Multi-source data fusion detection vehicle and detection method for shield tunnel underneath passing through south-to-north water transfer trunk canal

Abstract

The invention discloses a multi-source data fusion detection vehicle and detection method for a shield tunnel underneath passing through a south-to-north water transfer trunk canal. An industrial camera, a laser radar, an acceleration sensor, an angular velocity sensor and a data acquisition device are arranged on the detection vehicle; the industrial camera and the laser radar collect metro shield tunnel appearance data, and the metro shield tunnel appearance data is subjected to fusion processing through the data collection device to detect segment cracks, water leakage, segment staggering, convergence and deformation appearance diseases of the metro shield tunnel; the acceleration sensor is arranged on a bearing frame of a bogie of the detection vehicle and collects the acceleration of the two bearing frames in the vertical direction, the amplitude, phase difference and cross correlation coefficient of two acceleration change curves are analyzed through the data collection device, and whether local settlement occurs in a subway shield tunnel or not and whether a ballast bed is separated from a tunnel segment or not are judged. The local settlement amplitude of the metro shield tunnel is calculated through the secondary integration of the acceleration signal of the local settlement section to the time, so that the technical problem of comprehensive and rapid detection of metro shield tunnel diseases is solved.

Description

Technical Field

[0001] The present invention relates to the technical field of shield tunnel disease detection equipment, and particularly relates to a multi-source data fusion detection vehicle and a detection method for a shield tunnel passing under the South-to-North Water Diversion Main Canal. Background Art

[0002] During the operation of subway shield tunnels, due to the influence of geological conditions, external loads (such as adjacent engineering construction), and groundwater fluctuations, diseases such as segment cracks, water leakage, offset, convergence deformation, uneven settlement, and separation between the track bed and segments often occur, seriously threatening the safe operation of the subway. Especially when the subway tunnel passes under major water conservancy projects such as the South-to-North Water Diversion Main Canal, the change of water and soil pressure during the construction and operation period of the main canal may further exacerbate the risk of tunnel structure deformation. For example, when the South-to-North Water Diversion Main Canal passes under Beijing Subway's Wukesong Station, the ground disturbance induced by the construction caused the settlement of the existing station structure, and it was necessary to use information monitoring technology to real-time feedback deformation data to ensure safety. Such engineering cases show that the detection of tunnel diseases in complex environments needs to consider both apparent damage and hidden structure responses, and also consider the synergistic effects of multiple factor interferences.

[0003] In the prior art, "Research on Rapid Detection Method for Tunnel Apparent Diseases" proposed a multi-sensor integration method based on a CCD array camera, a laser scanner, an inertial navigation, and an angular velocity sensor, and achieved high-precision measurement of tunnel lining images and laser point clouds through dynamic collaboration. However, laser scanning control points for assisting inertial navigation to improve the absolute position accuracy were introduced into the tunnel. As a result, when the tunnel undergoes uneven settlement, the spatial positions of the laser scanning control points will also change accordingly. However, during the actual detection process of tunnel apparent diseases, since the positioning and attitude solution calculation of the spatial position change of the laser scanning control points was not introduced into the detection process, the detection results actually filtered out the uneven settlement diseases of the subway shield tunnel. At the same time, in this article, a method for detecting the separation disease between the track bed and the tunnel segments was not revealed either.

[0004] Another research, "Intelligent Identification of Uneven Settlement of Subway Tunnels Based on Vehicle-Track Vibration", analyzed the relationship between the vehicle body vibration signal and tunnel settlement through a vehicle-track dynamics model, and established an intelligent identification model based on a recurrent neural network. However, this method only relies on the vertical acceleration signal of the vehicle body, cannot quantitatively calculate the settlement value, and does not consider the interference of other structural damages to the vibration signal, making it difficult to adapt to the multi-disease coupling scenarios that may occur in the project of the South-to-North Water Diversion Main Canal passing under, and it also cannot be applied to the detection of the separation disease between the track bed and the tunnel segments.

[0005] In addition, the prior art lacks a multi-source data fusion detection method for the scenario of passing under major water conservancy projects, making it difficult to comprehensively evaluate the health status of the tunnel. Summary of the Invention

[0006] In order to overcome the deficiencies in the background technology, the present invention discloses a multi-source data fusion detection vehicle and a detection method for a shield tunnel passing under the South-to-North Water Diversion Main Canal, which are used to solve the technical problem of comprehensively and quickly detecting the diseases of subway shield tunnels and meet the major technical requirements for the safe operation of subway tunnels.

[0007] In order to achieve the above-mentioned invention purpose, the present invention adopts the following technical solutions: A multi-source data fusion detection vehicle for a shield tunnel passing under the South-to-North Water Diversion Main Canal, including a detection vehicle and an industrial camera, a lidar, an acceleration sensor, an angular velocity sensor, and a data acquisition device arranged on the detection vehicle; Among them, the industrial camera is used to collect the apparent images of the subway shield tunnel; Among them, the lidar is used to collect the apparent three-dimensional laser point cloud of the subway shield tunnel; Among them, the acceleration sensor is arranged on the bearing bracket of the bogie of the detection vehicle and is used to collect the acceleration in the vertical direction of the bearing bracket; Among them, the data acquisition device is used to control the operation of the industrial camera, lidar, and acceleration sensor, and collect, store, and process the data of the industrial camera, lidar, and acceleration sensor.

[0008] Furthermore, the detection vehicle includes a bogie and a car body, and the car body is arranged on the upper part of the bogie; the bogie is of a single-wheel pair structure and there are two; the bogie includes a wheel pair, a bearing bracket, a frame, and a car body connecting beam; the wheel pair is rotationally connected to the bearing bracket, the bearing bracket is connected to the frame through a primary suspension, and the frame is connected to the car body connecting beam through a secondary suspension and a traction frame; the acceleration sensor is fixedly arranged on the bearing bracket.

[0009] Furthermore, a sensor rigid frame is fixedly arranged on the car body; the sensor rigid frame includes an industrial camera support frame and a lidar support frame, and the industrial camera support frame is fixedly connected to the lidar support frame; there are several industrial cameras, and several industrial cameras are arranged in a semi-circular array on the industrial camera support frame and are parallel to the cross-section of the subway shield tunnel; the lidar is fixedly arranged on the lidar support frame, and its rotation scanning direction is parallel to the cross-section of the subway shield tunnel.

[0010] Furthermore, the angular velocity sensor is fixedly arranged on the bearing end cover on the bearing bracket.

[0011] Furthermore, several lighting sources are also arranged in an array on the industrial camera support frame.

[0012] Furthermore, a counterweight is also arranged on the car body.

[0013] A detection method for a multi-source data fusion detection vehicle of a shield tunnel passing under the main canal of the South-to-North Water Diversion Project. Three laser targets with known spatial coordinates are provided on both side walls and the top wall of the subway shield tunnel to be detected at several preset sections. When detecting the diseases of the subway shield tunnel, the detection vehicle travels along the subway track at a set speed. The lidar scans the subway shield tunnel with laser, several industrial cameras take pictures of the subway shield tunnel, and the acceleration sensor collects the acceleration signals of the bearing frame in the vertical direction. Among them, the lidar obtains the original spatial movement trajectory coordinates and attitude through its own inertial navigation system, performs lidar point cloud calculation on the lidar scan data, the original spatial movement trajectory coordinates and attitude, and obtains the original lidar point cloud data. In the original point cloud data, according to the lidar point cloud intensity information, the preset spatial coordinates of the three laser targets at several preset sections are extracted. According to the preset spatial coordinates of the three laser targets at several preset sections, the spatial movement trajectory coordinates, attitude and original point cloud data of the lidar are corrected for errors, and finally the corrected spatial movement trajectory coordinates, attitude and point cloud data of the lidar are obtained. Among them, the corrected point cloud data is processed to obtain the three-dimensional model of the subway shield tunnel. It is judged whether convergence, deformation and misalignment occur according to the three-dimensional model of the subway shield tunnel.

[0014] Furthermore, the lidar obtains its original spatial movement trajectory coordinates and attitude through its own inertial navigation system, performs lidar point cloud calculation on the lidar scan data, the original spatial movement trajectory coordinates and attitude, and obtains the original lidar point cloud data. In the original point cloud data, according to the lidar point cloud intensity information, the preset spatial coordinates of the three laser targets at several preset sections are extracted. According to the preset spatial coordinates of the three laser targets at several preset sections, the spatial movement trajectory coordinates, attitude and original point cloud data of the lidar are corrected for errors, and finally the corrected spatial movement trajectory coordinates, attitude and point cloud data of the lidar are obtained. Among them, the corrected point cloud data is processed to obtain the three-dimensional model of the subway shield tunnel. It is judged whether convergence, deformation and misalignment occur according to the three-dimensional model of the subway shield tunnel.

[0015] Furthermore, several industrial cameras obtain the spatial movement trajectory coordinates and attitude of several industrial cameras according to the corrected spatial movement trajectory coordinates and attitude of the lidar and their installation position relationship with the lidar. Several industrial cameras map the captured images onto the three-dimensional model of the subway shield tunnel according to the spatial movement trajectory coordinates and attitude, and obtain the three-dimensional apparent image of the subway shield tunnel. The three-dimensional apparent image of the subway shield tunnel is analyzed to determine whether cracks and water leakage occur.

[0016] Furthermore, the acceleration sensors on the two bearing frames collect acceleration signals to obtain two acceleration change curves in the vertical direction of the subway shield tunnel; by analyzing the amplitude, phase difference, and cross-correlation of the two acceleration change curves, it is determined whether local settlement has occurred in the subway shield tunnel and whether separation has occurred between the roadbed and the tunnel segment; the local settlement position and length of the subway shield tunnel are calculated through the correspondence between the acceleration change curve and the spatial movement trajectory after laser radar correction; the local settlement amplitude of the subway shield tunnel is calculated by quadratically integrating the acceleration signal in the vertical direction of the local settlement section over time.

[0017] Due to the adoption of the technical scheme as described above, the present invention has the following beneficial effects: the multi-source data fusion inspection vehicle for shield tunnels passing under the South-to-North Water Diversion Canal disclosed by the present invention is equipped with an industrial camera, a laser radar, an acceleration sensor, an angular velocity sensor, and a data acquisition device on the inspection vehicle; when the multi-source data fusion inspection vehicle for shield tunnels passing under the South-to-North Water Diversion Canal is working, it travels along the subway track at a set speed, and the industrial camera and the laser radar collect the apparent data of the subway shield tunnel, which is fused and processed by the data acquisition device to detect the apparent defects of the tube segment cracks, water leakage, misalignment, convergence, and deformation of the subway shield tunnel; wherein the acceleration sensor is arranged on the bearing frame of the bogie of the inspection vehicle, and collects the data of the two bearings. The acceleration in the vertical direction of the frame is measured, and the amplitude, phase difference, and mutual correlation coefficient of the two acceleration change curves are analyzed by the data acquisition device to determine whether local settlement of the subway shield tunnel has occurred and whether the roadbed and the tunnel segment have been separated. This can adapt to the multi-disease coupling scenarios that may occur in the South-to-North Water Diversion Canal Underpass Project; the local settlement position and length of the subway shield tunnel are calculated through the correspondence between the acceleration change curve and the spatial movement trajectory after the lidar correction; the local settlement amplitude of the subway shield tunnel is calculated by the secondary integration of the acceleration signal of the local settlement section over time, which can more comprehensively evaluate the health status of the tunnel, thereby solving the technical problem of comprehensive and rapid detection of subway shield tunnel diseases. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] Figure 1 This is a schematic diagram of the appearance of the multi-source data fusion detection vehicle for the shield tunnel that passes under the South-to-North Water Diversion Canal; Figure 2 This is a schematic diagram of the appearance of the bogie; Figure 3 This is a schematic diagram of the track model for the uneven settlement of a subway shield tunnel; Figure 4 Schematic diagram of the vehicle-track model when the roadbed and tunnel segment are separated.

[0019] In the figure: 1. Detection vehicle; 1.1 Bogie; 1.1.1 Wheel set; 1.1.2 Bearing housing; 1.1.3 Frame; 1.1.4 Car body connecting beam; 1.2 Car body; 3 Sensor rigid frame; 3.1 Industrial camera support frame; 3.2 LiDAR support frame; 4 Industrial camera; 5 LiDAR; 6 Acceleration sensor; 8 Data acquisition device; 9 Counterweight; 10 Track; 11 Ballast bed; 12 Tunnel segment; 13 Soil outside the tunnel. Detailed implementation mode

[0020] The present invention can be explained in detail through the following embodiments, and the purpose of disclosing the present invention is to protect all technical improvements within the scope of the present invention.

[0021] See the attached instruction Figure 1 、 2 : A multi-source data fusion detection vehicle for a shield tunnel passing under the South-to-North Water Diversion Main Canal, comprising a detection vehicle 1, an industrial camera 4, a LiDAR 5, an acceleration sensor 6, an angular velocity sensor 7, and a data acquisition device 8; the detection vehicle 1 includes two bogies 1.1 and a car body 1.2, the bogie 1.1 includes a wheel set 1.1.1, a bearing housing 1.1.2, a frame 1.1.3, and a car body connecting beam 1.1.4, the bearing housing 1.1.2 is rotatably arranged at both ends of the wheel set 1.1.1, the frame 1.1.3 is connected and arranged above the bearing housing 1.1.2 through a primary suspension, the car body connecting beam 1.1.4 is connected and arranged above the frame 1.1.3 through a secondary suspension and a traction frame, and the car body 1.2 is fixedly arranged above the car body connecting beam 1.1.4; the distance between the two bogies 1.1 (the contact point of the wheel set 1.1.1 and the rail) is 3 m; two counterweights 9 are also fixedly arranged on the car body 1.2, and the two counterweights 9 are respectively located above the left and right bogies 1.1, and the weight of the counterweight 9 cooperates with the primary suspension and secondary suspension of the bogie 1.1 to reduce the vibration of the detection vehicle car body 1.2 when walking along the subway track at a set speed, improve the cumulative error of the inertial navigation system of the LiDAR 5 itself, and at the same time ensure that the bogie 1.1 will not bounce when passing through the local settlement section of the subway shield tunnel and the section where the ballast bed is separated from the tunnel segment, and ensure the accuracy of the data collected by the acceleration sensor 6; a driving motor is arranged on one or two bogies 1.1 to drive the multi-source data fusion detection vehicle for the shield tunnel passing under the South-to-North Water Diversion Main Canal to automatically walk along the subway track; On the upper right end of the vehicle body 1.2, a sensor rigid frame 3 is fixedly arranged. The bogie 1.1 close to the sensor rigid frame 3 is the front bogie, and the bogie 1.1 far from the sensor rigid frame 3 is the rear bogie; the sensor rigid frame 3 includes an industrial camera support frame 3.1 and a lidar support frame 3.2. The industrial camera support frame 3.1 is semi-circular in shape, and the lidar support frame 3.2 is fixedly arranged at the right end of the industrial camera support frame 3.1 and is located at the center of the industrial camera support frame 3.1 to reduce the influence of the roll and pitch of the vehicle body 1.2 on the spatial pose of the lidar support frame 3.2; a number of industrial cameras 4 are provided, and the number of industrial cameras 4 is arranged in a semi-circular array on the industrial camera support frame 3.1 and is parallel to the cross-section of the subway shield tunnel; the lidar 5 is fixedly arranged on the outer end face of the lidar support frame 3.2, and its rotational scanning direction is parallel to the cross-section of the subway shield tunnel; a number of lighting sources are also arranged in an array on the industrial camera support frame 3.1 to supplement light for the industrial camera 4 during shooting; The acceleration sensor 6 is fixedly arranged directly above the bearing bracket 1.1.2 on one side of the bogie 1.1 and is aligned with the axis of the wheel set 1.1.1; in this inspection vehicle, the bogie 1.1 adopts a single-wheel set structure, and the bearing bracket 1.1.2 is a symmetric structure, avoiding the interference of the asymmetric bearing bracket rotating around the wheel set 1.1.1 on the data acquisition of the acceleration sensor 6 arranged on it when the bogie 1.1 adopts a double-wheel set structure design; the angular velocity sensor 7 is arranged on the bearing end cover of the bearing bracket 1.1.2 and is used to detect the real-time running speed of the vehicle during operation. By integrating the real-time running speed with respect to time, the driving mileage of the inspection vehicle can be obtained; The data acquisition device 8 is arranged on the vehicle body 1.2; the data acquisition device 8 is electrically connected to the industrial camera 4, the lidar 5, the acceleration sensor 6, and the angular velocity sensor 7.

[0022] A detection method for a multi-source data fusion detection vehicle for a shield tunnel passing under the South-to-North Water Diversion Main Canal. On the two side walls and the top wall of the subway shield tunnel to be detected at several preset cross-sections, three laser targets with known spatial coordinates are provided; when detecting the diseases of the subway shield tunnel, the detection vehicle travels along the subway track at a speed of 30 Km / h. The lidar 5 performs laser scanning on the subway shield tunnel, a number of industrial cameras 4 take pictures of the subway shield tunnel, and the acceleration sensor 6 collects data on the acceleration of the bearing bracket 1.1.2 of the bogie 1.1 in the vertical direction; The lidar 5 obtains its original spatial movement trajectory coordinates and attitude through its own inertial navigation system, performs lidar point cloud calculation on the lidar 5 scan data, original spatial movement trajectory coordinates and attitude, and obtains the original point cloud data of the lidar 5; in the original point cloud data, according to the lidar point cloud intensity information, extract the preset spatial coordinates of three laser targets at several preset sections, and according to the preset spatial coordinates of three laser targets at several preset sections, correct the errors of the lidar 5 spatial movement trajectory coordinates, attitude, original point cloud data and driving mileage, and finally obtain the corrected lidar 5 spatial movement trajectory coordinates, attitude, point cloud data and the driving mileage of the detection vehicle; among them, the corrected point cloud data is processed to obtain a three-dimensional model of the subway shield tunnel; analyze the three-dimensional model of the subway shield tunnel to judge whether the subway shield tunnel has convergence, deformation and misalignment of segments; when the acceleration sensor 6 collects the acceleration signal in the vertical direction of the bearing housing 1.1.2, mark the time stamp output by the lidar 5 on the time axis of the acceleration-time curve; According to the corrected spatial movement trajectory coordinates and attitude of the lidar 5, several industrial cameras 4 obtain the spatial movement trajectory coordinates and attitude of several industrial cameras 4 through their installation position relationship with the lidar 5; several industrial cameras 4 map and splice the captured images on the three-dimensional model of the subway shield tunnel according to the spatial movement trajectory coordinates and attitude, and obtain the three-dimensional apparent image of the subway shield tunnel; analyze the three-dimensional apparent image of the subway shield tunnel to judge whether there are cracks and water leakage in the segments of the subway shield tunnel; The acceleration sensors 6 on the front and rear bearing housings 1.1.2 collect acceleration signals, and correspondingly obtain the acceleration-time curves of the bearing housings 1.1.2 of the front and rear bogies 1.1 (actually the wheels on the corresponding axles 1.1.1) in the longitudinal direction of the subway shield tunnel; when the multi-source data fusion detection vehicle for the shield tunnel passing under the South-to-North Water Diversion Main Canal travels at a speed of 30 Km / h, there is a phase difference of 0.36 s between the acceleration-time curves corresponding to the front and rear bogies 1.1; Analyze the amplitude, phase difference and cross-correlation of the two acceleration change curves. The steps are as follows: S1: Calculate the average fluctuation change range of the acceleration change curve of the front bearing housing 1.1.2, and its calculation formula is:

[0023]

[0024] where is the average fluctuation change range of positive acceleration, n is the number of positive (or negative) acceleration extreme values in the acceleration-time curve, is the nth positive acceleration extreme value; where is the average fluctuation change range of negative acceleration, is the nth negative acceleration extreme value; S2: With a time window of 0.12 s (the width of the time window is 1 / 3 of the phase difference of the acceleration-time curves corresponding to the front and rear bogies 1.1), start scanning from the starting end of the acceleration-time curve of the front bearing housing 1.1.2, and calculate the average fluctuation value of the acceleration change curve within the time window. The calculation formula is as follows:

[0025]

[0026] Where is the average positive acceleration within the time window, m is the number of positive (or negative) acceleration extreme values within the time window, is the m-th positive acceleration extreme value within the time window; where is the average negative acceleration within the time window, is the m-th negative acceleration extreme value within the time window; S3: When or appears within the time window, lock the first time window; then start from the rear end of the first time window and continue to perform time window scanning along the two acceleration-time curves. When or appears within the time window again, lock the second time window; repeat the above process until or appears within three consecutive time windows, and several connected locked time windows are obtained; merge all the connected locked time windows to obtain the acceleration curve characteristic window of the front bearing housing; S4: Start from the rear end of the previous acceleration curve characteristic window of the front bearing housing and continue to perform time window scanning on the acceleration-time curve of the front bearing housing until the entire acceleration-time curve of the front bearing housing is scanned, and finally N acceleration curve characteristic windows of the front bearing housing are obtained; S5: Copy the finally obtained N acceleration curve characteristic windows of the front bearing housing to the acceleration-time curve corresponding to the rear bogie 1.1 and translate them 0.36 seconds to the right to obtain N corresponding acceleration curve characteristic windows of the rear bearing housing; S6: Perform cross-correlation analysis on the acceleration-time curves within the acceleration curve characteristic window of the front bearing housing and the corresponding acceleration curve characteristic window of the rear bearing housing in sequence. If the correlation coefficient , it is determined that there is local settlement at the subway shield tunnel corresponding to the acceleration curve characteristic window of the front bearing housing; if the correlation coefficient , it is determined that there is local settlement at the subway shield tunnel corresponding to the acceleration curve characteristic window of the front bearing housing and detachment occurs between the ballast bed and the tunnel segment; if the correlation coefficient , it is determined that there is a detachment between the track bed and the tunnel segment at the subway shield tunnel corresponding to the characteristic window of the front bearing housing acceleration curve; Calculation of the settlement position and length of the subway shield tunnel: The time scale output by the lidar 5 is marked on the time axis of the acceleration-time curve of the subway shield tunnel longitudinally; based on the time scale on the time axis of the acceleration signal curve, any moment on the time axis of the acceleration signal curve is mapped onto the spatial movement trajectory of the lidar 5, and then through the spatial position relationship between the lidar 5 and the wheel, the spatial position of the wheel at any moment on the time axis of the acceleration signal curve is calculated; after determining that there is local settlement at the subway shield tunnel corresponding to the characteristic window of the front bearing housing acceleration curve, the spatial position of the wheel calculated through the starting point of the characteristic window of the front bearing housing acceleration curve is the starting position of the local settlement, and the spatial position of the wheel calculated through the ending point of the characteristic window of the front bearing housing acceleration curve is the ending point position of the local settlement, so as to obtain the local settlement position and length of the subway shield tunnel; Calculation of the settlement amplitude of the subway shield tunnel: After determining that there is local settlement in the subway shield tunnel corresponding to the characteristic window of the front bearing housing acceleration curve, the acceleration signal within the characteristic window of the front bearing housing acceleration curve is integrated twice with respect to time to calculate the local settlement amplitude of the subway shield tunnel.

[0027] See the attached instruction Figure 3, this figure shows the dynamic model of the inspection vehicle and the track when local settlement occurs in the subway shield tunnel but there is no detachment between the track bed and the tunnel segment; in the article [Intelligent Identification of Uneven Settlement of Subway Tunnels Based on Vehicle-Track Vibration], it is determined that the vertical acceleration of the vehicle body is the most suitable characteristic dynamic response for identifying the uneven settlement of the tunnel. However, due to the influence of the transfer functions of the primary suspension and secondary suspension between the wheel set 1.1.1 and the vehicle body 1.2, the vertical acceleration of the vehicle body is not the vertical acceleration of the wheel set 1.1.1 when it travels along the track 10, but the output of the vertical acceleration of the wheel set 1.1.1 when it travels along the track 10 through the transfer functions of the primary suspension and secondary suspension. After being filtered by the primary suspension and secondary suspension, the high-frequency components of the vertical acceleration of the wheel set 1.1.1 when it travels along the track 10 are filtered out. Therefore, the vertical acceleration signal of the vehicle body can better reflect the characteristics of the local settlement of the subway shield tunnel, that is, the vertical acceleration of the vehicle body can be used as a sensitive index for detecting the uneven settlement of the tunnel. However, due to the complexity of the dynamic model of the inspection vehicle and the track, it is impossible to accurately establish the transfer functions of the primary suspension and secondary suspension. Therefore, the vertical acceleration signal of the vehicle body cannot be used to accurately calculate the value of the uneven settlement of the tunnel; in this patent, the acceleration sensor 6 is set on the bearing bracket 1.1.2, eliminating the influence of the transfer functions of the primary suspension and secondary suspension between the wheel set 1.1.1 and the vehicle body 1.2. The acceleration data collected by the acceleration sensor 6 is exactly the same as the vertical acceleration of the wheel set 1.1.1 when it travels along the track 10. Therefore, it can be used for the accurate calculation of the value of the uneven settlement of the tunnel; Supplementary note: When the acceleration sensor 6 is set on the bearing bracket 1.1.2 to collect the vertical acceleration of the wheel set 1.1.1 when it travels along the track 10 for calculating the value of the uneven settlement of the tunnel, the influence of the track dynamic model on the data collected by the acceleration sensor 6 is actually not considered. However, considering that the stiffness and damping of the track dynamic model are relatively high (regarded as a rigid body), its influence on the final calculation result of the uneven settlement of the tunnel can be ignored; See the attached drawings of the specification Figure 4, this figure shows the dynamic model of the inspection vehicle and the track when detachment occurs between the ballast bed and the tunnel segment; when detachment occurs between the ballast bed and the tunnel segment, the voided track 10 is actually in the state of a slender beam fixed at both ends; before the front bogie of the inspection vehicle enters the voided track 10, the voided track 10 is in a straight state; when the front bogie of the inspection vehicle enters the voided track 10, the voided track 10 undergoes bending deformation; when the front bogie of the inspection vehicle leaves the voided track 10, since the rear bogie of the inspection vehicle is still on the voided track 10, the voided track 10 remains in a bent deformation state; therefore, when the front bogie of the inspection vehicle passes over the voided track 10, the state of the voided track 10 is: straight - bent - bent; by the same analysis, when the rear bogie of the inspection vehicle passes over the voided track 10, the state of the voided track 10 is: bent - bent - straight; since the states of the voided track 10 are different when the front bogie and the rear bogie of the inspection vehicle pass over the voided track 10, the acceleration changes given to the wheels on the front and rear bogies by the voided track 10 are inconsistent, so the correlation of the acceleration signals of the bearing brackets 1.1.2 of the front and rear bogies is relatively low; while when local settlement occurs in the subway shield tunnel but detachment does not occur between the ballast bed and the tunnel segment, regarding the track as a rigid body, when the front and rear bogies of the inspection vehicle pass through the local settlement area, the acceleration changes given to the wheels on the front and rear bogies by the track in the settlement area are consistent, so the correlation of the acceleration signals of the bearing brackets 1.1.2 of the front and rear bogies is relatively high; based on the above, it is possible to determine whether local settlement occurs in the subway shield tunnel or whether detachment occurs between the ballast bed and the tunnel segment by performing cross-correlation analysis on the acceleration - time curves within the corresponding characteristic windows of the acceleration curves of the front bearing bracket and the rear bearing bracket.

[0028] The parts not detailed in the present invention are prior art.

[0029] Those skilled in the art should understand that those skilled in the art can realize variation examples in combination with the prior art and the above embodiments, and such variation examples do not affect the essence of the present solution, so they will not be elaborated here.

[0030] It should be understood that the present solution is not limited to the above specific implementation manners, and the structures and construction methods not detailed herein should be understood to be implemented in a common manner in the art; any person skilled in the art, without departing from the scope of the technical solution of the present solution, can make many possible changes and modifications to the technical solution of the present solution by using the methods and technical contents disclosed above, or modify it into an equivalent embodiment with equivalent changes, which does not affect the essence of the present solution. Therefore, any simple modification, equivalent change and modification made to the above embodiments according to the technical essence of the present solution without departing from the content of the technical solution of the present solution still belong to the scope of protection of the technical solution of the present solution.

Claims

1. A multi-source data fusion detection vehicle for a shield tunnel under the South-to-North Water Diversion Canal, characterized by: It comprises an inspection vehicle (1) and an industrial camera (4), a laser radar (5), an acceleration sensor (6), an angular velocity sensor (7), and a data acquisition device (8) arranged on the inspection vehicle (1); Among them, the industrial camera (4) is used to collect the surface image of the subway shield tunnel; Among them, the laser radar (5) is used to collect the apparent three-dimensional laser point cloud of the subway shield tunnel; The acceleration sensor (6) is arranged on the bearing frame (1.1.2) of the bogie (1.1) of the inspection vehicle (1) and is used to collect the vertical acceleration of the bearing frame (1.1.2); The data acquisition device (8) is used to control the operation of the industrial camera (4), the laser radar (5), and the acceleration sensor (6), and to acquire, store, and process data from the industrial camera (4), the laser radar (5), and the acceleration sensor (6).

2. According to claim 1, the multi-source data fusion detection vehicle for the shield tunnel passing under the South-to-North Water Diversion Canal is characterized by: The inspection vehicle (1) comprises a bogie (1.1) and a vehicle body (1.2), wherein the vehicle body (1.2) is arranged on the upper part of the bogie (1.1); the bogie (1.1) is a single wheel pair structure, and is provided with two wheels; the bogie (1.1) comprises a wheelset 1.1.1, a bearing frame (1.1.2), a frame (1.1.3), and a vehicle body connecting beam (1.1.4); the wheelset 1.1.1 is rotatably connected to the bearing frame (1.1.2), the bearing frame (1.1.2) and the frame (1.1.3) are connected via a primary suspension, and the frame (1.1.3) and the vehicle body connecting beam (1.1.4) are connected via a secondary suspension and a traction frame; and the acceleration sensor (6) is fixedly arranged on the bearing frame (1.1.2).

3. According to claim 2, the multi-source data fusion detection vehicle for the shield tunnel passing under the South-to-North Water Diversion Canal is characterized by: A sensor rigid frame (3) is fixedly arranged on the vehicle body (1.2); the sensor rigid frame (3) comprises an industrial camera support frame (3.1) and a laser radar support frame (3.2), and the industrial camera support frame (3.1) and the laser radar support frame (3.2) are fixedly connected; a plurality of industrial cameras (4) are provided, and the plurality of industrial cameras (4) are arranged on the industrial camera support frame (3.1) in a semi-circular array, parallel to the cross section of the subway shield tunnel; and the laser radar (5) is fixedly arranged on the laser radar support frame (3.2), and its rotation scanning direction is parallel to the cross section of the subway shield tunnel.

4. According to claim 1, the multi-source data fusion detection vehicle for the shield tunnel under the South-to-North Water Diversion Canal is characterized by: The angular velocity sensor (7) is fixedly arranged on a bearing end cover on the bearing frame (1.1.2).

5. According to claim 3, the multi-source data fusion detection vehicle for the shield tunnel passing under the South-to-North Water Diversion Canal is characterized by: A plurality of illumination light sources are also arranged in an array on the industrial camera support frame (3.1).

6. The multi-source data fusion detection vehicle for a shield tunnel passing under the South-to-North Water Diversion Canal according to claim 1 is characterized by: A counterweight block (9) is also provided on the vehicle body (1.2).

7. A detection method based on a multi-source data fusion detection vehicle for a shield tunnel passing under the South-to-North Water Diversion Canal according to any one of claims 1 to 6, characterized in that: Three laser targets with known spatial coordinates are arranged at a plurality of preset sections on the two side walls and the top wall of the subway shield tunnel to be inspected; when inspecting the defects of the subway shield tunnel, the inspection vehicle travels along the subway track at a set speed, the laser radar (5) performs laser scanning on the subway shield tunnel, a plurality of industrial cameras (4) photograph the subway shield tunnel, and the acceleration sensor (6) collects acceleration signals of the bearing frame (1.1.2) in the vertical direction; wherein the laser radar (5) obtains a three-dimensional model of the subway shield tunnel by laser scanning the subway shield tunnel, which is used for detecting convergence, deformation and misalignment of the subway shield tunnel; wherein the industrial camera (4) photographs the subway shield tunnel to obtain an apparent image of the subway shield tunnel, which is used for detecting cracks and water leakage in the subway shield tunnel segments; wherein the acceleration sensor (6) collects acceleration signals of the bearing frame (1.1.2) in the vertical direction, and obtains an acceleration variation curve of the bearing frame (1.1.2) in the vertical direction of the subway shield tunnel, which is used for detecting the settlement of the subway shield tunnel and the separation of the track structure from the tunnel segments.

8. The detection method of the shield tunnel multi-source data fusion detection vehicle under the South-to-North Water Diversion Canal according to claim 7 is characterized by: The laser radar (5) obtains its original spatial movement trajectory coordinates and posture through its own inertial navigation system, performs laser point cloud solution on the laser radar (5) scanning data and the original spatial movement trajectory coordinates and posture, and obtains the laser radar (5) original point cloud data; in the original point cloud data, according to the laser point cloud intensity information, extracts the preset spatial coordinates of three laser targets at a plurality of preset sections; according to the preset spatial coordinates of the three laser targets at the plurality of preset sections, performs error correction on the laser radar (5) spatial movement trajectory coordinates, posture and original point cloud data, and finally obtains the laser radar (5) corrected spatial movement trajectory coordinates, posture and point cloud data; wherein the corrected point cloud data is processed to obtain a three-dimensional model of a subway shield tunnel; Determine whether convergence, deformation and misalignment occur based on the three-dimensional model of the subway shield tunnel.

9. The detection method of the shield tunnel multi-source data fusion detection vehicle under the South-to-North Water Diversion Canal according to claim 7 is characterized by: The plurality of industrial cameras (4) obtain the spatial movement trajectory coordinates and postures of the plurality of industrial cameras (4) based on the corrected spatial movement trajectory coordinates and postures of the laser radar (5) and the installation position relationship between the plurality of industrial cameras (4) and the laser radar (5); the plurality of industrial cameras (4) map the images taken by the plurality of industrial cameras (4) onto the three-dimensional model of the subway shield tunnel based on the spatial movement trajectory coordinates and postures to obtain a three-dimensional surface image of the subway shield tunnel; and the three-dimensional surface image of the subway shield tunnel is analyzed to determine whether cracks or water leakage occur.

10. The detection method of the shield tunnel multi-source data fusion detection vehicle under the South-to-North Water Diversion Canal according to claim 7, characterized in that: The acceleration sensors (6) on the two bearing frames (1.1.2) collect acceleration signals to obtain two acceleration change curves in the vertical direction of the subway shield tunnel; by analyzing the amplitude, phase difference, and cross-correlation of the two acceleration change curves, it is determined whether local settlement has occurred in the subway shield tunnel and whether separation has occurred between the roadbed (11) and the tunnel segment (12); the local settlement position and length of the subway shield tunnel are calculated through the correspondence between the acceleration change curve and the corrected spatial movement trajectory of the laser radar (5); and the local settlement amplitude of the subway shield tunnel is calculated by quadratically integrating the acceleration signal in the vertical direction of the local settlement section over time.

Images