Tunnel steel belt grouting fullness intelligent detection robot

Abstract

The application discloses a tunnel steel belt grouting fullness intelligent detection robot, and relates to the field of tunnel nondestructive testing technology. The robot is magnetically attracted to walk and carries double excitation points to impact the same position in turn. Through the fusion of waveguide transfer function analysis and operation modal analysis, continuous detection of steel belt grouting fullness and intelligent identification of defect types are realized. The robot is driven by the magnetic walking and spatial positioning module to automatically and continuously walk along the steel belt. The double excitation point impact generation module applies standard impact to the same physical position on the steel belt in turn. The motion characteristics of the robot and the signal sensing process are deeply fused. A spatial and spectral joint sensing architecture based on the steel belt elastic waveguide effect is constructed. The structural defects of the prior art, such as dependence on manual point arrangement, mutual independence of data of each measuring point and inability to form continuous detection results, are solved. The spatial aliasing and missed detection risks caused by discrete sampling are avoided.

Description

Technical Field

[0001] This invention relates to the field of tunnel non-destructive testing technology, specifically to an intelligent robot for detecting the fullness of grouting in tunnel steel strips. Background Technology

[0002] Tunnel lining steel strip reinforcement is a new technology that has been widely used in recent years for tunnel defect treatment and structural reinforcement. Its principle involves fixing a steel strip circumferentially or spirally along the inner wall of the lining, and then injecting high-strength grout between the steel strip and the lining. This creates a composite load-bearing structure between the steel strip and the original lining, significantly improving the load-bearing capacity and durability of the tunnel lining. Grout fullness is a key indicator for evaluating the effectiveness of steel strip reinforcement, directly affecting the overall safety and service life of the reinforced structure.

[0003] Currently, the methods for detecting the grout fullness of tunnel lining steel strips are mainly divided into three categories:

[0004] (1) Manual tapping and listening method: The inspector taps the surface of the steel strip with a small hammer and judges whether the grouting is full by listening to the difference in sound. This method is highly subjective, cannot be quantified, and the test results vary from person to person. In addition, it cannot be used to test areas that are difficult for personnel to reach, such as the tunnel arch.

[0005] (2) Core drilling method: Drill holes in the steel strip and take out core samples to observe the grouting situation. This method is a destructive test, which will damage the reinforced structure, and can only be used for sampling inspection, which cannot reflect the grouting quality of the entire steel strip.

[0006] (3) Instrumental testing method: Represented by 'CN105136907A', "An intelligent testing system and method for grout compaction based on the flat test method", this method uses the impact echo principle for testing. Although this system achieves a certain degree of intelligence, it has the following fundamental defects:

[0007] First, the static discrete point detection architecture is based on the workflow of "point selection → point placement → point-by-point detection". There is no data correlation between the measurement points, and it is impossible to generate a spatial distribution profile of grouting quality along the steel strip direction, which poses a serious risk of spatial overlap and missed detection.

[0008] Second, the single-parameter determination model relies solely on two parameters: the initial acoustic time of the reflected wave and the diffracted wave. This completely loses the spectral and modal information contained in the signal and cannot distinguish between different types of grouting defects.

[0009] Third, the calibration dependency is serious: the void estimation model is calibrated under specific laboratory conditions. When the test object changes from the corrugated pipe in the laboratory to the grouting body of the tunnel steel strip, the original calibration model completely fails and the generalization ability of the scenario is extremely poor.

[0010] Therefore, there is an urgent need to develop an intelligent detection system for the fullness of grouting in tunnel lining steel strips that can achieve continuous detection throughout the entire section, multi-dimensional information fusion, and automatic identification of defect types. To this end, we propose an intelligent detection robot for the fullness of grouting in tunnel steel strips. Summary of the Invention

[0011] The purpose of this invention is to provide an intelligent robot for detecting the fullness of grouting in tunnel steel strips.

[0012] To achieve the above objectives, the present invention provides the following technical solution: an intelligent robot for detecting the fullness of grouting in tunnel steel strips, the robot comprising:

[0013] Magnetic walking and spatial positioning module: used to drive the entire robot to walk automatically and continuously along the surface of the tunnel lining steel strip, and to acquire the robot's position information and walking speed information on the steel strip in real time, and transmit the position information and walking speed information to the dual excitation point impact generation module;

[0014] Dual-excitation point impact generation module: Arranged at fixed intervals along the walking direction of the steel strip, it is used to apply standard impact force to the same physical position on the steel strip one after another according to the received walking speed information, and at the same time generate an impact trigger signal and transmit it to the signal acquisition and preprocessing module to provide a synchronous time reference for the acquisition and interception of vibration response signals.

[0015] Multi-channel vibration sensor array module: arranged between the dual-excitation point impact generation modules, used to synchronously acquire the vibration response signal generated by the excitation of the steel strip, and transmit the vibration response signal to the signal acquisition and preprocessing module;

[0016] Signal acquisition and preprocessing module: It is used to receive the impact trigger signal and use the trigger time as the time zero point to synchronously acquire and mark the vibration response signal, and then complete the amplification, filtering and analog-to-digital conversion processing. The preprocessed vibration signal with a unified time reference is simultaneously transmitted to the waveguide transfer function analysis module and the operating mode analysis and defect identification module.

[0017] Waveguide transfer function analysis module: Based on the received pre-processed vibration signal, it calculates the frequency domain transfer characteristics of vibration energy propagating in the composite waveguide section of steel strip and grout between two excitation points, generates waveguide transfer function data, and transmits the waveguide transfer function data to the grouting constraint stiffness profile reconstruction module.

[0018] Grouting constraint stiffness profile reconstruction module: used to generate continuous grouting constraint stiffness spatial distribution data along the entire length of the steel strip based on the received waveguide transfer function data, and transmit the grouting constraint stiffness spatial distribution data to the fullness grading and visualization module;

[0019] The modal analysis and defect identification module is used to extract the running mode information of the steel strip based on the received preprocessed vibration signal, calculate the modal confidence and identify the defect type, generate defect type and modal confidence data, and transmit the defect type and modal confidence data to the fullness grading and visualization module.

[0020] Fullness Grading and Visualization Module: This module integrates the received spatial distribution data of grouting constraint stiffness, as well as defect type and modal confidence data, to perform a three-level fullness grading and generate a visual inspection report.

[0021] As a further embodiment of the present invention: the magnetic walking and spatial positioning module includes a Halbach array permanent magnet adsorption unit, a servo drive wheel set, a flexible suspension mechanism, a steel belt edge tracking sensor, an odometer, an IMU inertial measurement unit, and a steel belt feature recognition unit;

[0022] The Halbach array permanent magnet adsorption unit is used to provide adsorption force so that the robot can be adsorbed onto the surface of the steel strip.

[0023] The steel strip edge tracking sensor senses the position of the steel strip edge in real time and feeds it back to the walking control unit to adjust the robot's direction of travel, so that the robot always stays on the center line of the steel strip.

[0024] The steel strip feature recognition unit identifies the inherent features of the steel strip, such as overlapping joints, welding points, and fixing bolts, based on the periodic structural noise features in the vibration signal collected by the multi-channel vibration sensor array module. These features are used as absolute position reference points to periodically correct the cumulative error of the odometer, achieving centimeter-level steel strip position coding accuracy.

[0025] As a further aspect of the present invention: the dual-excitation point impact generation module includes a front exciter and a rear exciter, each exciter including an impact hammer driven by a voice coil motor, a force sensor and a trigger control circuit;

[0026] The trigger control circuit determines the real-time walking speed based on the magnetic walking and spatial positioning module. Calculate the trigger time difference between the front and rear actuators. The two actuators are triggered in sequence, so that the front actuator and the rear actuator apply standard impact force to the same physical position on the steel strip in turn.

[0027] The force sensor is used to detect the impact force amplitude in real time and feed it back to the trigger control circuit for closed-loop control to maintain the consistency of excitation energy for each event.

[0028] in, This represents the real-time linear velocity of the robot as it moves along the steel belt, expressed in m / s. This represents the time interval (in seconds) between the front and rear actuators applying impact forces to the same location, satisfying the following condition: , This is the fixed distance between the front and rear actuators.

[0029] As a further aspect of the present invention: the multi-channel vibration sensing array module includes at least three acceleration sensors, which are spaced equally along the steel belt direction between the front exciter and the rear exciter. The accelerometers are arranged such that each accelerometer is tightly coupled to the surface of the steel strip being measured via a magnetic mount.

[0030] The frequency response range of the accelerometer covers 100Hz to 20kHz, and the sensitivity is not less than 100mV / g;

[0031] All accelerometers synchronously acquire vibration response signals generated by the excitation of the steel strip at a sampling rate of not less than 51.2 kS / s;

[0032] in:

[0033] The distance between the centers of two adjacent accelerometers is expressed in meters (m). The sampling rate is expressed in kS / s, which represents thousands of samples per second, i.e., the number of vibration signal sample points collected per second.

[0034] As a further aspect of the present invention: the specific working method of the waveguide transfer function analysis module is as follows: for each axial position of the steel strip... The time-domain window of the vibration response recorded by the sensor array after the exciter is triggered is captured. Time-domain window of vibration response recorded by sensor array after pre-exciter triggering ;

[0035] Perform short-time Fourier transforms on the two time-domain windows respectively to obtain the corresponding complex spectra. ;

[0036] Calculate position waveguide transfer function at The calculation formula is:

[0037] ;

[0038] in, This represents the axial position coordinates of the steel strip, corresponding one-to-one with the robot's current detection position, with the starting end of the steel strip as the origin, and the unit is meters (m). Time is expressed in seconds (s). Indicates the position of the post-exciter. After being triggered, the sensor array is at the location The time-domain signal window of the recorded vibration response. Indicates the position of the front actuator. After being triggered, the sensor array is at the location The time-domain signal window of the recorded vibration response. The frequency of the vibration signal is expressed in Hz. Indicates to The complex spectrum obtained after performing a short-time Fourier transform Indicates to The complex spectrum obtained after performing a short-time Fourier transform Indicates the position of the steel strip Location, frequency The waveguide transfer function describes the vibrational energy across a length of . The frequency domain propagation characteristics of the waveguide propagating in the composite waveguide section of steel strip and grouting body, including amplitude attenuation. Two components.

[0039] As a further aspect of the present invention, the specific working method of the grouting constraint stiffness profile reconstruction module is as follows:

[0040] Based on waveguide transfer function amplitude Calculate the position of the steel strip Grouting constraint stiffness index The calculation formula is:

[0041] ;

[0042] in, This indicates the lower limit of the sensitive frequency band for grout fullness testing, in Hz, and is taken as 0.5 times the first-order bending resonance frequency of the steel strip. This indicates the upper limit of the sensitive frequency band for grout fullness testing, in Hz, and is taken as twice the third-order bending resonance frequency of the steel strip. Indicates the frequency range [ , Definite integral operations within ] This indicates the grouting constraint stiffness index at position x of the steel strip, in N / m. 3 This value is used to quantitatively characterize the constraint strength of the grout body on the vibration of the steel strip; a larger value indicates a more complete grouting. The frequency domain weighting function is a dimensionless quantity, obtained through prior calibration experiments on standard specimens of steel strips and grout bodies with known saturation states. The calculation formula is:

[0043] ;

[0044] in, The amplitude of the waveguide transfer function of the standard specimen under fully compacted conditions. The waveguide transfer function amplitude of the standard specimen in a completely detached state is calculated by the grouting constraint stiffness profile reconstruction module based on the value obtained by the robot traveling along the steel strip. The sequence outputs a continuous curve of grouting constraint stiffness and position.

[0045] As a further aspect of the present invention, the specific working method of the operational modal analysis and defect identification module is as follows:

[0046] Operational modal analysis was performed on the N-channel vibration signals acquired by the multi-channel vibration sensor array module. The random subspace identification (SSI) algorithm was used to extract the forward motion of the steel strip under continuous robot excitation. First-order running mode vector ;

[0047] Calculate the modal confidence criterion values ​​between the measured mode shapes and the standard dense mode shapes of each order. The calculation formula is:

[0048] ;

[0049] Where N represents the total number of accelerometers in the multi-channel vibration sensing array, and the dimension of the operating mode vector is the same as the total number of accelerometers, which is N, a positive integer, and N≥3. The order of the extracted mode shape is a positive integer. ≤N, Indicates the order of the mode shape. , Indicates the first The measured mode shape vector is an N-dimensional complex vector. This indicates the first pre-marked section on a known fully compacted steel strip and grouting standard section. The first-order reference mode vector is an N-dimensional complex vector. express The conjugate transpose of . express The conjugate transpose of ;

[0050] It is a dimensionless quantity with a value range of [0,1]. The closer the value is to 1, the more similar the measured vibration mode is to the standard dense vibration mode.

[0051] when When <0.85, the mode is considered to have an anomaly, and based on... Identify defect types by the variation pattern of values ​​along the position of the steel strip:

[0052] like The value is generally below 0.85 and is evenly distributed along the steel strip, indicating insufficient overall grouting strength;

[0053] like A sudden drop in value at a local location (the change in value is greater than 0.3 within a length of 10 cm) is identified as a local void.

[0054] like The value shows a gradual drift trend along the direction of the steel strip, which is determined to be the density gradient of the grouting body.

[0055] As a further aspect of the present invention: the three-level fullness grading determination logic of the fullness grading determination and visualization module is as follows:

[0056] When the grouting constraint stiffness index at a certain location of the steel strip ≥ First threshold And all the previous Modal confidence When all values ​​are ≥0.90, the grouting fullness at that location is determined to be Grade I compact, and no warning is triggered.

[0057] When the grouting constraint stiffness index at a certain location of the steel strip ≥ second threshold And < first threshold And / or there exists a modal confidence level of no more than 2nd order. When the value is <0.85, the grouting fullness at this location is judged to be Grade II, slightly non-dense, triggering a yellow warning and recommending grouting reinforcement.

[0058] When the grouting constraint stiffness index at a certain location of the steel strip <Second threshold And / or there is a modal confidence level of 3 or higher. When the value is less than 0.85, the grouting fullness at this location is determined to be a Class III severe defect, triggering a red alert and recommending the removal of the steel strip and re-grouting.

[0059] in, This represents the threshold between Grade I dense and Grade II slightly loose, expressed in N / m. 3 Take the Class I dense section The 5th quantile of the value distribution. This represents the threshold between Grade II slight loosening and Grade III severe defects, expressed in N / m. 3 Take the 95th percentile of the K-value distribution of the Level III defect segment. A quantile is a value that is at a certain percentage position after a set of data is sorted from smallest to largest.

[0060] As a further aspect of the present invention: the visualization detection report generated by the fullness grading and visualization module includes:

[0061] A chromatogram of the spatial distribution of grout fullness along the entire length of the steel strip, with green representing Class I dense areas, yellow representing Class II slightly loose areas, and red representing Class III severely defective areas;

[0062] A detailed statistical table of each defective segment, including the start and end locations, length, average grouting constraint stiffness value, minimum modal confidence value, and recommended remedial measures;

[0063] Basic testing information includes the steel strip number, testing time, testing personnel, system configuration parameters, and testing environmental conditions.

[0064] Compared with the prior art, the beneficial effects of the present invention by adopting the above technical solution are as follows:

[0065] 1. This invention uses a magnetic walking and spatial positioning module to drive a robot to automatically and continuously walk along a steel strip. In conjunction with a dual-excitation point impact generation module, standard impacts are applied to the same physical location on the steel strip one after another. This deeply integrates the robot's motion characteristics with the signal perception process, and constructs a spatial and spectral joint perception architecture based on the elastic waveguide effect of the steel strip. This solves the structural defects of existing technologies that rely on manual point placement, independent data from each measuring point, and inability to form continuous detection results. It avoids the spatial aliasing and missed detection risks caused by discrete sampling. It can generate a spatial distribution profile of grouting quality along the entire length of the steel strip, clearly showing the start and end positions and transition characteristics of defects, and providing comprehensive spatial dimension information for engineering quality assessment.

[0066] 2. This invention calculates the frequency domain transmission characteristics of vibration energy in the composite waveguide section of steel strip and grout body through the waveguide transfer function analysis module, and combines the grout constraint stiffness profile reconstruction module to perform weighted fusion of multi-frequency vibration information. The judgment criteria are changed from the traditional single-point acoustic time parameter to the spatial propagation attenuation characteristics of vibration energy. This solves the problem that the existing technology only extracts the first wave acoustic time and systematically discards the spectrum and modal information. It breaks through the information bottleneck of the single-parameter judgment model, effectively reduces misjudgment caused by signal interference or local anomalies, and significantly improves the accuracy and reliability of grout fullness detection.

[0067] 3. This invention uses a relative measurement method that involves two excitation points impacting the same location sequentially. It calculates the waveguide transfer function by utilizing the vibration response generated at the same location by the two excitations. It uses relative attenuation characteristics instead of absolute response characteristics as the basis for judgment. This solves the problem that the existing technology's detection model is deeply bound to laboratory calibration conditions and cannot adapt to the differences in materials and environments in different engineering sites. It significantly reduces the system's dependence on pre-calibration and eliminates the need for complex model recalibration for different steel strip specifications and grouting material ratios in different engineering scenarios. It has stronger scenario generalization ability and engineering practicality.

[0068] 4. This invention extracts the operating vibration modes of the composite structure of steel strip and grout body through the modal analysis and defect identification module, calculates the modal confidence of the measured vibration mode and the standard compacted vibration mode, and performs two-dimensional cross-validation and three-level classification judgment in combination with the grouting constraint stiffness index. This solves the problem that the existing technology can only infer the size of voids and cannot distinguish different types of grouting defects. It can automatically identify a variety of defect types such as insufficient overall grouting strength, local voids and grout density gradient, and provide corresponding classification and treatment suggestions, providing more targeted technical support for engineering maintenance decisions. Attached Figure Description

[0069] Figure 1 This is a schematic diagram of the system flow in an embodiment of the present invention;

[0070] Figure 2 This is a schematic diagram illustrating the working principle steps of the system in an embodiment of the present invention. Detailed Implementation

[0071] The specific embodiments of the present invention will be further described below with reference to the accompanying drawings. It should be noted that the description of these embodiments is for the purpose of helping to understand the present invention, but does not constitute a limitation of the present invention.

[0072] Furthermore, the technical features involved in the various embodiments of the present invention described below can be combined with each other as long as they do not conflict with each other.

[0073] Please see the appendix Figure 1 -Appendix Figure 2 This invention discloses an intelligent detection robot for the fullness of grouting in tunnel steel strips. The robot includes:

[0074] Magnetic walking and spatial positioning module: used to drive the entire robot to walk automatically and continuously along the surface of the tunnel lining steel strip, and to acquire the robot's position information and walking speed information on the steel strip in real time, and transmit the position information and walking speed information to the dual excitation point impact generation module;

[0075] Dual-excitation point impact generation module: Arranged at fixed intervals along the walking direction of the steel strip, it is used to apply standard impact force to the same physical position on the steel strip one after another according to the received walking speed information, and at the same time generate an impact trigger signal and transmit it to the signal acquisition and preprocessing module to provide a synchronous time reference for the acquisition and interception of vibration response signals.

[0076] Multi-channel vibration sensor array module: arranged between the dual-excitation point impact generation modules, used to synchronously acquire the vibration response signal generated by the steel strip under excitation, and transmit the vibration response signal to the signal acquisition and preprocessing module;

[0077] Signal acquisition and preprocessing module: It is used to receive the impact trigger signal and use the trigger time as the time zero point to synchronously acquire and mark the vibration response signal, and then complete the amplification, filtering and analog-to-digital conversion processing. The preprocessed vibration signal with a unified time reference is simultaneously transmitted to the waveguide transfer function analysis module and the operating mode analysis and defect identification module.

[0078] Waveguide transfer function analysis module: Based on the received pre-processed vibration signal, it calculates the frequency domain transfer characteristics of vibration energy propagating in the composite waveguide section of steel strip and grout between two excitation points, generates waveguide transfer function data, and transmits the waveguide transfer function data to the grouting constraint stiffness profile reconstruction module.

[0079] Grouting constraint stiffness profile reconstruction module: It is used to generate continuous grouting constraint stiffness spatial distribution data along the entire length of the steel strip based on the received waveguide transfer function data, and transmit the grouting constraint stiffness spatial distribution data to the fullness grading and visualization module.

[0080] The modal analysis and defect identification module is used to extract the running mode information of the steel strip based on the received preprocessed vibration signal, calculate the modal confidence and identify the defect type, generate defect type and modal confidence data, and transmit the defect type and modal confidence data to the fullness grading and visualization module.

[0081] Fullness Grading and Visualization Module: This module integrates the received spatial distribution data of grouting constraint stiffness, as well as defect type and modal confidence data, to perform a three-level fullness grading and generate a visual inspection report.

[0082] Example 1

[0083] This embodiment is applicable to the standard tunnel steel strip reinforcement testing scenario with a steel strip width of 200mm and a thickness of 5mm, and the grouting material is CGM high-strength non-shrink grouting material.

[0084] The configuration parameters of the intelligent detection robot system for tunnel steel strip grouting fullness in this embodiment are as follows:

[0085] Dual excitation point spacing =200mm;

[0086] The multi-channel vibration sensing array contains five accelerometers, spaced apart. =50mm;

[0087] robot walking speed =0.05m / s;

[0088] Standard impact force amplitude = 50N;

[0089] Signal sampling rate = 51.2kHz;

[0090] Sensitive frequency bands =500Hz, =15kHz;

[0091] Extracted mode order =3;

[0092] First threshold =2.5×10 6 N / m 3 ;

[0093] Second threshold =1.2×10 6 N / m 3 .

[0094] The testing process is as follows:

[0095] Step 1: System Initialization and Calibration

[0096] Place the robot at the starting end of the steel strip to be inspected and start the system for self-testing. First, perform standard section calibration by acquiring vibration signals on a known fully compacted standard steel strip section and calculating the standard waveguide transfer function. and standard reference mode ;

[0097] Then, vibration signals were collected on a standard steel strip segment that was known to be completely detached, and the transfer function of the detached waveguide was calculated. ;

[0098] Finally, according to the formula Calculate the frequency domain weighting function .

[0099] Step 2: Robot's autonomous walking and impact stimulation

[0100] The magnetic walking and spatial positioning module provides adsorption force through Halbach array permanent magnet adsorption units, enabling the robot to be stably adsorbed on the steel strip surface.

[0101] The steel strip edge tracking sensor detects the position of the steel strip edge in real time and feeds it back to the walking control unit to adjust the direction of travel, ensuring that the robot always stays on the center line of the steel strip and walks at a stable speed of 0.05 m / s. The dual-excitation point impact generation module uses the real-time walking speed provided by the magnetic walking and spatial positioning modules. =0.05m / s, calculate the trigger time difference. 4s.

[0102] The trigger control circuit sequentially triggers the front and rear exciters with a 4-second time difference, causing both exciters to apply a standard impact force of 50N to the same physical location on the steel strip. A force sensor detects the impact force amplitude in real time and feeds it back to the trigger control circuit for closed-loop control, ensuring the consistency of excitation energy for each event.

[0103] Step 3: Vibration signal acquisition and preprocessing

[0104] The five accelerometers of the multi-channel vibration sensor array module synchronously acquire the vibration response signal generated by the excited steel strip at a sampling rate of 51.2kHz. The signal acquisition and preprocessing module receives the impact trigger signal and marks the zero point of time. It amplifies, filters, and performs analog-to-digital conversion on the acquired vibration response signal to remove power frequency interference and high-frequency noise, thus obtaining the preprocessed vibration signal.

[0105] Step 4: Waveguide transfer function calculation for each axial position of the steel strip. The waveguide transfer function analysis module extracts the vibration response time-domain window within 0.02 s after the exciter is triggered. Vibration response time-domain window within 0.02 s after triggering the pre-exciter. .

[0106] Perform short-time Fourier transforms on the two time-domain windows, with a window length of 1024 points and an overlap rate of 50%, to obtain the corresponding complex spectra. .

[0107] According to the formula Calculate position waveguide transfer function at The frequency domain propagation characteristics of vibration energy in a 0.2m long steel strip-grout composite waveguide section were obtained.

[0108] Step 5: Grouting Constraint Stiffness Profile Reconstruction

[0109] The grouting constraint stiffness profile reconstruction module is based on the waveguide transfer function. amplitude Combined with a pre-calibrated frequency domain weighting function Integral calculations were performed within the sensitive frequency band [500Hz, 15kHz] to obtain the grouting constraint stiffness index at position x of the steel strip. :

[0110] ;

[0111] The robot walks along the steel belt for the entire distance, and a calculation is performed at 1cm intervals. The value is used to generate a continuous curve of grouting constraint stiffness-position along the entire length of the steel strip.

[0112] Step 6: Perform modal analysis and defect identification

[0113] The operational modal analysis and defect identification module performs operational modal analysis on the 5-channel vibration signal and uses the Random Subspace Identification (SSI) algorithm to extract the first three operational mode vectors of the steel strip. .

[0114] Calculate the modal confidence criterion values ​​between the measured mode shapes and the standard reference mode shapes of each order:

[0115] ;

[0116] when When <0.85, the mode is determined to be abnormal, and based on... Identifying defect types based on value change patterns:

[0117] If 3rd order The values ​​are all between 0.7 and 0.8 and are evenly distributed along the steel strip, indicating that the overall grouting strength is insufficient;

[0118] If at 12.3m of the steel strip, It suddenly dropped from 0.95 to 0.52. It dropped sharply from 0.92 to 0.48. The value suddenly dropped from 0.89 to 0.55, with the magnitude of the change being greater than 0.3 / 10cm, indicating that there was a local void at this location;

[0119] If the MAC value gradually decreases from 0.92 at 25m of the steel strip to 0.68 at 30m, showing a gradual drift trend, it is determined that there is a grout density gradient in this section.

[0120] Step 7: Three-level fullness grading determination and report generation

[0121] The grouting constraint stiffness index is integrated into the fullness grading and visualization module. and modal confidence Perform a three-level judgment:

[0122] when ≥2.5×10 6 N / m 3 And all 3rd order When the density is ≥0.90, it is judged as Grade I dense.

[0123] When 1.2×10 6 N / m 3 ≤ <2.5×10 6 N / m 3 Or there exists no more than 2 orders When the density is <0.85, it is judged as Grade II, slightly loose.

[0124] when <1.2×10 6 N / m 3 Or there may be 3 or higher order When the value is less than 0.85, it is judged as a Level III serious defect.

[0125] Finally, a spatial distribution chromatogram of grout fullness along the entire length of the steel strip and a detailed test report are generated. The report includes the start and end positions, length, average K value, and minimum value of each defect segment. Values ​​and recommended handling measures.

[0126] This embodiment was verified on-site in the Xiamen Xiang'an Tunnel reconstruction and expansion project. A 50m long steel strip reinforced section was tested, in which a 30cm long local void section and a 50cm long grout density gradient section were artificially set.

[0127] Example 2

[0128] This embodiment is applicable to the reinforcement testing of wide steel strips with a width of 300mm and a thickness of 8mm.

[0129] The difference from Example 1 is as follows:

[0130] The multi-channel vibration sensor array is expanded into a double-row arrangement, with two rows of sensors set in the width direction of the steel strip. Each row contains 5 acceleration sensors along the length direction of the steel strip, for a total of 10 sensors.

[0131] The spacing L between the two excitation points was adjusted to 300mm;

[0132] The standard impact force amplitude has been adjusted to 80N;

[0133] Sensitive frequency bands adjusted to =300Hz, =12kHz;

[0134] The extracted mode order is M=4.

[0135] The remaining system configuration and testing procedures are the same as in Example 1. This example can adapt to the testing needs of wider steel strips, achieving full coverage testing in the width direction of the steel strip through a dual-row sensor array, thus avoiding blind spots caused by excessively wide steel strips.

[0136] This embodiment was also verified on-site in the Xiamen Xiang'an Tunnel reconstruction and expansion project, where a 50m long wide steel strip reinforcement section was tested, including a 30cm long local void section and a 50cm long grout density gradient section.

[0137] Comparative Example

[0138] This comparative example uses the "Intelligent Detection System and Method for Grouting Compactness Based on Flat Measurement Method" disclosed in CN105136907A to test the same 50m long steel strip reinforced section in Example 1.

[0139] The testing process is as follows:

[0140] Step 1: Model Calibration

[0141] Concrete specimens of the same specifications as those on site were prepared in the laboratory. Corrugated pipe channels with five density levels—fully dense, 3 / 4 dense, 1 / 2 dense, 1 / 4 dense, and completely void—were artificially set up to establish a void estimation model.

[0142] Step 2: Wave velocity calibration

[0143] Impact echo velocity was calibrated using standard concrete specimens with the same strength as the concrete specimens and prepared at the same time. =3800m / s.

[0144] Step 3: On-site site selection and data collection

[0145] A total of 5000 measuring points were randomly placed along the longitudinal direction of the steel strip surface at 1cm intervals. Coupling agent was applied to each measuring point, and data was collected using the planar measurement method. Three data points were collected at each measuring point, and the average was taken to obtain the first wave time of the reflected wave. and the first wave of diffraction .

[0146] Step 4: Void Size Calculation. Based on the void estimation model, by... and Calculate the void radius R at each measuring point, and output the grouting density judgment result by comparing it with the five density levels.

[0147] Comparison of test results

[0148]

[0149] As shown in the table above, the present invention exhibits excellent performance in detection scenarios for steel strips of different specifications. Compared with the prior art, it has significant improvements in detection efficiency, detection accuracy, defect identification capability, and detection coverage, and can better meet the actual needs of steel strip grouting fullness detection in tunnel engineering.

[0150] While the present invention has been disclosed above with reference to preferred embodiments, it is not intended to limit the invention. Any variations and modifications can be made by those skilled in the art without departing from the spirit and scope of the invention. Therefore, any modifications, equivalent changes, and alterations made to the above embodiments based on the technical essence of the present invention, without departing from the scope of the invention, fall within the protection scope defined by the claims of the present invention.

Claims

1. An intelligent robot for detecting the fullness of grouting in tunnel steel strips, characterized in that, The robot includes: Magnetic walking and spatial positioning module: used to drive the entire robot to walk automatically and continuously along the surface of the tunnel lining steel strip, and to obtain the robot's position information and walking speed information on the steel strip in real time; Dual-excitation point impact generation module: Arranged at fixed intervals along the walking direction of the steel belt, it is used to apply standard impact force to the same physical position on the steel belt one after another according to the received walking speed information, and at the same time generate an impact trigger signal and transmit it to the signal acquisition and preprocessing module to provide a synchronous reference for vibration signal acquisition; Multi-channel vibration sensor array module: arranged between the dual-excitation point impact generation modules, used to synchronously acquire the vibration response signal generated by the excitation of the steel strip; Signal acquisition and preprocessing module: used to receive the impact trigger signal and use the trigger moment as the time zero point to synchronously acquire and mark the vibration response signal, and then complete amplification, filtering and analog-to-digital conversion processing; Waveguide transfer function analysis module: Based on the received pre-processed vibration signal, it calculates the frequency domain transfer characteristics of vibration energy propagating in the composite waveguide section of steel strip and grout between two excitation points, and generates waveguide transfer function data; Grouting constraint stiffness profile reconstruction module: used to generate continuous grouting constraint stiffness spatial distribution data along the entire length of the steel strip based on the received waveguide transfer function data; The modal analysis and defect identification module is used to extract the running mode information of the steel strip based on the received preprocessed vibration signal, calculate the modal confidence score and identify the defect type, and generate defect type and modal confidence score data. Fullness Grading and Visualization Module: This module integrates the received spatial distribution data of grouting constraint stiffness, as well as defect type and modal confidence data, to perform a three-level fullness grading and generate a visual inspection report.

2. The intelligent detection robot for the fullness of tunnel steel strip grouting according to claim 1, characterized in that, The magnetic walking and spatial positioning module includes a Halbach array permanent magnet adsorption unit, a servo drive wheel set, a flexible suspension mechanism, a steel belt edge tracking sensor, an odometer, an IMU inertial measurement unit, and a steel belt feature recognition unit. The Halbach array permanent magnet adsorption unit is used to provide adsorption force so that the robot can be adsorbed onto the surface of the steel strip. The steel strip edge tracking sensor senses the position of the steel strip edge in real time and feeds it back to the walking control unit to adjust the robot's direction of travel, so that the robot always stays on the center line of the steel strip. The steel strip feature recognition unit identifies the inherent feature positions of steel strip lap joints, welding points, and fixing bolts based on the periodic structural noise features in the vibration signals collected by the multi-channel vibration sensor array module. These features are then used as absolute position reference points to periodically correct the cumulative error of the odometer.

3. The intelligent detection robot for the fullness of tunnel steel strip grouting according to claim 2, characterized in that, The dual-excitation point impact generation module includes a front exciter and a rear exciter. Each exciter includes an impact hammer driven by a voice coil motor, a force sensor, and a trigger control circuit. The trigger control circuit determines the real-time walking speed based on the magnetic walking and spatial positioning module. Calculate the trigger time difference between the front and rear actuators. The two actuators are triggered in sequence, so that the front actuator and the rear actuator apply standard impact force to the same physical position on the steel strip in turn. The force sensor is used to detect the impact force amplitude in real time and feed it back to the trigger control circuit for closed-loop control to maintain the consistency of excitation energy for each event. satisfy , This is the fixed distance between the front and rear actuators.

4. The intelligent detection robot for the fullness of tunnel steel strip grouting according to claim 3, characterized in that, The multi-channel vibration sensing array module includes at least three accelerometers, which are spaced equally along the steel belt direction between the front and rear exciters. The accelerometers are arranged such that each accelerometer is tightly coupled to the surface of the steel strip being measured via a magnetic mount. The frequency response range of the accelerometer covers 100Hz to 20kHz, and the sensitivity is not less than 100mV / g; All accelerometers synchronously acquire vibration response signals generated by the excitation of the steel strip at a sampling rate of no less than 51.2 kS / s.

5. The intelligent detection robot for the fullness of tunnel steel strip grouting according to claim 4, characterized in that, The specific working method of the waveguide transfer function analysis module is as follows: for each axial position of the steel strip... The time-domain window of the vibration response recorded by the sensor array after the exciter is triggered is captured. Time-domain window of vibration response recorded by sensor array after pre-exciter triggering ; Perform short-time Fourier transforms on the two time-domain windows respectively to obtain the corresponding complex spectra. ; Calculate position waveguide transfer function at The calculation formula is: ; in, This indicates the axial position of the steel strip, corresponding one-to-one with the robot's current detection position, with the starting end of the steel strip as the coordinate origin. Indicates time, Indicates the position of the post-exciter. After being triggered, the sensor array is at the location The time-domain signal window of the recorded vibration response. Indicates the position of the front actuator. After being triggered, the sensor array is at the location The time-domain signal window of the recorded vibration response. Indicates the frequency of the vibration signal. Indicates to The complex spectrum obtained after performing a short-time Fourier transform Indicates to The complex spectrum obtained after performing a short-time Fourier transform Indicates the position of the steel strip Location, frequency The waveguide transfer function describes the vibrational energy across a length of . The frequency domain propagation characteristics of the waveguide propagating in the composite waveguide section of steel strip and grouting body, including amplitude. Two components.

6. The intelligent detection robot for the fullness of tunnel steel strip grouting according to claim 5, characterized in that, The specific working method of the grouting constraint stiffness profile reconstruction module is as follows: Based on waveguide transfer function amplitude Calculate the position of the steel strip Grouting constraint stiffness index The calculation formula is: ； in, This indicates the lower limit of the sensitive frequency band for grout fullness testing. This indicates the upper limit of the sensitive frequency band for grout fullness testing. Indicates the frequency range [ , Definite integral operations within ] The frequency domain weighting function is obtained through prior calibration experiments on standard specimens of steel strips and grout bodies with known saturation states. The calculation formula is as follows: ； in, The amplitude of the waveguide transfer function of the standard specimen under fully compacted conditions. The waveguide transfer function amplitude of the standard specimen in a completely detached state is calculated by the grouting constraint stiffness profile reconstruction module based on the value obtained by the robot traveling along the steel strip. The sequence outputs a continuous curve of grouting constraint stiffness and position.

7. The intelligent detection robot for the fullness of tunnel steel strip grouting according to claim 6, characterized in that, The specific working method of the operational modal analysis and defect identification module is as follows: Operational modal analysis was performed on the N-channel vibration signals acquired by the multi-channel vibration sensor array module. A random subspace identification algorithm was used to extract the forward motion characteristics of the steel strip under continuous robot excitation. First-order running mode vector ; Calculate the modal confidence criterion values ​​between the measured mode shapes and the standard dense mode shapes of each order. The calculation formula is: ； Where N represents the total number of accelerometers in the multi-channel vibration sensing array, N is a positive integer and N≥3, and the dimension of the operating mode vector is the same as the total number of accelerometers, which is N. This indicates the order of the extracted mode shape. Indicates the order of the mode shape. , Indicates the first The measured mode shape vector is an N-dimensional complex vector. This indicates the first pre-marked section on a known fully compacted steel strip and grouting standard section. The reference mode vector is an N-dimensional complex vector. express The conjugate transpose of . express The conjugate transpose of ; when When <0.85, the mode is determined to be abnormal, and based on... Identify defect types by the variation pattern of values ​​along the position of the steel strip: like The value is generally below 0.85 and is evenly distributed along the steel strip, indicating insufficient overall grouting strength; like A sudden drop in value at a local location is identified as a local void. like The value shows a gradual drift trend along the direction of the steel strip, which is determined to be the density gradient of the grouting body.

8. The intelligent detection robot for the fullness of tunnel steel strip grouting according to claim 7, characterized in that, The three-level fullness grading logic of the fullness grading and visualization module is as follows: When the grouting constraint stiffness index at a certain location of the steel strip ≥ First threshold And all the previous Modal confidence When all values ​​are ≥0.90, the grouting fullness at that location is determined to be Grade I compact, and no warning is triggered. When the grouting constraint stiffness index at a certain location of the steel strip ≥ second threshold And < first threshold And / or there exists a modal confidence level of no more than 2nd order. When the value is <0.85, the grouting fullness at this location is judged to be Grade II, slightly non-dense, triggering a yellow warning and recommending grouting reinforcement. When the grouting constraint stiffness index at a certain location of the steel strip <Second threshold And / or there is a modal confidence level of 3 or higher. When the value is less than 0.85, the grouting fullness at this location is determined to be a Class III severe defect, triggering a red alert and recommending the removal of the steel strip and re-grouting.

9. The intelligent detection robot for the fullness of tunnel steel strip grouting according to claim 8, characterized in that, The visualization detection report generated by the fullness grading and visualization module includes: A chromatogram of the spatial distribution of grout fullness along the entire length of the steel strip, with green representing Class I dense areas, yellow representing Class II slightly loose areas, and red representing Class III severely defective areas; A detailed statistical table of each defective segment, including the start and end locations, length, average grouting constraint stiffness value, minimum modal confidence value, and recommended remedial measures; Basic testing information includes the steel strip number, testing time, testing personnel, system configuration parameters, and testing environmental conditions.

Images