A contactless identification method for voids behind shield tunnel segments

Abstract

The present invention provides a contactless identification method for voids behind shield tunnel segments, comprising the following steps: deploying a contactless identification system for voids behind shield tunnel segments and using the contactless identification system to collect data; wherein the contactless identification system includes a pulsed laser and a Doppler vibrometer; determining a target frequency domain interval and calculating a target time domain area; preprocessing the data; and analyzing the local characteristics of the signal in the target frequency interval to detect voids behind the segment walls. This method adopts the logic of the impact echo method, analyzes the stress wave reflection and refraction characteristics of different medium interfaces, and extracts features from the perspective of the signal time-frequency domain to achieve void identification behind shield tunnel segments. This method can detect voids behind the wall without contacting the structure surface, is highly practical, and is more adaptable to the tunnel environment. This method also has the advantages of being fast, non-destructive, and not easily affected by the tunnel environment.

Description

Technical Field

[0001] The present invention belongs to the field of cavity detection technology. Specifically, it relates to a contactless identification method for cavities behind shield tunnel segment walls. More specifically, it relates to a contactless identification method and electronic equipment for multiple types of cavities behind subway shield tunnel segment walls based on time-frequency analysis. Background Art

[0002] With the rapid development of urban rail transit in my country, many large and medium-sized cities have built trunk subway lines. However, due to the long-term operation of subways and the influence of geological conditions such as groundwater, many safety hazards have emerged during subway operation. Among them, cavities behind tunnel linings can induce various defects, such as deformation, cracking, and delamination of the lining, and even cause water leakage, damaging electrical wiring and equipment, and may also exacerbate the softening of the surrounding rock. Voids behind segmental walls can seriously affect the serviceability of subway systems.

[0003] Research on detecting defects within concrete structures is relatively mature. However, unlike internal cavities in concrete structures, tunnel surrounding rock is mostly composed of soil and rock. Compared to reinforced concrete segments, surrounding rock absorbs impact energy much more strongly, resulting in weaker reflected stress waves. This makes it more difficult to extract the characteristic values ​​of the time-frequency signal corresponding to the cavity. Furthermore, cavities are inherently inaccessible and invisible. Currently, a systematic, accurate, and efficient intelligent detection method for cavities behind shield tunnel segments has not yet been developed. Summary of the Invention

[0004] The purpose of the present invention is to provide a contactless identification method for cavities behind shield tunnel segment walls, aiming to solve the technical problem in the prior art that there is no systematic and accurate intelligent detection method for cavities behind shield tunnel segment walls.

[0005] To achieve the above object, the present invention adopts a technical solution of providing a contactless identification method for a cavity behind a shield tunnel segment wall, comprising the following steps:

[0006] Deploy a non-contact identification system for the voids behind shield tunnel segments and use it to collect data; wherein the non-contact identification system includes a pulsed laser and a Doppler vibrometer;

[0007] Determine the target frequency domain interval and calculate the target time domain area;

[0008] Preprocess the data;

[0009] Analyze the local characteristics of the signal in the target frequency range and detect the voids behind the segment wall.

[0010] Preferably, the deployment of a contactless identification system for the cavity behind the shield tunnel segment wall and the use of the contactless identification system for data collection include:

[0011] The inner surface of the segment is divided into measurement areas, with the intersection points as measurement points; the measurement areas are based on the distribution of cavities;

[0012] A pulsed laser is used to shoot a laser point onto a segment of a shield tunnel model and apply an impact; a Doppler vibrometer is used to receive vibration signals; wherein the laser point emitted by the Doppler vibrometer moves along with the movement of the laser point emitted by the pulsed laser.

[0013] Preferably, determining the target frequency domain interval and calculating the target time domain area includes:

[0014] Determine the target frequency domain interval, where the peak frequency is calculated as:

[0015]

[0016] Where: T i Indicates the depth of the target area from the surface of the segment. When i=1, 2, 3, and 4, it represents the distance of the cavity from the segment wall to 0m, 0.165m, 0.3m, and 0.6m respectively. α represents the cross-sectional coefficient. P v represents the wave velocity of P wave propagation; f represents the P wave frequency.

[0017] The target time domain interval can be calculated based on the wave velocity and the target cavity depth. The calculation formula for the time it takes for the P wave to propagate to the target depth is:

[0018]

[0019] Among them, t i It represents the time required for the P wave to propagate to the target area; T i Indicates the depth of the target area from the segment surface. When i=1, 2, 3, and 4, it represents the distance of the cavity from the segment wall to 0m, 0.165m, 0.3m, and 0.6m, respectively. v Indicates the speed of P-wave propagation.

[0020] Preferably, the target frequency domain interval is [2000 Hz, 7000 Hz]; the interval of time when the Doppler vibrometer receives the first echo after the impact is [1.56*10 -4 s, 4.56*10 -4 s].

[0021] Preferably, the pre-processing of data includes:

[0022] The peak recognition method was used to segment the repeated experimental data;

[0023] The data is denoised; the denoising method includes bandpass filtering, and the transfer function of the bandpass filter is:

[0024]

[0025] Among them, f L and f H are the low-frequency and high-frequency cutoff frequencies, respectively; f is the signal frequency; f0 represents the center frequency; Δf represents the bandwidth.

[0026] Preferably, the data is subjected to denoising, and the denoising method further includes: wavelet denoising based on wavelet transform; wherein the wavelet transform expression formula is:

[0027]

[0028] Among them, a refers to the scale factor, b refers to the translation factor, <x(t),ψ a,b (t)> represents the convolution operation, ψ(t) is the mother wavelet function of Daubechies 8 wavelet; W(a,b) represents the wavelet coefficient.

[0029] Preferably, analyzing the local characteristics of the signal in the target frequency range, and then sensing the cavity behind the segment wall, includes:

[0030] The time-frequency diagram is obtained by wavelet transform. The time domain expression of the wavelet function used in wavelet transform is:

[0031]

[0032] Where, f b is the bandwidth parameter, f c is the center frequency.

[0033] Preferably, the pulse laser is used to impact the surface of the segment of the shield tunnel model; the Doppler laser vibrometer is used to receive vibration signals, and the physical quantities collected by the Doppler laser vibrometer are one or more of velocity, acceleration, and displacement.

[0034] Preferably, when conducting the same segment experiment, the laser of the pulse laser is kept vertically incident on the shield tunnel.

[0035] An electronic device comprises: a bus, a transceiver, a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the transceiver, the memory, and the processor are connected via the bus. The electronic device is characterized in that when the computer program is executed by the processor, the steps of the contactless identification method for voids behind the wall of a shield tunnel segment as described in any one of the above items are implemented.

[0036] The beneficial effects of the contactless identification method for cavities behind shield tunnel segments provided by the present invention are as follows: Compared with existing technologies, the contactless identification method for cavities behind shield tunnel segments adopts the logic of the impact echo method, analyzes the stress wave reflection and refraction characteristics of different medium interfaces, and extracts features from the perspective of the signal time-frequency domain. This method can detect cavities behind shield tunnel segments without contacting the structural surface, thus being highly practical and more adaptable to the tunnel environment. This contactless identification method also has the advantages of being rapid, non-destructive, and not easily affected by the tunnel environment. BRIEF DESCRIPTION OF THE DRAWINGS

[0037] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0038] Figure 1 This is a schematic diagram of the shield tunnel model; Figure 2 Schematic diagram of measuring area division on the inner surface of the segment; 1-30 represents the number of columns; 1-12 represents the number of rows; Figure 3 This is the time domain diagram of 5Hz experimental data; Figure 4 This is the time domain diagram of the 5Hz repeated experiment with a peak; Figure 5 This is the time domain diagram of the segmented 5Hz repeated experiment; Figure 6 This is the time-frequency diagram before denoising of the denoising comparison example diagram; Figure 7Figure 8 is a denoised time-frequency diagram of the denoising comparison example; Figure 8 is the cavity placement area behind the wall of each specimen in the first and second groups; among them, Figure 8(a) represents the first group of experimental specimen block A; Figure 8(b) represents the first group of experimental specimen block B; Figure 8(c) represents the first group of experimental specimen block C; Figure 8(d) represents the first group of experimental specimen block D; Figure 8(e) represents the first group of experimental specimen block E; Figure 8(f) represents the second group of experimental specimen block A; Figure 8(g) represents the second group of experimental specimen block B; Figure 8(h) represents the second group of experimental specimen block C; Figure 8(i) represents the second group of experimental specimen block D; Figure 8(j) represents the second group of experimental specimen block E; Figure 8(k) Figure 9 shows the time-frequency diagram of the measurement points at the edge of the cavity, where Figure 9(a) shows 1-A0612, Figure 9(b) shows 1-A0613, Figure 9(c) shows 1-A0617, Figure 9(d) shows 1-A0712, and Figure 9(e) shows 1-A0717. Figure 10 shows the time-frequency diagram of the measurement points in the center of the cavity, where Figure 10(a) shows 1-A0614, Figure 10(b) shows 1-A0616, Figure 10(c) shows 1-A0713, Figure 10(d) shows 1-A0714, Figure 10(e) shows 1-A0715, and Figure 10(f) shows 1-A0716. Figure 11 Schematic diagram of the principle of the contactless identification system for the cavity behind the shield tunnel segment wall. DETAILED DESCRIPTION

[0039] The following will be combined with the drawings in the embodiments of this application to clearly and completely describe the technical solutions in the embodiments of this application. Obviously, the embodiments described are part of the embodiments of this application, not all of them. Based on the embodiments in this application, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of this application.

[0040] The terms "first," "second," and the like in the specification and claims of this application are used to distinguish similar objects, and are not used to describe a specific order or precedence. It should be understood that the terms used in this manner are interchangeable where appropriate, so that the embodiments of this application can be implemented in an order other than that illustrated or described herein, and that the objects distinguished by "first," "second," and the like are generally of the same type, and do not limit the number of objects; for example, the first object can be one or more. In addition, the term "and / or" in the specification and claims refers to at least one of the connected objects, and the character " / " generally indicates that the objects connected are in an "or" relationship.

[0041] In the following, in combination with the accompanying drawings, a contactless identification method for cavities behind the wall of a shield tunnel segment provided by an embodiment of the present application is described in detail through specific embodiments and application scenarios.

[0042] When stress waves encounter interfaces with different acoustic impedances in a structure, they will be reflected, transmitted, and refracted. Among them, stress waves include body waves (P waves, longitudinal waves and S waves, transverse waves) and surface waves (R waves, Rayleigh waves). P waves propagate faster, and the echoes are mainly P waves. Interfaces with different acoustic impedances will result in different echo signal characteristics, which can realize non-destructive detection of the thickness and defects of the structure. When elastic waves propagate inside the structure, they will bypass the steel structure and then re-converge after bypassing the steel bars, which will increase the propagation time. In general, the steel mesh inside the structure has little effect on the propagation of elastic waves.

[0043] Please also refer to Figures 1 to 11 The present invention provides a non-contact method for identifying a cavity behind a shield tunnel segment wall. The non-contact method for identifying a cavity behind a shield tunnel segment wall comprises the following steps:

[0044] Step S1: deploying a non-contact identification system for the cavity behind the shield tunnel segment wall and using the non-contact identification system to collect data; the non-contact identification system includes a pulse laser 1 and a Doppler vibrometer 2 arranged at a certain distance from the pulse laser 1;

[0045] In an optional embodiment of the present invention, step S1 may include the following sub-steps:

[0046] Substep S1.1: Divide the inner surface of the segment into measurement zones. Measure lines of 0.1m in length in both the horizontal and vertical directions, with the intersections serving as measurement points. The measurement zones are based on the distribution of voids. Skip measurement points at bolt holes or grouting holes.

[0047] Sub-step S1.2: using a pulsed laser 1 to emit a laser point onto a segment of the shield tunnel model and apply an impact; using a Doppler vibrometer 2 to receive a vibration signal;

[0048] The laser point of the Doppler vibrometer 2 on the pipe segment is set at a certain distance from the laser point of the pulse laser 1 on the pipe segment.

[0049] The laser point emitted by the Doppler vibrometer moves along with the movement of the laser point emitted by the pulse laser. When conducting the same segment experiment, the laser of the pulse laser 1 always penetrates the shield tunnel vertically.

[0050] Step S2: determine the target frequency domain interval and calculate the target time domain area;

[0051] In an optional embodiment of the present invention, step S2 may include the following sub-steps:

[0052] Sub-step S2.1: Determine the target frequency domain interval, where the interval value is [2000 Hz, 7000 Hz].

[0053] Through the specific depth range where defects may exist and affect the structural performance, the peak frequency corresponding to the defect is deduced. The calculation formula is:

[0054]

[0055] Where: T i Indicates the depth of the target area from the surface of the segment. When i=1, 2, 3, and 4, it represents the distance of the cavity from the segment wall to 0m, 0.165m, 0.3m, and 0.6m respectively. α represents the cross-sectional coefficient, which is generally 0.96. v It represents the velocity of P-wave propagation, and the maximum value of P-wave propagation velocity is 3955.54 m / s; f represents the frequency of P-wave.

[0056] Taking T1 = 0.3m, we calculate f1 = 6328.86Hz. At a position 0.165m behind the segment wall, T2 = 0.465m, we calculate f2 = 4083.13Hz. At a position 0.3m behind the segment wall, T3 = 0.6m, we calculate f3 = 3164.43Hz. At a position 0.6m behind the segment wall, T4 = 0.9m, we calculate f4 = 2109.62Hz. Therefore, the preset interval is [2000Hz, 7000Hz].

[0057] Sub-step S2.2: The target time domain interval can be calculated based on the wave velocity and the target cavity depth. The calculation formula for the time it takes for the P wave to propagate to the target depth is:

[0058]

[0059] Among them, t i It represents the time required for the P wave to propagate to the target area; T i Indicates the depth of the target area from the segment surface. When i=1, 2, 3, and 4, it represents the distance of the cavity from the segment wall to 0m, 0.165m, 0.3m, and 0.6m, respectively. v Indicates the speed of P-wave propagation. The maximum value of P-wave propagation speed is 3955.54m / s.

[0060] Take T1 = 0.3m, and calculate the time it takes for the P wave to propagate to the target area: t1 = 7.58*10 -5 / s.

[0061] When T2=0.465m, we can calculate t2=1.18*10 -4 / s.

[0062] When T3 = 0.6m, the calculated value is t3 = 1.52*10 -4 / s.

[0063] When T4=0.9m, the calculated value is t4=2.28*10 -4 / s.

[0064] The time it takes for Doppler vibrometer 2 to receive the first echo after the impact should be doubled based on this. The target time domain interval is [1.56*10 -4 s, 4.56*10 -4 s].

[0065] Step S4: Preprocess the data

[0066] In an optional embodiment of the present invention, step S3 may include the following sub-steps:

[0067] Sub-step S4.1: Segment the repeated experimental data using a peak recognition method.

[0068] In order to focus on the part of the data that is of interest, we only take 1*10 before and after the peak value in combination with the calculation of the target time domain interval. -3 s time range. It can achieve the effect of magnifying the target area and facilitate the observation of the basic characteristics of the echo signal.

[0069] Sub-step S4.2: denoising the data. The denoising method includes one or more of band-pass filtering and wavelet denoising based on wavelet transform.

[0070] The specific implementation of bandpass filtering denoising is as follows: Bandpass filtering allows signals in a specific frequency range to pass through while suppressing signals below or above this frequency range, which is used to filter the frequency band. The bandpass filter has two main parameters: center frequency f0 and bandwidth Δf. Its transfer function is:

[0071]

[0072] Among them, f L and f H They are low frequency and high frequency cutoff frequencies, respectively, set to 10 3 and 5 4 , f is the signal frequency, that is, only the frequency within f is retained L and f H The signal between.

[0073] Wavelet denoising based on wavelet transform is implemented by performing multi-scale wavelet decomposition on the original signal x(t) to obtain a series of wavelet coefficients. The present invention selects Daubechies 8 wavelet for decomposition, and the obtained high-frequency and low-frequency coefficients represent the details and approximate parts of the signal respectively. The wavelet basis function is convolved with the signal to obtain the wavelet coefficients W(a, b). The corresponding wavelet transform is expressed as:

[0074]

[0075] Among them, a refers to the scale factor, b refers to the translation factor, <x(t),ψ a,b (t)> represents the convolution operation, and ψ(t) is the mother wavelet function of Daubechies 8 wavelet.

[0076] Adaptive soft threshold λ is used to denoise the high frequency part, and the threshold is automatically adjusted according to the noise estimation. The denoised wavelet coefficient c′ is obtained according to the following denoising formula n :

[0077] c′ n =sign(c n )·max(|c n |-λ,0)

[0078] Threshold processing is performed on the high-frequency coefficients. For each wavelet coefficient, denoising is performed according to the soft threshold formula mentioned above. The denoised wavelet coefficients (including the high-frequency coefficients that have been thresholded and the unprocessed low-frequency coefficients) are used to perform inverse wavelet transform to reconstruct the denoised signal. The calculation expression of the reconstruction process is:

[0079]

[0080] Among them, C ψ is the normalization constant.

[0081] Step S5: Analyze the local characteristics of the signal in the target frequency range to detect the cavity behind the segment wall.

[0082] The time-frequency plot is generated using the wavelet transform. The wavelet transform is a signal processing method based on the wavelet function, which decomposes the signal into different frequency scales. The wavelet transform not only provides time-frequency domain conversion and frequency domain analysis, but also provides combined analysis of time-frequency domain information. By observing signals at different time and frequency scales, it excels at revealing local characteristics of the signal and is particularly friendly to non-stationary signals with sudden changes.

[0083] The present invention uses a complex Morlet wavelet function, which has a relatively balanced resolution in the time domain and the frequency domain, and generates a smooth time-frequency graph, which is convenient for observation and analysis.

[0084] The time domain expression of the complex Morlet wavelet is as follows:

[0085]

[0086] Where, f b is the bandwidth parameter, f c is the center frequency, f b and f cThe smaller it is, the better the frequency domain resolution is, and vice versa, the better the time domain resolution is. b =3 / 2π, f c =1 / 2π. j is the decomposition scale of wavelet transform.

[0087] The present invention provides a non-contact method for identifying voids behind shield tunnel segments. Compared to existing technologies, this method leverages the logic of the impact echo method to analyze the stress wave reflection and refraction characteristics at interfaces between different media. This method extracts features from the signal's time-frequency domain, enabling the identification of voids behind shield tunnel segments. This method can detect voids behind the segment walls without contacting the structure's surface, making it highly practical and adaptable to tunnel environments. This non-contact identification method is also rapid, non-destructive, and less susceptible to the effects of the tunnel environment.

[0088] Example 1. The method proposed in this example conducted an outdoor experiment on detecting cavities in the surrounding rock of a "segment-surrounding rock" model. Segments of different models and the same size were used, and cavities of different volumes, burial depths, and special positions were set. The vibration signals of different cavity conditions were analyzed in the frequency domain and time-frequency domain. Wavelet transform was used to qualitatively determine whether there were cavities behind the segment wall, and an identification rate evaluation index was set.

[0089] A non-contact identification method for a cavity behind a shield tunnel segment wall comprises the following steps:

[0090] Step SS1: construct a shield tunnel model and preset a cavity structure on the shield tunnel model;

[0091] In this step, a shield tunnel model is constructed using a "segment-surrounding rock" combination. Specifically, six segments are placed in a longitudinal arrangement. From left to right, they are named A, B, C, D, E, and F. Adjacent segments are connected end to end, and a wall is built between each segment to enhance segment stability. The surrounding rock behind the segment wall is repeatedly compacted by a roller, simulating the actual structure of a shield tunnel.

[0092] The surrounding rock behind the segment was 1.5m wide, allowing for different cavity depths. Commonly available mineral water bottles of varying sizes were used for simulation. The sizes included 1.5L, 4.5L, 5L, 12L, and 16L. To ensure experimental integrity, cavities were also constructed using combinations of 5L+5L, 12L+12L, and 1.5L*2+4.5L*2 mineral water bottles. The material parameters for the segment are shown in Table 1.

[0093] Table 1. Example segment material parameters

[0094]

[0095] Step SS2: deploying a non-contact identification system for the cavity behind the shield tunnel segment wall and using the non-contact identification system for data collection; the non-contact identification system includes a pulse laser 1 and a Doppler vibrometer 2 arranged at a certain distance from the pulse laser 1;

[0096] In an optional embodiment of the present invention, step S2 may include the following sub-steps:

[0097] Sub-step SS2.1: Divide the inner surface of the segment into a grid of 0.1m squares. Use grid intersections as measurement points. The measurement area is based on the distribution of voids. Skip measurement points at bolt holes or grouting holes.

[0098] Substep SS2.2: Use pulsed laser 1 to project laser light onto the segment and impact it; use Doppler vibrometer 2 to receive vibration signals. Specifically, pulsed laser 1 is used to impact the surface of the segment of the shield tunnel model. The Doppler laser vibrometer receives vibration signals. The Doppler laser vibrometer has a frequency bandwidth range of 0-25 MHz and can measure three physical quantities: velocity, acceleration, and displacement, facilitating subsequent processing. Acceleration is selected as the physical quantity to be collected. The entire experiment utilizes non-contact impact and sensing. The laser point of Doppler vibrometer 2 on the segment is located a certain distance from the laser point of pulsed laser 1 on the segment. The distance between the laser point of Doppler vibrometer 2 on the segment and the laser point of pulsed laser 1 on the segment is 0.01-0.05 m, preferably 0.02 m. The laser point emitted by Doppler vibrometer 2 moves in conjunction with the movement of the laser point emitted by pulsed laser 1.

[0099] During the same segment experiment, the laser from pulsed laser 1 was always emitted perpendicularly into the shield tunnel. Specifically, the frequency of pulsed laser 1 was set at 5 Hz, and multiple experiments were performed. The laser emission point was at the center of the inner ring of the segment. From any two points on the segment surface, a line segment perpendicular to the inner ring at those points was drawn. The two line segments were extended to obtain the intersection point, which was the center of the inner ring of the segment. The pulsed laser rotated around the emission point to ensure that the laser always entered the shield tunnel perpendicularly.

[0100] This experiment was divided into two groups, with six segments in each group. For economic reasons, the cavities and surrounding rock soil were excavated in the secondary experiment, and the cavities of the secondary experiment were reburied according to this method. The experimental settings of the two groups are shown in Table 2 below.

[0101] Table 2 Arrangement of cavity pre-embedding in two experiments

[0102]

[0103] Two groups of experimental data were used. This study used 120 data from the first group of experiments and 100 data from the second group of experiments. There were six segments in each group, and the cavities behind each segment wall were set with different conditions. A total of 220 data were measured for single impact and multiple impact, 110 data for 1Hz single experiment, and 110 data for 5Hz multiple experiments.

[0104] Step SS3: Position the time-frequency interval, specifically, determine the target frequency domain interval and calculate the target time domain area;

[0105] This step includes:

[0106] Sub-step SS3.1: Determine the target frequency domain interval, where the interval value is [2000 Hz, 7000 Hz].

[0107] The specific determination process is: through the specific depth range where defects may exist and affect structural performance, the peak frequency corresponding to the defect is deduced. The calculation formula is:

[0108]

[0109] Where: T i Indicates the depth of the target area from the surface of the segment. When i=1, 2, 3, and 4, it represents the distance of the cavity from the segment wall to 0m, 0.165m, 0.3m, and 0.6m respectively. α represents the cross-sectional coefficient, which is generally 0.96. v It represents the velocity of P-wave propagation, and the maximum value of P-wave propagation velocity is 3955.54 m / s; f represents the frequency of P-wave.

[0110] Taking T1 = 0.3m, we calculate f1 = 6328.86Hz. At a position 0.165m behind the segment wall, T2 = 0.465m, we calculate f2 = 4083.13Hz. At a position 0.3m behind the segment wall, T3 = 0.6m, we calculate f3 = 3164.43Hz. At a position 0.6m behind the segment wall, T4 = 0.9m, we calculate f4 = 2109.62Hz. Therefore, the preset interval is [2000Hz, 7000Hz].

[0111] Sub-step SS3.2: Calculate the target time domain interval based on the wave velocity and the target cavity depth. The calculation formula for the time it takes for the P wave to propagate to the target depth is:

[0112]

[0113] Among them, t i It represents the time required for the P wave to propagate to the target area; T iIndicates the depth of the target area from the segment surface. When i=1, 2, 3, and 4, it represents the distance of the cavity from the segment wall to 0m, 0.165m, 0.3m, and 0.6m, respectively. v Indicates the speed of P-wave propagation. The maximum value of P-wave propagation speed is 3955.54m / s.

[0114] Take T1 = 0.3m, and calculate the time it takes for the P wave to propagate to the target area: t1 = 7.58*10 -5 / s.

[0115] When T2=0.465m, we can calculate t2=1.18*10 -4 / s.

[0116] When T3 = 0.6m, the calculated value is t3 = 1.52*10 -4 / s.

[0117] When T4=0.9m, the calculated value is t4=2.28*10 -4 / s.

[0118] The time it takes for Doppler vibrometer 2 to receive the first echo after the impact should be doubled based on this. The target time domain interval is approximately [1.56*10 -4 s, 4.56*10 -4 s].

[0119] Step SS4, attach Figure 3 The collected data were pre-processed. 1 Hz single test data and 5 Hz repeated test data were collected. Comprehensive evaluation of 5 Hz repeated test data helps to reduce the influence of errors. Therefore, the main research object of this embodiment is 5 Hz repeated test data.

[0120] In an optional embodiment of the present invention, step S4 may include the following sub-steps:

[0121] Sub-step S4.1: Use the peak recognition method to segment the repeated experimental data. Extract the points with obvious larger values ​​as the peaks of each interval segment, and divide the 10 points before and after each peak into segments. -3 The data within the time range of s is regarded as an interval segment, and the segmented graph refers to Figure 4 shown.

[0122] In order to focus on the part of the data that is of interest, we only take 1*10 before and after the peak value in combination with the calculation of the target time domain interval. -3 s time range, refer to Figure 5 It can achieve the effect of magnifying the target area, making it easier to observe the basic characteristics of the echo signal.

[0123] Sub-step S4.2: De-noising the data. De-noising methods include band-pass filtering and wavelet denoising based on wavelet transform.

[0124] The specific implementation of bandpass filtering denoising is as follows: Bandpass filtering allows signals in a specific frequency range to pass through while suppressing signals below or above this frequency range, which is used to filter the frequency band. The bandpass filter has two main parameters: center frequency f0 and bandwidth Δf. Its transfer function is:

[0125]

[0126] Among them, f L and f H They are low frequency and high frequency cutoff frequencies, respectively, set to 10 3 and 5 4 , f is the signal frequency, that is, only the frequency within f is retained L and f H The signal between.

[0127] Wavelet denoising based on wavelet transform is implemented by performing multi-scale wavelet decomposition on the original signal x(t) to obtain a series of wavelet coefficients. The present invention selects Daubechies 8 wavelet for decomposition, and the obtained high-frequency and low-frequency coefficients represent the details and approximate parts of the signal respectively. The wavelet basis function is convolved with the signal to obtain the wavelet coefficients W(a, b). The corresponding wavelet transform is expressed as:

[0128]

[0129] Among them, a refers to the scale factor, b refers to the translation factor, <x(t),ψ a,b (t)> represents the convolution operation, and ψ(t) is the mother wavelet function of Daubechies 8 wavelet.

[0130] Adaptive soft threshold λ is used to denoise the high frequency part, and the threshold is automatically adjusted according to the noise estimation. The denoised wavelet coefficient c′ is obtained according to the following denoising formula n :

[0131] c′ n =sign(c n )·max(|c n |-λ,0)

[0132] Threshold processing is performed on the high-frequency coefficients. For each wavelet coefficient, denoising is performed according to the soft threshold formula mentioned above. The denoised wavelet coefficients (including the high-frequency coefficients that have been thresholded and the unprocessed low-frequency coefficients) are used to perform inverse wavelet transform to reconstruct the denoised signal. The calculation expression of the reconstruction process is:

[0133]

[0134] Among them, C ψ is the normalization constant.

[0135] Reference Figure 6 、 Figure 7 Observation and comparison show that the denoised time-frequency graph in the red frame has well suppressed high-frequency data between 0.8MHz and 1MHz, and the low-frequency data appears smoother. In addition, some discrete noise signals are removed without excessively weakening the basic characteristics of the time-frequency signal.

[0136] Step SS5: Extract the frequency components corresponding to the specific scale, analyze the local characteristics of the signal in the target frequency range, and then perceive the voids behind the segment wall.

[0137] The time-frequency plot is generated using the wavelet transform. The wavelet transform is a signal processing method based on the wavelet function, which decomposes the signal into different frequency scales. The wavelet transform not only provides time-frequency domain conversion and frequency domain analysis, but also provides combined analysis of time-frequency domain information. By observing signals at different time and frequency scales, it excels at revealing local characteristics of the signal and is particularly friendly to non-stationary signals with sudden changes.

[0138] The present invention uses a complex Morlet wavelet function, which has a relatively balanced resolution in the time domain and the frequency domain, and generates a smooth time-frequency graph, which is convenient for observation and analysis.

[0139] The time domain expression of the complex Morlet wavelet is as follows:

[0140]

[0141] Where, f b is the bandwidth parameter, f c is the center frequency, f b and f c The smaller it is, the better the frequency domain resolution is, and vice versa, the better the time domain resolution is. b =3 / 2π, f c =1 / 2π.

[0142] The experiment was divided into two groups, with different cavity locations, burial depths, sizes, and contents. A segment without a cavity was also set up for comparison. The cavity placement area behind the wall of each specimen is shown in Figures 8(a) to (e).

[0143] All experimental data were collated using a comparative analysis method with controlled variables. Data were categorized based on measurement point location, the presence of cavities, their location, depth, and volume. Each measurement point had different segmented data, and the recognition success rate was quantified using the proportion of data that met the recognition signal characteristics. To distinguish the first and second sets of experimental data, the numbers 1 and 2 were added before the measurement point numbers.

[0144] The first group of A# test blocks was used as the control group. The control group consisted of small (5L) wall-attached air voids behind the grouting hole in the center of the segment. The recognition rate of the data was first analyzed using the measurement point location as the classification standard, as shown in Table 3.

[0145] Table 3. Classification and recognition rate of control group data

[0146]

[0147] After comparison, it was found that in the first group of A# test block experiments, due to the irregular shape of the water bottle and the fact that the incident laser line was not guaranteed to be perpendicular to the tangent of the pipe surface, the time-frequency diagrams of most measuring points at the edge of the cavity had almost no echo response in the target time-frequency range.

[0148] The survey area is concentrated in the A0612-A0717 area, including eleven measuring points. The data analysis results of the cavity edge measuring points 1-A0612, 1-A0613, 1-A0617, 1-A0712 and 1-A0717 are shown in Figure 9.

[0149] In contrast, the time-frequency diagrams of the measurement points in the center of the cavity exhibit double or multi-peak patterns within the target time-frequency interval, reflecting the echo and oscillation performance of the P wave when it encounters the air interface. The data analysis results of the cavity center measurement points 1-A0614, 1-A0616, 1-A0713, 1-A0714, 1-A0715, and 1-A0716 are shown in Figure 10.

[0150] The first set of experimental data for Block A# served as a control group. The control group consisted of a small (5L) air cavity attached to the wall after the central grouting hole in the segment. We analyzed the recognition rates for other experimental data sets with or without cavities, as well as with varying cavities' locations, depths, volumes, and contents. The recognition rates for the different experimental data sets are shown in Table 4 below (recognition rates are rounded to integers).

[0151] Table 4 Void recognition rate of each experimental group under different conditions

[0152]

[0153]

[0154] In summary, the system has a high recognition rate for cavities behind the segment wall. This includes both water-rich and air-filled cavities. It can identify cavities within 30 centimeters behind the segment wall, and has a high accuracy rate for cavities larger than 1.5 liters, effectively meeting the requirements of nondestructive testing for deep-seated defects in tunnel structures.

[0155] An embodiment of the present invention also provides an electronic device, comprising: a processor, a memory, and a computer program stored in the memory and runnable on the processor. When the computer program is executed by the processor, the various processes of the above-mentioned embodiment of the contactless identification method for cavities behind the wall of a shield tunnel segment are implemented, and the same technical effect can be achieved. To avoid repetition, it will not be repeated here.

[0156] Although preferred embodiments of the present invention have been described, those skilled in the art may make additional changes and modifications to these embodiments once they are aware of the basic creative concepts. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments and all changes and modifications that fall within the scope of the embodiments of the present invention. The above is a detailed introduction to a contactless identification method and electronic device for the void behind the wall of a shield tunnel segment provided by the present invention. Specific examples are used herein to illustrate the principles and implementation methods of the present invention. The description of the above embodiments is only used to help understand the method of the present invention and its core idea. At the same time, for those skilled in the art, according to the idea of ​​the present invention, there will be changes in the specific implementation methods and application scope. In summary, the content of this specification should not be understood as limiting the present invention.

Claims

1. A contactless identification method for cavities behind shield tunnel segments, characterized in that: The following steps are involved: Deploy a non-contact identification system for the voids behind shield tunnel segments and use it to collect data; wherein the non-contact identification system includes a pulsed laser and a Doppler vibrometer; Determine the target frequency domain interval and calculate the target time domain area; Preprocess the data; Analyze the local characteristics of the signal in the target frequency domain interval to detect the voids behind the shield tunnel segment wall; The method of deploying a contactless identification system for the cavity behind the shield tunnel segment wall and using the contactless identification system for data collection includes: The inner surface of the shield tunnel segment is divided into measurement areas. The measurement lines are divided into 0.1m lengths in both the horizontal and vertical directions, and the intersections are used as measurement points. The measurement areas are based on the distribution of voids. A pulsed laser is used to shoot a laser spot onto a shield tunnel segment and apply an impact. A Doppler laser vibrometer is used to receive the vibration signal. The laser spot emitted by the Doppler vibrometer moves as the laser spot emitted by the pulsed laser moves. The determining of the target frequency domain interval and the calculating of the target time domain area include: Determine the target frequency domain interval, where the peak frequency is calculated as: Where: T i represents the target cavity depth from the shield tunnel segment surface. When i=1, 2, 3, and 4, it represents the case where the cavity is 0m, 0.165m, 0.3m, and 0.6m away from the shield tunnel segment wall, respectively. α represents the section coefficient. P v represents the wave velocity of P wave propagation; f represents the P wave frequency; The target time domain interval can be calculated based on the wave velocity and the target cavity depth. The time required for the P wave to propagate to the target cavity depth can be calculated as follows: Among them, t i Indicates the time required for the P wave to propagate to the target cavity depth; The analysis of the local characteristics of the signal in the target frequency domain interval, and then sensing the void behind the shield tunnel segment wall, includes: The time-frequency diagram is obtained by wavelet transform. The time domain expression of the wavelet function used in wavelet transform is: Where, f b is the bandwidth parameter, f c is the center frequency; j is the decomposition scale of wavelet transform.

2. The non-contact identification method for the cavity behind the shield tunnel segment wall according to claim 1, characterized in that: The data preprocessing includes: The data were segmented using the peak recognition method; The data is denoised; the denoising method includes bandpass filtering, and the transfer function of the bandpass filtering is: Among them, f L and f H are the low-frequency and high-frequency cutoff frequencies, respectively; f is the signal frequency; f0 represents the center frequency; Δf represents the bandwidth.

3. The non-contact identification method for the cavity behind the shield tunnel segment wall according to claim 2, characterized in that: The pulse laser is used to impact the surface of the shield tunnel segment; the Doppler laser vibrometer is used to receive vibration signals, and the physical quantity collected by the Doppler laser vibrometer is one or more of velocity, acceleration, and displacement.

4. An electronic device comprising: A bus, a transceiver, a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the transceiver, the memory, and the processor are connected via the bus, and wherein the computer program, when executed by the processor, implements the steps of a contactless identification method for a cavity behind a shield tunnel segment wall as described in any one of claims 1 to 3.

Images