of the above-mentioned detection unit, calculate and output the surface feature data of the foundation pit support, and the human-computer interaction unit is used for real-time display and data recording of the surface features of the foundation pit support, and is used to install the detection unit.

Description

Technical Field

[0001] The present invention relates to the technical field of foundation pit support surface feature detection, and in particular to a foundation pit support surface feature detection method and system. Background Art

[0002] To ensure construction quality during tunnel construction, it is necessary to inspect the surface characteristics of the foundation pit support. These characteristics primarily include the flatness, inclination, and distance between the two walls of the foundation pit's sprayed concrete walls. To inspect the flatness and inclination of the foundation pit's sprayed concrete walls, tools such as levels, three-meter rulers, and MERLIN beams are primarily used. Cross-sectional inspection methods are used to directly measure the longitudinal section, and the flatness and inclination are output based on the longitudinal section curve. A ruler is typically used to inspect the distance between the two walls. While these methods can detect the surface characteristics of the pit support, they suffer from slow speed, low efficiency, and large errors.

[0003] In summary, current technology cannot meet the requirements of high-precision detection of foundation pit support surface features. Therefore, there is an urgent need for a fast, efficient, and high-precision foundation pit support surface feature detection method and system to solve the problems in traditional measurement methods. Summary of the Invention

[0004] The purpose of the present invention is to solve the problems of slow speed, low efficiency and large error in the traditional surface feature detection method of foundation pit support in the above background, and propose a surface feature detection method of foundation pit support, which includes the following steps:

[0005] S1. Install the lidar and inertial sensor on the tripod head, and adjust the tripod to a horizontal level based on the output information of the inertial sensor.

[0006] S2, using the straight-through filtering and outlier filtering algorithms for preprocessing to eliminate noise interference and extract the point cloud of the wall to be measured;

[0007] S3, divide the wall point cloud into n intervals with equal intervals, and perform fine measurement on each interval;

[0008] S4, using the random sampling consensus algorithm to perform plane fitting on the point cloud data of n intervals to obtain a plane model;

[0009] S5. By counting the distance between the original point cloud and the plane model, the wall flatness information is obtained. The angle θ between the corresponding vector V1 of the inertial sensor's gravity direction in the lidar coordinate system and the plane model's normal vector V2 is calculated, which is the wall inclination angle. The distance between the lidar origin and the plane model is calculated, which is the distance from the lidar to one side of the wall.

[0010] In the preferred solution, step S2 is specifically as follows: collecting the original point cloud data P output by the laser radar x,y,z(t), and perform through-filtering and outlier filtering to eliminate noise interference. The coordinates of the nth point cloud at time t are P(x, y, z) n (t), the processing method of the straight-through filtering is:

[0011]

[0012] Where, It is the point cloud data after straight-through filtering.

[0013] The distances of all points in the point cloud obey Gaussian distribution, and the distance S from the nth point cloud to the mth point cloud at time t is n for:

[0014]

[0015] The mean of the Gaussian distribution is represented by the average distance μ from each point to its k neighbors:

[0016]

[0017] The standard deviation is represented by the standard deviation σ of the distance from each point to its k neighbors:

[0018]

[0019] The outlier filtering method is as follows:

[0020]

[0021] Where, It is the point cloud data after outlier filtering.

[0022] In the preferred solution, step S4 specifically includes: performing plane fitting on the point cloud data of n intervals using a random sampling consensus algorithm to obtain the following plane model:

[0023] A i x+B i y+C i z+D i =0i=1,2,..,n,

[0024] Where A i 、B i 、C i 、D i is the parameter of plane i;

[0025] Any point on the wall The flatness calculation method is:

[0026]

[0027] Where, s is the flatness of the wall to be measured, d n for point Distance to the fitting plane;

[0028] In the preferred solution, in step S5, when calculating the wall inclination angle, firstly, the three-axis acceleration information is obtained based on the inertial sensor built into the radar: (ACC x ,ACC y ,ACC z ), the pitch angle and yaw angle of the current attitude are:

[0029]

[0030] roll=atan2(ACC y ,ACC z );

[0031] The radar data is then calibrated according to the Pitch and Roll parameter values, where the calibrated point cloud coordinates (x′, y′, z′) are:

[0032] x′=xcos({pitch})+zsin({pitch}),

[0033] y′=ycos({roll})-xsin({pitch})sin({roll})+zcos({pitch})sin({roll}),

[0034] z′=ysin({roll})+xsin({pitch})cos({roll})-zcos({pitch})cos({roll});

[0035] Assume that the fitted plane model is A i x′+B i y′+C i z′+D i = 0, plumb line direction vector a = (0,0,1), plane normal vector b = (A i ,B i ,C i ), then the angle between vectors a and b is:

[0036]

[0037] Under normal circumstances, the wall is perpendicular to the ground plane, so the calculation method of the wall inclination angle φ is:

[0038] φ=90-θ.

[0039] In the preferred solution, let the fitting plane model be A i x+Bi y+C i z+D i = 0, then the distance d from the laser radar to the wall o for:

[0040]

[0041] The present invention also provides a foundation pit support surface feature detection system, which includes a detection unit, a control unit, a human-computer interaction unit and a tripod;

[0042] The detection unit is used to collect point cloud data of the wall to be tested;

[0043] The control unit is used to process the original point cloud data of the above detection unit, calculate and output the surface feature data of the foundation pit support;

[0044] The human-computer interaction unit is used for real-time display and data recording of foundation pit support surface characteristics;

[0045] The tripod is used to mount the detection unit.

[0046] In the preferred solution, the detection unit consists of an inertial sensor and a lidar and is powered by a battery. The inertial sensor is used to calculate the corresponding vector of the gravity direction in the lidar coordinate system. The lidar is used to collect point cloud data within the detection range and transmit the data to the control unit via TCP. The data transmission frequency is 1 / 3 Hz.

[0047] In the preferred solution, the control unit includes an industrial computer, which is used to process the detection data transmitted by the detection unit. First, the point cloud data is pre-processed, and then the wall point cloud partitions are divided into fine-grained measurements. Then, a plane model is established based on the point cloud data. Finally, the surface feature detection results of the foundation pit support are calculated and output to the human-computer interaction unit.

[0048] In the preferred solution, the human-computer interaction unit is mainly composed of a touch screen, which displays and records the characteristics of the foundation pit support surface in real time;

[0049] The detection unit is installed on the tripod's gimbal and uses the output information of the inertial sensor to adjust the gimbal to a horizontal level;

[0050] The tripod is placed on the bottom of the foundation pit.

[0051] In the preferred solution, the excavation process of the foundation pit is from top to bottom. When the excavation reaches a certain depth, steel supports will be added to the wall surface.

[0052] The detection range in the elevation direction is from the bottom of the foundation pit to the nearest steel support above, and the horizontal detection range is from the detection system to the two walls.

[0053] The beneficial effects of the present invention are as follows: the present invention uses a laser radar to collect point cloud data of the sprayed wall area of ​​the foundation pit in real time and establish a plane model, and outputs surface feature data of the foundation pit support based on the plane model and inertial sensor data, which can realize real-time detection of the surface features of the foundation pit support, and has the characteristics of fast, efficient, and high precision. BRIEF DESCRIPTION OF THE DRAWINGS

[0054] Figure 1 It is a cross-sectional view of the foundation pit;

[0055] Figure 2 This is the structural diagram of the foundation pit support surface feature detection system;

[0056] Figure 3 It is a detection flow chart of the surface feature detection method of foundation pit support.

[0057] In the figure, there are the wall to be measured 1, steel support 2, foundation pit bottom 3, inertial sensor 4, lidar 5, battery 6, tripod 7, control unit 8, and touch screen 9. DETAILED DESCRIPTION

[0058] Example 1

[0059] In order to better understand the present invention, the content of the present invention is further illustrated below with reference to examples, and the detection method is not limited to this embodiment. Figure 1 The figure shows a cross-sectional view of the foundation pit, which includes the wall surface 1 to be tested, the steel support 2 and the foundation pit bottom surface 3. The construction process of the foundation pit excavation is from top to bottom. When the excavation reaches a certain depth, the steel support 2 will be added to the wall surface. The foundation pit support surface feature detection system proposed in this invention is always placed on the foundation pit bottom surface 3. The detection range in the elevation direction is from the foundation pit bottom surface 3 to the nearest steel support above. The longitudinal detection range is x meters, and the lateral detection range is from the above detection system to the two walls. The foundation pit support surface feature detection system proposed in this article is as follows Figure 2 As shown, it mainly consists of a detection unit, a control unit, a human-computer interaction unit and a tripod 7. The principles of each unit are explained as follows:

[0060] Detection unit: The detection unit consists of an inertial sensor 4 and a lidar 5 and is powered by a battery 6. The inertial sensor 4 is used to calculate the corresponding vector of the gravity direction in the lidar coordinate system. The lidar 5 is used to collect point cloud data within the detection range and transmit the data to the control unit via TCP. The data transmission frequency is 1 / 3 Hz.

[0061] 2. Control unit: The control unit is composed of an industrial computer 8, which is used to process the detection data transmitted by the detection unit. First, the point cloud data is pre-processed, and then the wall point cloud is partitioned for fine measurement. Then, a plane model is established based on the point cloud data. Finally, the surface feature detection results of the foundation pit support are calculated and output to the human-computer interaction unit.

[0062] 3. Human-computer interaction unit: The human-computer interaction unit is mainly composed of a touch screen 9, which is used for real-time display and data recording of surface characteristics of foundation pit support.

[0063] 4. Mounting bracket: The detection unit is mounted on the pan-tilt platform of the tripod 7. The pan-tilt platform can be adjusted to a horizontal level based on the output information of the inertial sensor 4. The tripod is always placed on the bottom of the foundation pit.

[0064] Figure 3 Shown is a detection flow chart of the above-mentioned foundation pit support surface feature detection method, which introduces the detection method for any wall surface in the foundation pit.

[0065] First, the tripod 7 is adjusted to be horizontal according to the output information of the inertial sensor;

[0066] Secondly, collect the original point cloud data P output by the laser radar x,y,z (t), and perform through-filtering and outlier filtering to eliminate noise interference. Assume that the coordinates of the nth point cloud at time t are P(x, y, z) n (t), the processing method of straight-through filtering is shown in formula (1):

[0067]

[0068] Where, It is the point cloud data after straight-through filtering.

[0069] Assuming that the distances of all points in the point cloud obey Gaussian distribution, the distance S from the nth point cloud to the mth point cloud at time t is n for:

[0070]

[0071] The mean of the Gaussian distribution is represented by the average distance μ from each point to its k neighbors:

[0072]

[0073] The standard deviation is represented by the standard deviation σ of the distance from each point to its k neighbors:

[0074]

[0075] The outlier filtering processing method is shown in formula (5).

[0076]

[0077] Where, It is the point cloud data after outlier filtering.

[0078] Then, the wall point cloud is divided into n intervals with equal intervals for fine measurement, and the point cloud data of the n intervals are respectively fitted with a random sampling consensus algorithm to obtain a plane model as shown in formula (6):

[0079] A i x+B i y+C i z+D i =0 i=1,2,..,n (6)

[0080] Where A i 、B i 、C i 、D i are the parameters of plane i.

[0081] Any point on the wall The calculation method of flatness is shown in formula (7)-(8):

[0082]

[0083] Where, s is the flatness of the wall to be measured, d n for point Distance to the fitting plane.

[0084] When calculating the wall inclination angle, firstly, the three-axis acceleration information is obtained based on the inertial sensor built into the radar: (ACC x ,ACC y ,ACC z ), the pitch angle (Pitch) and yaw angle (Roll) of the current posture are shown in equations (9)-(10):

[0085]

[0086] roll=atan2(ACC y ,ACC z ) (10)

[0087] Afterwards, the radar data is calibrated according to the Pitch and Roll parameter values, where the calibrated point cloud coordinates (x′, y′, z′) are shown in Equation (11):

[0088]

[0089] Assume that the fitted plane model is A i x′+B i y′+C i z′+D i = 0, plumb line direction vector a = (0, 0, 1), (10) plane normal vector b = (A i ,Bi ,C i ), then the angle between vectors a and b is:

[0090]

[0091] Considering that the wall is perpendicular to the ground plane under normal circumstances, the calculation method of the wall inclination angle φ is shown in formula (13):

[0092] φ=90-θ (13)

[0093] Assume that the fitted plane model is A i x+B i y+C i z+D i = 0, then the distance d from the laser radar to the wall o As shown in formula (14):

[0094]

[0095] Example 2

[0096] The axial force of the support and the displacement of the soil in the foundation pit are a reflection of the stability of the foundation pit structure. If the stability of the foundation pit is affected, it may cause changes in the surface characteristics of the foundation pit, such as uneven settlement and cracks on the wall. If the fluctuation of the internal support axial force causes uneven displacement of the foundation pit support structure, cracks or uneven settlement may occur on the surface of the foundation pit. These changes are a direct reflection of the surface characteristics of the foundation pit. The failure mode of shallow soil, which is mainly vertical deformation and settlement, may cause uneven stress distribution and possible cracks on the surface of the foundation pit support, affecting the integrity and function of the support structure. If abnormal changes are detected in the monitoring of the internal support axial force and soil displacement, a more detailed inspection of the foundation pit surface may be required to assess potential risks. Therefore, a monitoring module for the axial force and soil displacement of the internal support in the foundation pit can also be introduced into the foundation pit support surface feature detection system of the present invention. The influence of the axial force and soil displacement of the internal support in the foundation pit on the surface characteristics of the foundation pit serves as an early warning mechanism for the safety of foundation pit construction. The specific implementation method includes the following steps:

[0097] Add monitoring of foundation pit support surface characteristics (such as cracks, uneven settlement, tilt, etc.) to the monitoring system to ensure simultaneous collection of internal support axial force and soil displacement data;

[0098] When conducting data analysis, the internal support axial force and soil displacement data are considered, and the correlation between these data and the surface characteristics of the foundation pit support is analyzed to identify potential surface deformation risks;

[0099] Incorporate changes in the surface characteristics of foundation pit support into the early warning system. When abnormal changes in surface characteristics are detected, an early warning is triggered, indicating possible internal mechanical problems.

[0100] Develop a comprehensive assessment model that combines data on internal support axial forces, soil displacements, and pit support surface characteristics to more comprehensively assess pit stability and safety;

[0101] The specific guiding significance of changes in the surface characteristics of foundation pit support for construction adjustments should be clearly pointed out in the response measures. For example, the increase in surface cracks may require additional support or changes in construction methods.

[0102] In risk assessment and decision support, the changes in the surface characteristics of foundation pit support are used as an important factor in risk assessment to more accurately judge the construction safety status;

[0103] During the construction adjustment and optimization step, the specific impact of changes in the surface characteristics of the foundation pit support on the construction plan should be considered. For example, uneven surface settlement may require adjustments to the design of the support structure.

[0104] In the reports and records, the changes in the surface characteristics of the foundation pit support and the correlation analysis results of these changes with the internal support axial force and soil displacement data shall be recorded in detail.

[0105] By implementing this plan, it is possible to ensure that the safety of the foundation pit construction process is effectively monitored, and possible risks can be discovered and responded to in a timely manner, thereby improving construction quality and safety.

[0106] The above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit the same. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the various embodiments of the present invention.

Claims

1. A method for detecting surface characteristics of foundation pit support, characterized by: The following steps are involved: S1. Install the lidar and inertial sensor on the tripod head, and adjust the tripod to a horizontal level based on the output information of the inertial sensor. S2, using the straight-through filtering and outlier filtering algorithms for preprocessing to eliminate noise interference and extract the point cloud of the wall to be measured; S3, divide the wall point cloud into n intervals with equal intervals, and perform fine measurement on each interval; S4, using the random sampling consensus algorithm to perform plane fitting on the point cloud data of n intervals to obtain a plane model; S5. By counting the distance between the original point cloud and the plane model, the wall flatness information is obtained. The angle θ between the corresponding vector V1 of the inertial sensor's gravity direction in the lidar coordinate system and the plane model's normal vector V2 is calculated, which is the wall inclination angle. The distance between the lidar origin and the plane model is calculated, which is the distance from the lidar to one side of the wall.

2. The method for detecting surface characteristics of foundation pit support according to claim 1, characterized in that: Step S2 is specifically: collecting the original point cloud data P output by the laser radar x,y,z (t), and perform through-filtering and outlier filtering to eliminate noise interference. The coordinates of the nth point cloud at time t are P(x, y, z) n (t), the processing method of the straight-through filtering is: Where, Point cloud data after straight-through filtering; The distances of all points in the point cloud obey Gaussian distribution, and the distance S from the nth point cloud to the mth point cloud at time t is n for: The mean of the Gaussian distribution is represented by the average distance μ from each point to its k neighbors: The standard deviation is represented by the standard deviation σ of the distance from each point to its k neighbors: The outlier filtering processing method is: Where, It is the point cloud data after outlier filtering.

3. The method for detecting surface characteristics of foundation pit support according to claim 1, characterized in that: Step S4 is specifically as follows: using the random sampling consensus algorithm to perform plane fitting on the point cloud data of n intervals to obtain the following plane model: A i x+B i y+C i z+D i =0i=1,2,..,n, Where A i 、B i 、C i 、D i is the parameter of plane i; Any point on the wall The flatness calculation method is: Where, s is the flatness of the wall to be measured, d n for point Distance to the fitting plane.

4. The method for detecting surface characteristics of foundation pit support according to claim 1, wherein: In step S5, when calculating the wall inclination angle, the three-axis acceleration information is first measured based on the inertial sensor built into the radar: (ACC x ,ACC y ,ACC z ), the pitch angle and yaw angle of the current attitude are: roll=atan2(ACC y ,ACC z ); The radar data is then calibrated according to the Pitch and Roll parameter values, where the calibrated point cloud coordinates (x′, y′, z′) are: x′=xcos({pitch})+zsin({pitch}), y′=ycos({roll})-xsin({pitch})sin({roll})+zcos({pitch})sin({roll}), z′=ysin({roll})+xsin({pitch})cos({roll})-zcos({pitch})cos({roll}); Assume that the fitted plane model is A i x′+B i y′+C i z′+D i = 0, plumb line direction vector a = (0,0,1), plane normal vector b = (A i ,B i ,C i ), then the angle between vectors a and b is: Under normal circumstances, the wall is perpendicular to the ground plane, so the calculation method of the wall inclination angle φ is: φ=90-θ.

5. The surface feature detection method of foundation pit support according to claim 4 is characterized in that: The fitted plane model is A i x+B i y+C i z+D i = 0, then the distance d from the laser radar to the wall o for:

6. A foundation pit support surface feature detection system, characterized by: The foundation pit support surface feature detection system includes a detection unit, a control unit, a human-computer interaction unit and a tripod (7); The detection unit is used to collect point cloud data of the wall surface (1) to be tested; The control unit is used to process the original point cloud data of the detection unit, calculate and output the surface feature data of the foundation pit support; The human-computer interaction unit is used for real-time display and data recording of foundation pit support surface characteristics; The tripod (7) is used to install the detection unit.

7. The foundation pit support surface feature detection system according to claim 6, characterized in that: The detection unit is composed of an inertial sensor (4) and a laser radar (5) and is powered by a battery (6). The inertial sensor (4) is used to calculate the corresponding vector of the gravity direction in the laser radar coordinate system. The laser radar (5) is used to collect point cloud data within the detection range and transmit the data to the control unit through TCP. The data transmission frequency is 1 / 3 Hz.

8. The foundation pit support surface feature detection system according to claim 7, characterized in that: The control unit includes an industrial computer (8) for processing the detection data transmitted by the detection unit, first pre-processing the point cloud data, then partitioning the wall point cloud for fine measurement, then establishing a plane model based on the point cloud data, and finally calculating and outputting the surface feature detection results of the foundation pit support to the human-computer interaction unit.

9. The foundation pit support surface feature detection system according to claim 6, characterized in that: The human-computer interaction unit includes a touch screen (9) for real-time display and data recording of features of the foundation pit support surface; The detection unit is mounted on a pan / tilt platform of a tripod (7), and the pan / tilt platform is adjusted to be horizontal according to the output information of the inertial sensor 4; The tripod (7) is placed on the bottom surface of the foundation pit (3).

10. The foundation pit support surface feature detection system according to claim 6, characterized in that: The construction process of foundation pit excavation is from top to bottom. When the excavation reaches a certain depth, steel supports are added to the wall (2); The detection range in the elevation direction is from the bottom of the foundation pit (3) to the nearest steel support above, and the horizontal detection range is from the detection system to two opposite walls.

Images