TBM tunneling disturbance area acoustic detection strong noise structure guided total variation denoising method

Abstract

The disclosure provides a TBM tunneling disturbance zone acoustic wave detection strong noise structure guided total variation denoising method, relates to the field of tunnel advanced geological prediction noise suppression technology, and comprises the following steps: acquiring an acoustic wave signal of array acoustic wave detection in a TBM disturbance zone; using a multi-scale structure tensor method to estimate the dip angle information and structural coherence of the acoustic wave signal, setting a coherence threshold, and dividing the acoustic wave signal into a signal region and a noise region; for the signal region, the dip angle information of the signal region is stable, a data dip angle guided directional total variation denoising method is used for noise suppression, the denoising intensity of the signal direction is weakened, and therefore the signal can be effectively protected; for the noise region, the noise region has no obvious dip angle information, and a total variation denoising method is directly used for noise suppression, so that the signal leakage and noise suppression are effectively balanced. The disclosure can effectively balance the signal fidelity and noise removal problem.

Description

Technical Field

[0001] This disclosure relates to the field of tunnel advanced geological prediction noise suppression technology, specifically to a structurally guided total variation denoising method for strong noise detection in the TBM excavation disturbance zone. Background Technology

[0002] The statements in this section are merely background information relating to this disclosure and do not necessarily constitute prior art.

[0003] Due to the tight construction process of TBM tunnels, which involves multiple trades working together, the data acquired on-site is subject to severe interference from complex mixed noise, including but not limited to equipment noise, mechanical noise, and construction noise. This noise significantly reduces the signal-to-noise ratio of the acquired data, blurring the effective reflection wave phase axis and hindering subsequent processing such as seismic wave velocity inversion, imaging, and interpretation. Therefore, denoising research on array acoustic wave detection data is crucial.

[0004] Currently, common noise reduction methods used in tunnel seismic wave advanced geological prediction technology include bandpass filtering, FK transform, and Radon transform. These methods primarily utilize the characteristics of noise and signal in terms of frequency, apparent velocity, and in-phase axis curvature to suppress noise. However, these methods have the following problems in application: (1) Reliance on expert experience. Data processing personnel need to have a certain understanding of the signal frequency band, apparent velocity range and in phase axis curvature in order to obtain better data processing results; traditional filtering algorithms mainly rely on spectral characteristics to suppress random noise, which is difficult to adapt to the non-stationary noise and multiple waves in the complex environment of TBM construction.

[0005] (2) It only utilizes a single feature of the signal and performs poorly in the complex noise environment of the tunnel; (3) The use of a single filter parameter globally makes it difficult to achieve a good balance between noise removal and signal protection, which can easily damage the effective signal. Summary of the Invention

[0006] To address the aforementioned issues, this disclosure proposes a structure-guided total variational denoising method for acoustic wave detection in the disturbance zone of TBM tunneling. This method utilizes the multi-scale structure tensor method to estimate the dip angle information and structural coherence of the acoustic data. Then, it leverages the correlation between local coherence and the acoustic signal to delineate signal and noise regions. For both signal and noise regions, an adaptive constraint denoising method based on the data dip angle signal effectively balances signal fidelity and noise removal.

[0007] According to some embodiments, the present disclosure adopts the following technical solutions: TBM tunneling disturbance zone acoustic detection of strong noise structure-guided total variation denoising method, including: Acquire acoustic signals from the TBM disturbance region array acoustic detection; The tilt angle information and structural coherence of the acoustic signal are estimated by using the multi-scale structural tensor method, and the coherence threshold is set by using the correlation between structural coherence and acoustic signal to divide the acoustic signal into signal region and noise region. A hybrid total variation denoising method based on data tilt angle is adopted to perform adaptive constraint denoising on the signal region and the noise region, thereby suppressing background noise and protecting the effective signal. The hybrid total variation denoising method based on data tilt angle guidance performs adaptive constraint denoising on the signal region and the noise region, including: For the signal region, where the tilt information is stable, a total variational denoising method guided by the data tilt angle is used for noise suppression. By reducing the denoising intensity in the signal direction, the signal can be effectively protected. For the noise region, where there is no obvious tilt information, a total variational denoising method is directly used for noise suppression, thereby effectively balancing signal leakage and noise suppression.

[0008] According to some embodiments, the present disclosure adopts the following technical solutions: A TBM tunneling disturbance zone acoustic detection system with strong noise structure-guided total variational denoising includes: The signal acquisition module is used to acquire the acoustic signals of the TBM disturbance zone array acoustic detection. The effective signal extraction module is used to estimate the tilt angle information and structural coherence of the acoustic signal using the multi-scale structural tensor method, and to set a coherence threshold using the correlation between structural coherence and acoustic signal to divide the acoustic signal into signal region and noise region. The adaptive denoising module is used to perform adaptive constraint denoising on the signal region and the noise region using a hybrid total variation denoising method guided by data tilt angle, so as to suppress background noise and protect the effective signal. The hybrid total variation denoising method based on data tilt angle guidance performs adaptive constraint denoising on the signal region and the noise region, including: For the signal region, where the tilt information is stable, a total variational denoising method guided by the data tilt angle is used for noise suppression. By reducing the denoising intensity in the signal direction, the signal can be effectively protected. For the noise region, where there is no obvious tilt information, a total variational denoising method is directly used for noise suppression, thereby effectively balancing signal leakage and noise suppression.

[0009] According to some embodiments, the present disclosure adopts the following technical solutions: A computer program product includes a computer program that, when executed by a processor, implements the TBM tunneling disturbance zone acoustic wave detection strong noise structure-guided total variation denoising method.

[0010] According to some embodiments, the present disclosure adopts the following technical solutions: A non-transitory computer-readable storage medium is provided for storing computer instructions, which, when executed by a processor, implement the TBM tunneling disturbance zone acoustic wave detection strong noise structure-guided total variation denoising method.

[0011] According to some embodiments, the present disclosure adopts the following technical solutions: An electronic device includes a processor, a memory, and a computer program; wherein the processor is connected to the memory, the computer program is stored in the memory, and when the electronic device is running, the processor executes the computer program stored in the memory to enable the electronic device to implement the TBM tunneling disturbance zone acoustic wave detection strong noise structure-guided total variation denoising method.

[0012] Compared with the prior art, the beneficial effects of this disclosure are as follows: This disclosure discloses a structure-guided total variational denoising method for strong noise detection in the disturbance zone of TBM tunneling. It utilizes the multi-scale structure tensor method to solve for the tilt angle information and structural coherence of the acoustic data. Then, by leveraging the correlation between structural coherence and the acoustic signal and setting a coherence threshold, the acoustic data is divided into signal and noise regions. In the signal region, where the tilt angle information is stable, a tilt-guided directional total variational denoising method is used. In the noise region, where there is no significant tilt angle information, a direct total variational denoising method is applied. Through this adaptive constraint denoising method, a balance can be effectively struck between signal fidelity and noise removal.

[0013] This disclosure presents a structure-guided total variational denoising method for acoustic wave detection in the disturbance zone of TBM tunneling, which introduces the multi-scale structure tensor (MST) method to estimate the dip angle of acoustic wave data. This method can accurately capture the dip angle information of the acoustic wave phase axis at different scales, exhibits strong noise robustness, and can still obtain stable dip angle and coherence information under the complex noise interference of TBM tunnels. It can provide reliable structural information for subsequent hybrid total variational denoising methods.

[0014] This disclosure presents a structure-guided total variational denoising method for strong noise detection in TBM tunneling disturbance zones. Specifically, it proposes a hybrid total variational denoising method (MST-HTV) based on structural tilt angle guidance. This method, building upon traditional total variational denoising, introduces the structural tilt angle and coherence information of the acoustic data. By utilizing the structural differences between noise and signal, it first achieves effective segmentation of the signal and noise regions. Then, in the signal region, it uses signal tilt angle information for constrained denoising, effectively preserving signal fidelity by reducing the denoising intensity along the signal direction. In the noise region, the traditional total variational denoising method is used to effectively remove noise. Therefore, this hybrid total variational denoising method effectively removes strong noise in TBM tunnels and protects the effective signal. Attached Figure Description The accompanying drawings, which form part of this disclosure, are used to provide a further understanding of this disclosure. The illustrative embodiments of this disclosure and their descriptions are used to explain this disclosure and do not constitute an undue limitation of this disclosure.

[0015] Figure 1 This is a flowchart of the TBM tunneling disturbance zone acoustic wave detection strong noise structure-guided total variation denoising method according to an embodiment of the present disclosure; Figure 2 The embodiments of this disclosure include real noise data and added noise; wherein, Figure 2 (a) in the data represents noisy and complex data. Figure 2 (b) in the figure represents the noise data that was added; Figure 3 The structure tilt angle and coherence are solved by the multi-scale structure tensor method of this disclosure in embodiments of the present disclosure; wherein, Figure 3 In the diagram, (a) represents the tilt angle of the complex data structure. Figure 3 (b) in the text represents the coherence of complex data; Figure 4 The present disclosure provides the MF denoising results and MF denoising residuals for complex data containing real noise, as described in this embodiment. Figure 4 (a) in the image represents the MF denoising result. Figure 4 In the diagram, (b) represents the residual from MF denoising; Figure 5 This disclosure provides TV denoising results and TV denoising residuals for complex data containing real noise, as described in embodiments of this disclosure. Figure 5 (a) in the image represents the TV denoising result. Figure 5 (b) in the figure represents the residual noise reduction in TV; Figure 6 The present disclosure provides the ST-DTV denoising results and ST-DTV denoising residuals for complex data containing real noise, as described in this embodiment. Figure 6(a) in the figure represents the ST-DTV denoising result. Figure 6 (b) in the figure represents the residual noise reduction of ST-DTV; Figure 7 The table below shows the MST-HTV denoising results and MST-HTV denoising residuals for complex data containing real noise, according to embodiments of this disclosure; wherein, Figure 7 (a) in the image represents the MST-HTV denoising result. Figure 7 (b) in the figure represents the residual noise reduction of MST-HTV. Detailed Implementation

[0016] The present disclosure will be further described below with reference to the accompanying drawings and embodiments.

[0017] It should be noted that the following detailed descriptions are illustrative and intended to provide further explanation of this disclosure. 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 disclosure pertains.

[0018] It should be noted that the terminology used herein is for the purpose of describing particular embodiments only and is not intended to limit the exemplary embodiments according to this disclosure. As used herein, the singular form is intended to include the plural form as well, unless the context clearly indicates otherwise. Furthermore, it should be understood that when the terms “comprising” and / or “including” are used in this specification, they indicate the presence of features, steps, operations, devices, components, and / or combinations thereof.

[0019] Example 1 One embodiment of this disclosure provides a structurally guided total variational denoising method for acoustic wave detection of strong noise in the disturbance zone of a TBM tunneling machine, the method steps of which include: Step 1: Acquire the acoustic signal from the TBM disturbance region array acoustic detection; Step 2: Estimate the tilt angle information and structural coherence of the acoustic signal using the multi-scale structural tensor method, and set a coherence threshold using the correlation between structural coherence and acoustic signal to divide the acoustic signal into signal region and noise region; Step 3: Adaptive constraint denoising is performed on the signal and noise regions using a data tilt-guided hybrid total variation denoising method to suppress background noise and protect the effective signal. The hybrid total variation denoising method based on data tilt angle guidance performs adaptive constraint denoising on the signal region and the noise region, including: For the signal region, where the tilt information is stable, a total variational denoising method guided by the data tilt angle is used for noise suppression. By reducing the denoising intensity in the signal direction, the signal can be effectively protected. For the noise region, where there is no obvious tilt information, a total variational denoising method is directly used for noise suppression, thereby effectively balancing signal leakage and noise suppression.

[0020] As one embodiment, this disclosure proposes a data tilt-guided hybrid total variation denoising method (MST-HTV). Specifically, this disclosure uses coherence as an evaluation index. By setting a coherence threshold, a data tilt-constrained total variation denoising method is used in signal regions containing acoustic data. This effectively protects the signal by reducing the denoising intensity in the signal direction. In unstructured noise regions, a traditional TV denoising method is used to avoid interference from erroneous structural tilt information on the TV denoising results. The specific implementation process is as follows: Step 1: Acquire the acoustic signal from the TBM disturbance region array acoustic detection; This disclosure describes a sensor array that acquires raw acoustic signals through high-precision synchronous sampling. The acquired acoustic signals contain valid signals. s and noise data n It can be represented as: (1) in, d This indicates observational data that contains noise.

[0021] Step 2: Use the multi-scale structure tensor method to estimate the tilt angle information and structural coherence of the acoustic signal; Specifically, acoustic data exhibits significant multi-scale and multi-dip characteristics, primarily manifested in the following ways: deeper structures are more gently sloping, while shallower structures possess rich layering details; dip angles vary dramatically in complex structural regions, even exhibiting multi-directional superposition. Therefore, dip angle estimation at a single scale is often insufficient to accurately describe the true geological structure. To address this, this disclosure introduces the Multi-Scale Structure Tensor (MST) method to estimate the dip angle of acoustic data. This method can adaptively capture acoustic phase axes at different scales and possesses strong noise robustness, providing reliable dip angle information for subsequent total variation denoising methods.

[0022] The multi-scale structure tensor method primarily quantifies the local structural features of a pixel by analyzing the gradients in its neighborhood. For a pixel with a tilt angle of... θ Regarding the phase axis of the acoustic wave data, the acoustic wave amplitude is along the normal direction of the phase axis. The changes are drastic, and along the same phase axis. The change is the most stable. Assume there are two points along the same phase axis, namely ( x , t)and( x +Δ x , t +Δ t If the energy change is such that the energy change is expressed by the following formula: (2) in, x , t The offset and sampling time of the acoustic wave data. E ( x , t ) represents a pixel ( x , t Relative changes in energy I ( x, t ) is a pixel ( x, t The amplitude of the sound wave signal at point ) I x , I t For signal energy in x , t partial derivative of direction ， superscript T This represents the transpose of a matrix.

[0023] For two-dimensional acoustic data, as shown in formula (2), the energy gradient determines the relative change in energy between two points. Let the intermediate term of formula (2) equal to: (3) in, J For the structure tensor, , and In practical applications, to avoid the influence of noise, windowing is usually required when obtaining the structure tensor. First, the standard deviation is used... σ The derivative of the Gaussian window function G σ After filtering the acoustic data, the first-order gradient is: (4) in, This indicates solving for the first-order partial derivative. This is a convolution operation. Based on this, a ring filter is used to filter the structure tensor. This filter aligns with the feature edges, accurately normalizing the edge energy to the feature centers. The filtered set of structure tensors can then be represented as: (5) in, R This represents an environmental filter. Based on this, the structure tensors at different scales are normalized, and at each acoustic data location ( x ,t The algorithm calculates the trace of the structure tensor at each scale. The trace represents the integral of the "gradient energy" in the neighborhood and measures the structural saliency of the pixel at that scale. Therefore, the larger the trace, the better it represents the structural saliency of the pixel at that scale. σ This allows for better representation of local information. Therefore, by using a one-dimensional maximum search, the optimal filter scale for each pixel can be obtained. σ : (6) Here, tr(·) represents the solution trace over the normalized structure tensor matrix. The optimal structure tensor obtained by extracting the structure tensor corresponding to the optimal filter scale pixel-by-pixel is: (7) The eigenvalues ​​of the optimal structure tensor can be solved using the following formula (8): (8) in λ 1. λ 2 represents the maximum and minimum eigenvalues ​​of the structure tensor, respectively, characterizing the maximum and minimum energy gradients of the local acoustic signal, and their corresponding eigenvectors. , These correspond to the directions of most drastic energy change (normal direction of the phase axis) and the gentlest change (tilt direction of the phase axis), respectively. The signal tilt angle can then be calculated. (9) Step 3: Set a coherence threshold based on the correlation between structural coherence and acoustic signal, divide the acoustic signal into signal region and noise region, and use a hybrid total variation denoising method based on data tilt angle to adaptively constrain denoising the signal region and noise region to suppress background noise and protect effective signal. First, denoising acoustic data can be viewed as an inverse problem, that is, denoising acoustic data containing noisy sound... d Recovering the effective signal s can be viewed as an optimization problem that minimizes the following formula: (10) in, J Represents the loss function. For the estimated effective signal, R The regularization term utilizes prior information to achieve better noise reduction. λ is the regularization factor, and ||·||² represents the regularization term. L 2-norm. Traditional total variation denoising methods (TV) use total variation to measure the first-order gradient sparsity of the data profile, which can preserve the edge information of the image relatively well. For a two-dimensional dataset... (t , x ),in t For the timeline, x Using the spatial axis, the TV denoising model can be expressed as: (11) Where ||·||1 represents the L1 norm. As shown in the above equation, the TV denoising method primarily penalizes the L1 norm of the gradient, thus exhibiting strong noise suppression while being relatively lenient towards the acoustic wave phase axis. This allows it to effectively remove noise while preserving the edge characteristics of the acoustic wave phase axis. However, the total variation method is isotropic, meaning its denoising capability is the same in every direction. When processing acoustic wave data with a significant phase axis, it is difficult to effectively balance signal leakage and noise suppression.

[0024] Therefore, this disclosure introduces the tilt angle of the acoustic data structure obtained in step 2, which weakens the denoising intensity along the phase axis direction and increases the denoising intensity in the normal direction of the phase axis. This method can be called the Data Tilt Angle Constrained Total Variation Denoising Method (DTV), specifically: (12) in, D θ This is the tilt direction matrix of the local structure of the in-phase axis. This is the matrix of normal directions of the phase axes. ω This is a scaling factor that can adjust the noise reduction intensity in the tangential and normal directions of the in-phase axis. ω ≤1. ω The smaller the value, the weaker the noise reduction strength in the direction of the same phase axis, and the better it can protect the same phase axis. When ω When = 1, the denoising intensity is consistent in both directions, making the DTV method equivalent to the TV method. The DTV denoising method can be expressed as the following optimization problem: (13) in, This is a scaling factor matrix. Let be the local rotation matrix formed by the structural tilt field. Both the TV and DTV denoising methods described above can be solved using the primal-dual algorithm. Compared to the TV denoising method, DTV, by introducing the structural tilt field, forms anisotropic constraints along / perpendicular to the in-phase axis, thus possessing both stronger noise robustness and better structural preservation capability.

[0025] Secondly, the tilt angle of acoustic data obtained by the multi-scale structure tensor method is severely affected by noise interference. In pure noise regions, the estimated tilt angle oscillates within the range of -90° to 90°. Directly using this tilt angle result would mislead the total variation method in noisy regions. However, in the coherence results, the phase axis of the acoustic data has stable directional characteristics, while the noise signal is highly random, resulting in a much greater coherence result for the phase axis of the acoustic data compared to the noise signal. Utilizing these characteristics, this disclosure proposes a hybrid total variation denoising method (MST-HTV) guided by data tilt angle. Specifically, coherence is used as an evaluation index. By setting a coherence threshold, a data tilt angle-constrained total variation denoising method is applied in signal regions containing acoustic data. By reducing the denoising intensity in the signal direction, the signal can be effectively protected. In unstructured noise regions, a traditional TV denoising method is used to avoid interference from erroneous structural tilt angle information.

[0026] The coherence of local structures can be evaluated using the eigenvalues ​​of the structure tensor. (14) Coherence is an evaluation index distributed in the range [0, 1]. When a correlation threshold is set, if the structural coherence C=0, the local structure exhibits obvious isotropy, that is, there is basically no change in the amplitude of the signal, and therefore the region is judged as noise; if the structural coherence C=1, the local structure exhibits obvious anisotropy, and then it is judged as a valid signal.

[0027] Furthermore, a coherence threshold is used to convert the coherence result into a binary mask matrix, thereby distinguishing between the in-phase region and the noise region of the acoustic data. The specific formula is as follows: (15) in α It is a binary mask matrix. τ This is the coherence threshold. It can be appropriately selected based on the coherence results. Based on this, the scaling factor in formula (13) is further optimized: (16) in, The scaling factor for hybrid total variation denoising, For the minimum value of the scaling factor, when α When =1, it proves that there is an in-phase axis for acoustic data. DTV is used for noise reduction; when α When =0, there is no obvious structure here, therefore In this case, TV noise reduction method is used. This disclosure sets... Finally, the hybrid total variation denoising method based on data tilt angle guidance proposed in this disclosure can be expressed as: (17) This disclosure employs a hybrid total variation denoising and random noise suppression method based on structure tilt angle guidance. The data is divided into signal and noise regions by coherence. Since the tilt angle of the signal region is stable, a total variation denoising method guided by data tilt angle is adopted, while a traditional TV denoising method is adopted in the unstructured noise region. This effectively suppresses random noise while protecting the effective signal.

[0028] Simulation Experiment In one embodiment, the data containing real noise used in this disclosure are as follows: Figure 2 As shown. Among them. Figure 2 The signal-to-noise ratio of the data shown in (a) is approximately 3.5 dB, as... Figure 2 As shown in (b) of the figure, the added noise has obvious non-stationarity and strong randomness, which puts forward higher requirements for the denoising method proposed in this disclosure.

[0029] Figure 3 This demonstrates the use of the multi-scale structure tensor method to obtain results on the tilt angle and coherence of complex data structures. For example... Figure 3 (a) and Figure 3 As shown in (b), Figure 3 (a) represents the tilt angle of the complex data structure. Figure 3 (b) shows the coherence of complex data, indicating that under real noise interference, the local stability of the structure tilt angle and coherence results, especially in the deep reflection region, decreases slightly. However, at the main reflection phase axis, the structure tilt angle and coherence maintain good continuity, and the overall reflection trend is clearly discernible. Therefore, although real noise may introduce some local disturbances, it does not destroy the identifiability of the main reflection wave phase axis. This verifies that the multi-scale structure tensor method still maintains good adaptability and effectiveness in real noisy and complex data.

[0030] Figure 4 , Figure 5 , Figure 6 as well as Figure 7 The denoising results and their corresponding residuals are shown respectively on the data containing real noise, obtained by applying median filtering (MF), total variation (TV), ST-DTV, and MST-HTV. Figure 4 As shown in (a), the MF method can preserve the main reflection events, but its denoising effect is limited, and strong random noise still exists in the background. Its denoising residual ( Figure 4 (b) shows a large amount of irregularly distributed noise texture, proving that some noise was not effectively removed, and there is also leakage of effective signal. Figure 5As shown in (a), the TV method significantly suppresses background noise and makes the data phase axis clearer, but the phase axis boundary shows slight blurring; its residual ( Figure 5 (b) is relatively clean, but a weak in-phase axis component can still be observed, indicating some signal leakage. Figure 6 As shown in (a), the ST-DTV method, with the help of structural tilt angle constraints, can better protect the effective signal and reduce leakage. Its residual plot ( Figure 6 In (b) of the sample, there is almost no obvious acoustic phase axis structure, only weak noise texture, indicating that the effective signal and noise are well separated. In comparison, the MST-HTV method performs the most balanced among the three methods: Figure 7 As shown in (a), the obtained data has a sharper in-phase axis, effectively alleviating the problem of edge blurring, and noise suppression is more sufficient. In the corresponding residuals ( Figure 7 (b) The information of the in-phase axis is almost invisible, and the noise distribution is uniform and stable, indicating that MST-HTV has the most ideal effect in protecting the effective signal and eliminating noise.

[0031] Overall, the MST-HTV method exhibits the best performance in maintaining the in-phase structure of the reflected wave and suppressing background noise, thus verifying that the hybrid total variation denoising method based on data tilt angle guidance proposed in this paper can be applied to suppressing real TBM construction noise.

[0032] Example 2 One embodiment of this disclosure provides a structurally guided total variation denoising system for acoustic detection of strong noise in the disturbance zone of a TBM tunneling machine, comprising: The signal acquisition module is used to acquire the acoustic signals of the TBM disturbance zone array acoustic detection. The effective signal extraction module is used to estimate the tilt angle information and structural coherence of the acoustic signal using the multi-scale structural tensor method, and to set a coherence threshold using the correlation between structural coherence and acoustic signal to divide the acoustic signal into signal region and noise region. The adaptive denoising module is used to perform adaptive constraint denoising on the signal region and the noise region using a hybrid total variation denoising method guided by data tilt angle, so as to suppress background noise and protect the effective signal. The hybrid total variation denoising method based on data tilt angle guidance performs adaptive constraint denoising on the signal region and the noise region, including: For the signal region, where the tilt information is stable, a total variational denoising method guided by the data tilt angle is used for noise suppression. By reducing the denoising intensity in the signal direction, the signal can be effectively protected. For the noise region, where there is no obvious tilt information, a total variational denoising method is directly used for noise suppression, thereby effectively balancing signal leakage and noise suppression.

[0033] Example 3 One embodiment of this disclosure provides a computer program product, including a computer program that, when executed by a processor, implements the TBM tunneling disturbance zone acoustic wave detection strong noise structure-guided total variation denoising method.

[0034] Example 4 One embodiment of this disclosure provides a non-transitory computer-readable storage medium for storing computer instructions. When the computer instructions are executed by a processor, they implement the TBM tunneling disturbance zone acoustic wave detection strong noise structure-guided total variation denoising method.

[0035] Example 5 One embodiment of this disclosure provides an electronic device, including: a processor, a memory, and a computer program; wherein the processor is connected to the memory, and the computer program is stored in the memory. When the electronic device is running, the processor executes the computer program stored in the memory to enable the electronic device to implement the TBM tunneling disturbance zone acoustic wave detection strong noise structure-guided total variation denoising method.

[0036] This disclosure is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of this disclosure. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create a machine for implementing the flowchart illustrations and / or block diagrams. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0037] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0038] While the specific embodiments of this disclosure have been described above in conjunction with the accompanying drawings, this is not intended to limit the scope of protection of this disclosure. 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 this disclosure are still within the scope of protection of this disclosure.

Claims

1. A structurally guided total variation denoising method for acoustic detection of strong noise in the disturbance zone of TBM tunneling, characterized in that, include: Acquire acoustic signals from the TBM disturbance region array acoustic detection; The tilt angle information and structural coherence of the acoustic signal are estimated by using the multi-scale structural tensor method, and the coherence threshold is set by using the correlation between structural coherence and acoustic signal to divide the acoustic signal into signal region and noise region. A hybrid total variation denoising method based on data tilt angle is adopted to perform adaptive constraint denoising on the signal region and the noise region, thereby suppressing background noise and protecting the effective signal. The hybrid total variation denoising method based on data tilt angle guidance performs adaptive constraint denoising on the signal region and the noise region, including: For the signal region, where the tilt information is stable, a total variational denoising method guided by the data tilt angle is used for noise suppression. By reducing the denoising intensity in the signal direction, the signal can be effectively protected. For the noise region, where there is no obvious tilt information, a total variational denoising method is directly used for noise suppression, thereby effectively balancing signal leakage and noise suppression.

2. The TBM tunneling disturbance zone acoustic wave detection strong noise structure-guided total variation denoising method as described in claim 1, characterized in that, The acoustic signal obtained from the TBM disturbance region array acoustic wave detection includes valid signal and noise data; the acoustic signal denoising process is regarded as an inverse problem, that is, recovering the valid signal from the noisy acoustic signal.

3. The TBM tunneling disturbance zone acoustic wave detection strong noise structure-guided total variation denoising method as described in claim 1, characterized in that, The method of estimating the tilt angle information and structural coherence of acoustic signals using the multi-scale structure tensor method includes: quantifying the local structural features of a pixel by analyzing the gradient in its neighborhood; and further, using the multi-scale structure tensor method to estimate the tilt angle information and structural coherence of the acoustic signal. First, the acoustic data is filtered using a derivative Gaussian window function. Then, a ring filter is used to filter the structure tensor. This filter can align with the feature edges and accurately normalize the edge energy to the feature center, thus obtaining the filtered set of structure tensors. The filtered structure tensors at different scales are normalized, and the trace of the structure tensor at each scale is solved at each acoustic data location. The optimal filter scale at each pixel is obtained through one-dimensional maximum search. By extracting the structure tensor corresponding to the optimal filter scale pixel by pixel, the optimal structure tensor is obtained, and the eigenvalues ​​of the optimal structure tensor are solved to obtain the signal structure tilt angle information. The coherence of the local structure is calculated using the eigenvalues ​​of the structure tensor, yielding the results of the local structure coherence.

4. The TBM tunneling disturbance zone acoustic wave detection strong noise structure-guided total variation denoising method as described in claim 1, characterized in that, The method of setting a coherence threshold based on the correlation between structural coherence and acoustic signal to divide the acoustic signal into a signal region and a noise region includes: Coherence is an evaluation index distributed between [0, 1]. When a correlation threshold is set, when the local coherence C=0, the local structure exhibits obvious isotropy, that is, there is basically no change in the amplitude of the signal, so the region is judged as noise; when the local coherence C=1, the local structure exhibits obvious anisotropy, so it is judged as a valid signal. By employing a coherence threshold, the coherence results are converted into a binary mask matrix, thereby distinguishing between the in-phase region and the noise region of the acoustic data.

5. The TBM tunneling disturbance zone acoustic wave detection strong noise structure-guided total variation denoising method as described in claim 1, characterized in that, For the signal region, where the tilt angle information is stable, a data tilt angle-guided total variation denoising method is used for noise suppression, including: The obtained acoustic signal structure tilt angle is introduced to reduce the noise reduction intensity along the phase axis direction, while increasing the noise reduction intensity in the phase axis normal direction; The scaling factor is set to adjust the denoising intensity in the tangential and normal directions of the in-phase axis. The smaller the scaling factor, the weaker the denoising intensity in the in-phase axis direction, which can protect the in-phase axis. When the scaling factor = 1, the denoising intensity in both directions is the same.

6. The TBM tunneling disturbance zone acoustic wave detection strong noise structure-guided total variation denoising method as described in claim 1, characterized in that, The method employs a hybrid total variation denoising approach guided by data tilt angle to adaptively constrain denoising of the signal and noise regions. The formula is as follows: Where J is the loss function, s , These are the valid signal and the valid signal obtained after denoising, respectively. d The original data, λ It is a scaling factor; TV and DTV denotes the traditional total variation denoising method and the total variation denoising method based on data tilt angle, respectively; C is structural coherence, and τ is the coherence threshold.

7. The TBM tunneling disturbance zone acoustic wave detection strong noise structure-guided total variation denoising method as described in claim 4, characterized in that, A coherence threshold is used to convert the coherence result into a binary mask matrix, thereby distinguishing between in-phase regions and noise regions in the acoustic data. The specific formula is as follows: in, α It is a binary mask matrix. τ This is the coherence threshold.

8. A computer program product, comprising a computer program, characterized in that, When the computer program is executed by the processor, it implements the TBM tunneling disturbance zone acoustic wave detection strong noise structure-guided total variation denoising method as described in any one of claims 1-7.

9. A non-transitory computer-readable storage medium, characterized in that, The non-transitory computer-readable storage medium is used to store computer instructions, which, when executed by a processor, implement the TBM tunneling disturbance zone acoustic wave detection strong noise structure-guided total variation denoising method as described in any one of claims 1-7.

10. An electronic device, characterized in that, include: The device includes a processor, a memory, and a computer program; wherein the processor is connected to the memory, the computer program is stored in the memory, and when the electronic device is running, the processor executes the computer program stored in the memory to enable the electronic device to implement the TBM tunneling disturbance zone acoustic wave detection strong noise structure-guided total variation denoising method as described in any one of claims 1-7.

Images