Shield tunnel sound wave detection migration imaging method based on low frequency prior constraint

Abstract

The application provides a shield tunnel sound wave detection migration imaging method based on low-frequency prior constraint, and belongs to the field of tunnel engineering sound wave detection imaging. The method comprises the following steps: collecting frequency sound wave data and performing pretreatment to obtain low-frequency effective reflection wave data and high-frequency effective reflection wave data; obtaining a low-frequency reverse time migration imaging result based on the low-frequency effective reflection wave data and performing feature transformation to obtain a low-frequency prior probability graph; taking the low-frequency prior probability graph as an imaging condition, obtaining a high-frequency imaging result under the low-frequency prior constraint based on the high-frequency effective reflection wave data, performing feature transformation to obtain a high-frequency feedback probability graph, and weighting and fusing the high-frequency feedback probability graph and the low-frequency prior probability graph to obtain an updated prior probability graph; iteratively until a convergence criterion is met, and outputting a multi-frequency sound wave joint reverse time migration imaging result. The application realizes the organic unification of low-frequency stable positioning and high-frequency fine resolution, and effectively suppresses the pseudo abnormal response under a strong scattering background.

Description

Technical Field

[0001] This invention belongs to the field of acoustic wave detection and imaging technology in tunnel engineering, and particularly relates to a method for acoustic wave detection migration imaging of shield tunnels based on low-frequency prior constraints. Background Technology

[0002] The statements in this section are merely background information related to the present invention and do not necessarily constitute prior art.

[0003] During shield tunneling, large boulders, isolated rocks, and fractured zones are often randomly distributed ahead of the tunnel face. These geological bodies vary greatly in scale and have complex boundaries, easily causing engineering accidents such as cutterhead wear, shield jamming, and tunneling stoppage. Therefore, high-precision advanced geological exploration ahead of the tunnel face is crucial for construction safety management. Acoustic wave detection, which relies on differences in strata impedance to identify underground anomalies, is the mainstream technology for advanced shield tunneling prediction, but existing imaging methods have significant shortcomings.

[0004] Single-frequency acoustic imaging struggles to balance detection stability and imaging resolution: low-frequency acoustic waves have strong penetrating power and can stably locate large-scale anomalies, but their long wavelengths often result in wide energy bands that cannot depict fine boundaries; high-frequency acoustic waves have short wavelengths and high resolution, which can better characterize the details of anomaly boundaries, but they are easily affected by scattered waves from small pebbles and random noise in strong scattering media such as pebble and sandy gravel strata, leading to more false anomaly responses in the imaging results.

[0005] Furthermore, current multi-frequency imaging methods only perform simple superposition or parallel comparison of imaging results at different frequencies. There is a lack of bidirectional constraints and iterative feedback mechanisms between low-frequency and high-frequency information. The stability advantage of low frequencies cannot be transferred to high-frequency imaging, and the fine details of high frequencies cannot be used to correct the defects of low-frequency imaging. It is impossible to simultaneously achieve reliable positioning of anomalies and accurate boundary characterization. Summary of the Invention

[0006] To overcome the shortcomings of the prior art, this invention provides a shield tunnel acoustic wave detection migration imaging method based on low-frequency prior constraints, aiming to solve the problem that traditional single-frequency acoustic wave imaging is difficult to simultaneously take into account detection stability and boundary resolution.

[0007] To achieve the above objectives, one or more embodiments of the present invention provide the following technical solutions: The first aspect of this invention provides a method for acoustic wave detection migration imaging of shield tunnels based on low-frequency prior constraints; A migration imaging method for acoustic detection in shield tunnels based on low-frequency prior constraints includes: Multi-frequency acoustic wave data were collected and preprocessed during the tunnel boring process to obtain low-frequency effective reflected wave data and high-frequency effective reflected wave data. Low-frequency reverse time migration imaging is performed based on low-frequency effective reflected wave data to obtain low-frequency reverse time migration imaging results. Feature transformation is then performed on the low-frequency reverse time migration imaging results to obtain a low-frequency prior probability map. Using the low-frequency prior probability map as the imaging condition, the high-frequency imaging result under the low-frequency prior constraint is obtained based on the high-frequency effective reflected wave data. The high-frequency imaging result is then subjected to feature transformation to obtain the high-frequency feedback probability map. The high-frequency feedback probability map and the low-frequency prior probability map are then weighted and fused to obtain the updated prior probability map. Repeat the high-frequency imaging, high-frequency feedback probability map construction, and prior probability map update process under low-frequency prior constraints until the convergence criterion is met, and output the multi-frequency acoustic wave joint reverse time migration imaging results.

[0008] As a further technical solution, low-frequency reverse-time migration imaging is performed based on low-frequency effective reflected wave data to obtain low-frequency reverse-time migration imaging results, including: Based on low-frequency effective reflected wave data, the source wavefield is extended forward and the receiver wavefield is extended backward using the acoustic wave equation. The low-frequency reverse time migration imaging results are obtained through cross-correlation imaging conditions.

[0009] In the formula, This is a low-frequency reverse time migration imaging result. For low-frequency source wavefield, For low-frequency detector wave field, For spatial coordinates, For time.

[0010] As a further technical solution, feature transformation is performed on the low-frequency reverse-time migration imaging results to obtain a low-frequency prior probability map, including: Envelope extraction is performed on the low-frequency reverse time migration imaging results, and the positive and negative oscillatory responses are converted into energy-type distributions. The extracted low-frequency imaging envelope is normalized, and the amplitude is mapped to a standard numerical range to obtain the normalized low-frequency energy map. A nonlinear enhancement function is introduced to nonlinearly enhance the normalized low-frequency energy map, and the enhanced result is smoothed and diffused to obtain a continuously distributed low-frequency prior probability map.

[0011] As a further technical solution, the low-frequency prior probability map is used as the imaging condition. Based on the high-frequency effective reflected wave data, the high-frequency imaging results under the low-frequency prior constraint are obtained, as shown below:

[0012] In the formula, This represents the high-frequency imaging result after low-frequency prior constraints in the k-th round. This is the low-frequency prior probability graph for the k-th round; Forward propagation wave field of high-frequency source, This is a high-frequency detector backpropagation wave field.

[0013] As a further technical solution, feature transformation is performed on the high-frequency imaging results to obtain a high-frequency feedback probability map. The high-frequency feedback probability map is then weighted and fused with a low-frequency prior probability map to obtain an updated prior probability map, including: Envelope extraction is performed on the high-frequency imaging results after prior constraints, and the oscillatory response is transformed into an energy-type distribution. The extracted high-frequency imaging envelope is locally normalized. Different local regions are normalized within a fixed window to preserve the relative response intensity of weak signal regions. Nonlinear enhancement is performed on the locally normalized high-frequency energy map to construct a high-frequency feedback probability map; The high-frequency feedback probability map is weighted and fused with the current low-frequency prior probability map, and the fusion weight is dynamically adjusted with each iteration to obtain the updated prior probability map.

[0014] As a further technical solution, the high-frequency imaging, high-frequency feedback probability map construction, and prior probability map update process under low-frequency prior constraints are repeatedly executed until the convergence criterion is met, and the multi-frequency acoustic wave joint reverse time migration imaging results are output, including: After each round of prior probability map update is completed, the next round of high-frequency imaging, high-frequency feedback probability map construction and prior probability map update process under low-frequency prior constraints is immediately executed. After each iteration, the rate of change between two adjacent high-frequency imaging results and the rate of change between two adjacent prior probability maps are calculated. When the rate of change of the imaging result is less than the preset first tolerance threshold and the rate of change of the prior probability map is less than the preset second tolerance threshold, the iteration is determined to be converged, the update is stopped, and the high-frequency imaging result obtained in the current round is output as the final multi-frequency acoustic wave joint reverse time migration imaging result.

[0015] The second aspect of the present invention provides a shield tunnel acoustic wave detection migration imaging system based on low-frequency prior constraints.

[0016] A shield tunnel acoustic wave detection migration imaging system based on low-frequency prior constraints includes: The data acquisition and preprocessing module is configured to: acquire multi-frequency acoustic wave data during the shield tunneling process and preprocess it to obtain low-frequency effective reflected wave data and high-frequency effective reflected wave data; The low-frequency imaging and prior construction module is configured to: perform low-frequency reverse time migration imaging based on low-frequency effective reflected wave data to obtain low-frequency reverse time migration imaging results, and perform feature transformation on the low-frequency reverse time migration imaging results to obtain a low-frequency prior probability map. The high-frequency constrained imaging and feedback fusion module is configured to: use the low-frequency prior probability map as the imaging condition, obtain the high-frequency imaging result under the low-frequency prior constraint based on the high-frequency effective reflected wave data, perform feature transformation on the high-frequency imaging result to obtain the high-frequency feedback probability map, and weightedly fuse the high-frequency feedback probability map with the low-frequency prior probability map to obtain the updated prior probability map. The iterative control and imaging output module is configured to repeatedly execute the high-frequency imaging, high-frequency feedback probability map construction and prior probability map update process under low-frequency prior constraints until the convergence criterion is met, and output the multi-frequency acoustic wave joint reverse time migration imaging results.

[0017] A third aspect of the present invention provides a computer-readable storage medium having a program stored thereon, which, when executed by a processor, implements the steps in the shield tunnel acoustic wave detection migration imaging method based on low-frequency prior constraints as described in the first aspect of the present invention.

[0018] A fourth aspect of the present invention provides an electronic device, including a memory, a processor, and a program stored in the memory and executable on the processor, wherein the processor executes the program to implement the steps in the shield tunnel acoustic wave detection migration imaging method based on low-frequency prior constraints as described in the first aspect of the present invention.

[0019] The fifth aspect of the present invention provides a computer program product, including a computer program / instruction that, when executed by a processor, implements the steps in the shield tunnel acoustic wave detection migration imaging method based on low-frequency prior constraints described in the first aspect of the present invention.

[0020] The above one or more technical solutions have the following beneficial effects: (1) This invention utilizes the advantage of low-frequency acoustic wave detection in locating large-scale anomalies. It constructs a prior probability map of the target anomaly through low-frequency reverse time migration imaging results, and transforms the stable position information in low-frequency imaging into a spatial constraint weight field that can be used for high-frequency imaging. This allows the stability advantage of low-frequency detection to be effectively transferred to the high-frequency imaging process, laying the foundation for subsequent multi-frequency information fusion.

[0021] (2) By introducing the prior probability map as a spatial weighting factor into the high-frequency reverse time migration imaging condition, the high-frequency imaging energy is preferentially focused on the high-confidence anomalous region indicated by the low-frequency imaging. This can effectively suppress the arc-shaped stripe artifacts and stray energy generated by the scattering waves of small pebbles and random noise in high-frequency imaging under strong scattering background, thereby improving the focus and signal-to-noise ratio of the target anomalous body.

[0022] (3) This invention uses high-frequency imaging results to construct a feedback probability map in reverse, and uses a weighted fusion method to iteratively correct the low-frequency prior probability map, so that the local structure and boundary details contained in the high-frequency imaging results can be gradually transmitted to the prior probability map, realizing the feedback correction of the low-frequency prior by the high-frequency boundary details, and finally realizing the organic unity of low-frequency positioning stability and high-frequency resolution advantage.

[0023] (4) The present invention sets up a dual convergence criterion of the rate of change of high frequency imaging results and the rate of change of prior probability map, which can accurately determine whether the iterative process has reached a stable state, effectively avoid the problem of insufficient fusion caused by early stopping and the waste of computing resources caused by excessive iteration, and ensure the stability of imaging results and the rationality of computing efficiency.

[0024] Advantages of additional aspects of the invention will be set forth in part in the description which follows, and in part will be obvious from the description, or may be learned by practice of the invention. Attached Figure Description

[0025] The accompanying drawings, which form part of this invention, are used to provide a further understanding of the invention. The illustrative embodiments of the invention and their descriptions are used to explain the invention and do not constitute an improper limitation of the invention.

[0026] Figure 1 This is a flowchart of the method in the first embodiment.

[0027] Figure 2 This is a schematic diagram of the frequency division acoustic wave detection and observation method in the first embodiment.

[0028] Figure 3 A schematic diagram of the process for constructing a low-frequency prior probability map based on low-frequency reverse time migration imaging results for the first embodiment.

[0029] Figure 4 This is a schematic diagram comparing the results of the method of the present invention with those of single-frequency RTM imaging under the dual boulder model in the first embodiment.

[0030] Figure 5 This is a schematic diagram comparing the results of the method of the present invention with those of single-frequency RTM imaging under the three-boulder model in the first embodiment.

[0031] Figure 6 This is a system structure diagram of the second embodiment. Detailed Implementation

[0032] It should be noted that the following detailed descriptions are exemplary and intended to provide further illustration of the invention. Unless otherwise specified, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains.

[0033] It should be noted that the terminology used herein is for the purpose of describing particular implementations only and is not intended to limit the exemplary implementations of the present invention.

[0034] Where there is no conflict, the embodiments and features in the embodiments of the present invention can be combined with each other.

[0035] Example 1 This embodiment discloses a shield tunnel acoustic wave detection migration imaging method based on low-frequency prior constraints. A prior probability map is constructed using low-frequency imaging results to spatially constrain high-frequency reverse-time migration imaging. At the same time, the prior probability map is updated using high-frequency imaging results as feedback, so that low-frequency stable positioning information and high-frequency fine boundary information are gradually fused during the iteration process, thereby improving the target focusing, boundary continuity and identification reliability of imaging of adverse geological bodies under complex scattering backgrounds.

[0036] like Figure 1 As shown, the shield tunnel acoustic wave detection migration imaging method based on low-frequency prior constraints includes: Step S1: Collect multi-frequency acoustic wave data during the shield tunneling process and preprocess it to obtain low-frequency effective reflected wave data and high-frequency effective reflected wave data.

[0037] Step S11, combined Figure 2 A sound wave excitation device and a sound wave receiving device are installed in the cutterhead area of ​​the shield tunnel to excite low-frequency sound wave signals and high-frequency sound wave signals respectively facing the detection area in front of the shield tunneling, and collect corresponding multi-channel sound wave records.

[0038] Low-frequency acoustic signals are used to acquire large-scale spatial responses and stable position indications of anomalies such as large boulders, isolated rocks, and boulder clusters, while high-frequency acoustic signals are used to enhance local interface reflections and boundary detail responses of the anomalies. Depending on the detection distance, target scale, and stratum scattering intensity, mid-frequency acoustic signals can also be set as a transition frequency band between low-frequency and high-frequency acoustic waves.

[0039] Step S12: Preprocess the acquired low-frequency sound wave data and high-frequency sound wave data respectively. The preprocessing process includes DC removal, trend removal, bandpass filtering, channel equalization, random noise suppression, and abnormal channel removal.

[0040] After preprocessing, low-frequency effective reflected wave data are obtained. and high-frequency effective reflected wave data This step can improve the signal-to-noise ratio of target reflected waves in acoustic recordings of different frequencies, and reduce the impact of construction environment noise, pebble scattering waves, and instrument noise on subsequent reverse time migration imaging.

[0041] Step S2: Perform low-frequency reverse time migration imaging based on low-frequency effective reflected wave data to obtain low-frequency reverse time migration imaging results. Perform feature transformation on the low-frequency reverse time migration imaging results to obtain a low-frequency prior probability map.

[0042] Step S21, based on low-frequency effective reflected wave data The source wavefield is extended in the forward direction and the receiver wavefield is extended in the reverse direction using the acoustic wave equation. Low-frequency reverse time migration imaging results are obtained through cross-correlation imaging conditions.

[0043] In the formula, This is a low-frequency reverse time migration imaging result. For low-frequency source wavefield, For low-frequency detector wave field, For spatial coordinates, For time.

[0044] Low-frequency reverse time migration imaging results are used to characterize the large-scale spatial location and macroscopic energy distribution of target anomalies. Due to the long wavelength and strong penetrating power of low-frequency sound waves, the imaging results usually have good positional stability. However, due to the long wavelength of low-frequency waves, the imaging profile usually shows a wide energy band, which is difficult to use directly for the interpretation of target body boundaries.

[0045] Step S22, combined Figure 3 In order to transform the low-frequency imaging results into a spatial weight field that can participate in the high-frequency imaging constraint, the low-frequency reverse time migration imaging results are first processed. Envelope extraction:

[0046] In the formula, Represents the Hilbert transform. This represents the low-frequency imaging envelope. By extracting the envelope, the positive and negative oscillatory responses in low-frequency RTM images can be transformed into a stable energy distribution, reducing the influence of phase reversal, local ringing, and sidelobe effects on prior construction.

[0047] Subsequently, the low-frequency imaging envelope was normalized:

[0048] In the formula, This is the normalized low-frequency energy map, with values ​​ranging from [0,1].

[0049] To enhance the contrast between the target anomaly area and the background area, a nonlinear enhancement function is introduced:

[0050] In the formula, The image shows the low-frequency energy map after nonlinear enhancement, where α is the nonlinear enhancement coefficient, and preferably α>1. When α>1, the high-energy anomaly region is further highlighted, while the low-energy background region is suppressed.

[0051] Furthermore, anisotropic Gaussian diffusion kernels are used to smooth the diffusion of the enhanced low-frequency energy map:

[0052] In the formula, This is the initial low-frequency prior probability map. For Gaussian diffusion nuclei, For convolution operations, This represents the normalization operator.

[0053] The low-frequency prior probability map is used to characterize spatial location. The initial probability that the target anomalous entity exists at a given location. When the value is close to 1, it indicates that the region has a strong and stable anomalous response in low-frequency imaging, and the probability of the presence of a target is high; when... When the value is close to 0, it indicates that the area is more likely to be a background area or a weak scattering interference area.

[0054] Through the above processing, the low-frequency reverse time migration imaging results are transformed from the original imaging profile into a spatial prior probability map that can participate in the constraints of high-frequency reverse time migration imaging. This prior probability map can retain the stable positioning advantage of low-frequency imaging and provide target area guidance for subsequent high-frequency imaging.

[0055] Step S3: Using the low-frequency prior probability map as the imaging condition, obtain the high-frequency imaging result under the low-frequency prior constraint based on the high-frequency effective reflected wave data. Perform feature transformation on the high-frequency imaging result to obtain the high-frequency feedback probability map. Weightedly fuse the high-frequency feedback probability map with the low-frequency prior probability map to obtain the updated prior probability map.

[0056] Step S31, based on high-frequency effective reflected wave data High-frequency reverse time migration imaging (RTM) is performed. The results of conventional high-frequency RTM imaging are as follows:

[0057] In the formula, Forward propagation wave field of high-frequency source, For high-frequency detector backpropagation wave field, This is the result of conventional high-frequency reverse time migration imaging.

[0058] Because high-frequency sound waves are more sensitive to scattering objects such as small pebbles, random noise, and multiples, conventional high-frequency imaging results are prone to artifacts in non-target areas, streak interference, and energy diffusion. Therefore, the prior probability map of the k-th round is used... By introducing high-frequency reverse time migration imaging conditions, high-frequency imaging results under low-frequency prior constraints are obtained:

[0059] In the formula, This represents the high-frequency imaging result after low-frequency prior constraints in the k-th round.

[0060] when When the area is large, it is considered to have high target confidence, and the high-frequency imaging response is preserved or enhanced; when When the frequency is small, the region is more likely to be a background scattering area or a non-target region, and the high-frequency imaging response is weakened. Therefore, the low-frequency prior plays a positional constraint and energy guiding role in high-frequency reverse time migration, so that the high-frequency imaging energy is preferentially focused on the high-confidence anomaly region indicated by the low-frequency imaging.

[0061] Step S32: Using only low-frequency priors to constrain high-frequency imaging constitutes a one-way constraint. Due to the limited resolution of low-frequency imaging results, the low-frequency prior probability map may suffer from problems such as an overly wide target range, blurred boundaries, and multi-target response connectivity. Therefore, it is necessary to utilize the local structure and boundary details in the high-frequency imaging results to perform reverse correction on the low-frequency prior probability map.

[0062] Specifically, after the k-th round of low-frequency prior constraint high-frequency imaging, the current high-frequency imaging results are... Envelope extraction:

[0063] in, For high-frequency imaging envelope, This represents the Hilbert transform.

[0064] In complex, heterogeneous layers, high-frequency imaging results may exhibit localized strong responses that mask other truly weak responses. To avoid global normalization suppressing weak targets, this embodiment further introduces a local normalization strategy, normalizing different local regions separately within a fixed window:

[0065] In the formula, This is the normalized high-frequency energy diagram. For A local window centered on the center. This is a stability coefficient used to avoid the denominator being zero.

[0066] Further, a nonlinear enhancement function is introduced to construct the feedback probability graph:

[0067] In the formula, The high-frequency feedback probability map is constructed from the high-frequency imaging results of the kth round, where β is the high-frequency feedback enhancement coefficient, and preferably β>1.

[0068] Through the above processing, the high-frequency imaging results are transformed into a high-frequency feedback probability map. This feedback probability map reflects the local structure, boundary details, and target energy distribution information provided by high-frequency imaging, and can be used to correct problems such as excessively wide response, blurred boundaries, or inaccurate local positioning in low-frequency prior maps.

[0069] Step S33, after obtaining the high-frequency feedback probability map Then, it is compared with the prior probability diagram from the previous round. We perform weighted fusion to obtain the updated prior probability map:

[0070] In the formula, This is the prior probability graph for the (k+1)th round. Let be the feedback fusion weight for the kth round.

[0071] To ensure both stability and the ability to gradually absorb high-frequency boundary details during the iteration process, the fusion weights are... Set it to gradually increase with each iteration round:

[0072] In the formula, To achieve the maximum fusion weight, These are the parameters for controlling the iteration speed. With the above settings, more low-frequency priors are retained in the early stages of iteration to ensure the stability of the target position constraint; as iteration progresses, the role of high-frequency feedback is gradually enhanced, allowing the local structural and detail information contained in the high-frequency imaging results to progressively correct the prior probability map.

[0073] Step S4: Repeat the high-frequency imaging, high-frequency feedback probability map construction, and prior probability map update process under low-frequency prior constraints until the convergence criterion is met, and output the multi-frequency acoustic wave joint reverse time migration imaging results.

[0074] Repeatedly perform high-frequency reverse time migration imaging, high-frequency feedback probability map construction, and prior probability map update under low-frequency prior constraints until the change region of the prior probability map or high-frequency imaging result is stable.

[0075] Among them, the rate of change of imaging results is defined. for:

[0076] Define the rate of change of the prior plot for:

[0077] Set the first tolerance threshold. Second tolerance threshold When satisfied and When the iteration converges, the update stops, and the high-frequency imaging result obtained in the current round is output as the final multi-frequency acoustic wave joint reverse time migration imaging result:

[0078] In the formula, This represents the results of multi-frequency acoustic wave joint reverse time migration imaging under low-frequency a priori feedback iterative constraints, where K is the final iteration number. This represents the high-frequency imaging result after low-frequency prior constraints in the Kth round.

[0079] To verify the effectiveness of the method proposed in this invention, a double-drift stone model and a triple-drift stone model with a random background of small pebbles were constructed, and a comparative experiment of multi-frequency acoustic reverse time migration imaging was carried out.

[0080] In the dual-boulder model, two large boulder targets are set up with their spatially separated positions. The background medium consists of randomly distributed small pebbles of different sizes, used to simulate the complex pebble strata environment containing isolated boulders ahead of the tunnel face. Single-frequency reverse-time migration imaging was performed using 300Hz, 600Hz, and 800Hz acoustic data, and the results were compared with those of the method of this invention. The results are as follows: Figure 4 As shown.

[0081] Figure 4 (a) shows that the 300Hz low-frequency reverse time migration imaging results can reflect the position of the main boulder target relatively stably, and form obvious abnormal response near the target area. However, the range of abnormal response is wide, and the boundaries and local contours are relatively blurred. Figure 4 (b) shows that the local interface response of the target is enhanced in the 600Hz reverse time migration imaging results, and the energy near the target is relatively more concentrated. However, as the frequency increases, the interference of scattered waves caused by the background pebbles is significantly enhanced, and more arc-shaped oscillation stripes and local artifacts appear in the imaging profile. Figure 4 Figure (c) shows that the 800Hz high-frequency reverse time migration imaging results further improve the local boundary resolution, but it is more sensitive to small-scale inhomogeneities. The striped artifacts and stray energy caused by background pebble scattering are significantly enhanced, and some artifacts are superimposed with the real target response, making the interpretation of pebble boundaries uncertain. The multi-frequency joint imaging results obtained using the method of this invention are as follows: Figure 4 As shown in (d), the arc-shaped scattering artifacts and striped interference in the background area are significantly suppressed, the target's abnormal response is more concentrated, the spatial position of the double boulder target is clearer, and no obvious pseudo-connection phenomenon is observed between the targets.

[0082] Furthermore, the three-boulder model includes three large boulder targets located at different spatial positions, including a large-scale main anomaly and two smaller-scale anomalies. This is used to analyze the response differences of targets of different scales in acoustic imaging at different frequencies and to verify the invention's ability to locate and spatially separate multiple targets. Figure 5 As shown. Figure 5 In the 300Hz low-frequency reverse time migration imaging results, a more obvious anomalous response is formed near the large-scale boulder in the middle, indicating that the low-frequency sound wave can reflect the location of the main anomalous body relatively stably, but the response of small-scale targets is weak and the local boundary is difficult to be clearly identified. Figure 5 (b) shows the results of the 600Hz reverse time-shift imaging. The local interface response of the large boulder in the middle is enhanced, and the abnormal response of the small boulder is also improved. The spatial relationship of the three targets is clearer. However, the background scattering interference increases with the frequency and more striped oscillations appear in the imaging profile. Figure 5 Figure (c) shows that the 800Hz high-frequency reverse time migration imaging results further enhance the local response of the target, and some boundary details are more prominent. However, high-frequency sound waves are more sensitive to scattering by small pebbles, and more striped artifacts and stray energy are generated in non-target areas, indicating that weak target identification still has instability. The multi-frequency joint imaging results obtained by the method of this invention are as follows: Figure 5 As shown in (d), the striped artifacts and local stray energy in the non-target area are significantly suppressed, the abnormal response of the target body is more concentrated, the imaging focus of the large boulder in the middle is significantly improved, the small boulder in the lower left forms a relatively independent local abnormal display, and the area where the small boulder in the upper right is located also retains a certain abnormal indication. No obvious pseudo-connection phenomenon is found between multiple targets.

[0083] The experimental results above show that the method of the present invention, through low-frequency prior constraints and high-frequency feedback iteration mechanism, can effectively utilize the stable positioning capability of low-frequency imaging and the boundary resolution capability of high-frequency imaging, significantly suppress the pseudo-anomaly response under strong scattering background, improve the imaging focusing, boundary continuity and spatial separation capability of anomalies under multi-target conditions, and verify the effectiveness and reliability of the present invention in fine detection of adverse geological bodies in complex pebble strata.

[0084] Example 2 This embodiment discloses a shield tunnel acoustic wave detection migration imaging system based on low-frequency prior constraints; like Figure 6 As shown, the shield tunnel acoustic wave detection migration imaging system based on low-frequency prior constraints includes: The data acquisition and preprocessing module is configured to: acquire multi-frequency acoustic wave data during the shield tunneling process and preprocess it to obtain low-frequency effective reflected wave data and high-frequency effective reflected wave data; The low-frequency imaging and prior construction module is configured to: perform low-frequency reverse time migration imaging based on low-frequency effective reflected wave data to obtain low-frequency reverse time migration imaging results, and perform feature transformation on the low-frequency reverse time migration imaging results to obtain a low-frequency prior probability map. The high-frequency constrained imaging and feedback fusion module is configured to: use the low-frequency prior probability map as the imaging condition, obtain the high-frequency imaging result under the low-frequency prior constraint based on the high-frequency effective reflected wave data, perform feature transformation on the high-frequency imaging result to obtain the high-frequency feedback probability map, and weightedly fuse the high-frequency feedback probability map with the low-frequency prior probability map to obtain the updated prior probability map. The iterative control and imaging output module is configured to repeatedly execute the high-frequency imaging, high-frequency feedback probability map construction and prior probability map update process under low-frequency prior constraints until the convergence criterion is met, and output the multi-frequency acoustic wave joint reverse time migration imaging results.

[0085] Example 3 The purpose of this embodiment is to provide a computer-readable storage medium.

[0086] A computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps in the shield tunnel acoustic wave detection migration imaging method based on low-frequency prior constraints as described in Embodiment 1.

[0087] Example 4 The purpose of this embodiment is to provide an electronic device.

[0088] An electronic device includes a memory, a processor, and a program stored in the memory and executable on the processor. When the processor executes the program, it implements the steps in the shield tunnel acoustic wave detection migration imaging method based on low-frequency prior constraints as described in Embodiment 1.

[0089] Example 5 Embodiment 5 of the present invention provides a computer program product, including a computer program / instruction, which, when executed by a processor, implements the steps in the shield tunnel acoustic wave detection migration imaging method based on low-frequency prior constraints as described in Embodiment 1.

[0090] The steps and methods involved in the apparatuses of Embodiments 2, 3, 4, and 5 above correspond to those in Embodiment 1. For specific implementation details, please refer to the relevant description section of Embodiment 1. The term "computer-readable storage medium" should be understood as a single medium or multiple media including one or more instruction sets; it should also be understood as including any medium capable of storing, encoding, or carrying an instruction set for execution by a processor and enabling the processor to perform any of the methods in this invention.

[0091] Those skilled in the art will understand that the modules or steps of the present invention described above can be implemented using general-purpose computer devices. Optionally, they can be implemented using computer-executable program code, thereby allowing them to be stored in a storage device for execution by a computer device, or they can be fabricated as separate integrated circuit modules, or multiple modules or steps can be fabricated as a single integrated circuit module. The present invention is not limited to any particular combination of hardware and software.

[0092] While the specific embodiments of the present invention have been described above in conjunction with the accompanying drawings, this is not intended to limit the scope of protection of the present invention. Those skilled in the art should understand that various modifications or variations that can be made by those skilled in the art without creative effort based on the technical solutions of the present invention are still within the scope of protection of the present invention.

Claims

1. A method for acoustic wave detection migration imaging of shield tunnels based on low-frequency prior constraints, characterized in that, include: Multi-frequency acoustic wave data were collected and preprocessed during the tunnel boring process to obtain low-frequency effective reflected wave data and high-frequency effective reflected wave data. Low-frequency reverse-time migration imaging (RTM) is performed based on low-frequency effective reflected wave data to obtain RTM imaging results. Feature transformation is then applied to these results to obtain a low-frequency prior probability map. Specifically: the envelope of the RTM imaging results is extracted, and the positive and negative oscillatory responses are converted into energy-type distributions; the extracted low-frequency imaging envelope is normalized, mapping the amplitude to a standard numerical range to obtain a normalized low-frequency energy map; a nonlinear enhancement function is introduced to nonlinearly enhance the normalized low-frequency energy map, and the enhanced result is smoothed and diffused to obtain a continuously distributed low-frequency prior probability map. Using the low-frequency prior probability map as an imaging condition, the high-frequency imaging results under the low-frequency prior constraint are obtained based on the high-frequency effective reflected wave data, as shown below: In the formula, This represents the high-frequency imaging result after low-frequency prior constraints in the k-th round. This is the low-frequency prior probability graph for the k-th round; Forward propagation wave field of high-frequency source, For high-frequency detector backpropagation wave field; The high-frequency imaging results are subjected to feature transformation to obtain a high-frequency feedback probability map. This high-frequency feedback probability map is then weighted and fused with a low-frequency prior probability map to obtain an updated prior probability map. Specifically: the envelope of the high-frequency imaging results under prior constraints is extracted, and the oscillating response is converted into an energy distribution; the extracted high-frequency imaging envelope is locally normalized, with different local regions normalized within a fixed window, preserving the relative response intensity of weak signal regions; the locally normalized high-frequency energy map is nonlinearly enhanced to construct the high-frequency feedback probability map; and the high-frequency feedback probability map is weighted and fused with the current low-frequency prior probability map, with the fusion weight dynamically adjusted with each iteration to obtain the updated prior probability map. Repeat the high-frequency imaging, high-frequency feedback probability map construction, and prior probability map update process under low-frequency prior constraints until the convergence criterion is met, and output the multi-frequency acoustic wave joint reverse time migration imaging results.

2. The shield tunnel acoustic wave detection migration imaging method based on low-frequency prior constraints as described in claim 1, characterized in that, Low-frequency reverse-time migration imaging was performed based on low-frequency effective reflected wave data to obtain low-frequency reverse-time migration imaging results, including: Based on low-frequency effective reflected wave data, the source wavefield is extended forward and the receiver wavefield is extended backward using the acoustic wave equation. The low-frequency reverse time migration imaging results are obtained through cross-correlation imaging conditions. In the formula, This is a low-frequency reverse time migration imaging result. For low-frequency source wavefield, For low-frequency detector wave field, For spatial coordinates, For time.

3. The shield tunnel acoustic wave detection migration imaging method based on low-frequency prior constraints as described in claim 1, characterized in that, Repeat the high-frequency imaging, high-frequency feedback probability map construction, and prior probability map update process under low-frequency prior constraints until the convergence criterion is met, and output the multi-frequency acoustic wave joint reverse time migration imaging results, including: After each round of prior probability map update is completed, the next round of high-frequency imaging, high-frequency feedback probability map construction and prior probability map update process under low-frequency prior constraints is immediately executed. After each iteration, the rate of change between two adjacent high-frequency imaging results and the rate of change between two adjacent prior probability maps are calculated. When the rate of change of the imaging result is less than the preset first tolerance threshold and the rate of change of the prior probability map is less than the preset second tolerance threshold, the iteration is determined to be converged, the update is stopped, and the high-frequency imaging result obtained in the current round is output as the final multi-frequency acoustic wave joint reverse time migration imaging result.

4. A shield tunnel acoustic wave detection migration imaging system based on low-frequency prior constraints, characterized in that, include: The data acquisition and preprocessing module is configured to: acquire multi-frequency acoustic wave data during the shield tunneling process and preprocess it to obtain low-frequency effective reflected wave data and high-frequency effective reflected wave data; The low-frequency imaging and prior construction module is configured to: perform low-frequency reverse time migration imaging based on low-frequency effective reflected wave data to obtain low-frequency reverse time migration imaging results; perform feature transformation on the low-frequency reverse time migration imaging results to obtain a low-frequency prior probability map; specifically: extract the envelope of the low-frequency reverse time migration imaging results and convert the positive and negative oscillation responses into energy-type distributions; normalize the extracted low-frequency imaging envelope and map the amplitude to a standard numerical range to obtain a normalized low-frequency energy map; introduce a nonlinear enhancement function to nonlinearly enhance the normalized low-frequency energy map, and smooth the enhanced result to obtain a continuously distributed low-frequency prior probability map; The high-frequency constrained imaging and feedback fusion module is configured to: use the low-frequency prior probability map as the imaging condition, and obtain the high-frequency imaging result under the low-frequency prior constraint based on the high-frequency effective reflected wave data, as shown below: In the formula, This represents the high-frequency imaging result after low-frequency prior constraints in the k-th round. This is the low-frequency prior probability graph for the k-th round; Forward propagation wave field of high-frequency source, For high-frequency detector backpropagation wave field; The high-frequency imaging results are subjected to feature transformation to obtain a high-frequency feedback probability map. This high-frequency feedback probability map is then weighted and fused with a low-frequency prior probability map to obtain an updated prior probability map. Specifically: the envelope of the high-frequency imaging results under prior constraints is extracted, and the oscillating response is converted into an energy distribution; the extracted high-frequency imaging envelope is locally normalized, with different local regions normalized within a fixed window, preserving the relative response intensity of weak signal regions; the locally normalized high-frequency energy map is nonlinearly enhanced to construct the high-frequency feedback probability map; and the high-frequency feedback probability map is weighted and fused with the current low-frequency prior probability map, with the fusion weight dynamically adjusted with each iteration to obtain the updated prior probability map. The iterative control and imaging output module is configured to repeatedly execute the high-frequency imaging, high-frequency feedback probability map construction and prior probability map update process under low-frequency prior constraints until the convergence criterion is met, and output the multi-frequency acoustic wave joint reverse time migration imaging results.

5. A computer-readable storage medium having a program stored thereon, characterized in that, When the program is executed by the processor, it implements the steps in the shield tunnel acoustic wave detection migration imaging method based on low-frequency prior constraints as described in any one of claims 1-3.

6. An electronic device comprising a memory, a processor, and a program stored in the memory and executable on the processor, characterized in that, When the processor executes the program, it implements the steps in the shield tunnel acoustic wave detection migration imaging method based on low-frequency prior constraints as described in any one of claims 1-3.

7. A computer program product comprising a computer program / instructions, characterized in that, When the computer program / instruction is executed by the processor, it implements the steps in the shield tunnel acoustic wave detection migration imaging method based on low-frequency prior constraints as described in any one of claims 1-3.

Images