Intelligent surface wave noise suppression method and device
By constructing a large-scale sample label library and introducing a deep learning network with surface wave frequency and velocity constraints for the target work area, combined with an adaptive subtraction algorithm, the problem of poor surface wave noise suppression across work areas was solved, achieving high-precision intelligent surface wave noise suppression, which is suitable for seismic data processing.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- CHINA PETROLEUM & CHEMICAL CORP
- Filing Date
- 2024-12-26
- Publication Date
- 2026-06-26
AI Technical Summary
Existing intelligent surface wave noise suppression technology does not perform well in cross-processing applications. It requires the preparation of a large number of sample labels in the target processing area for network optimization, which is difficult to be accepted by production applications.
By constructing a large-scale sample label library, training a deep learning network, and introducing surface wave frequency and velocity constraints for the target work area, combined with an adaptive subtraction algorithm, high-precision denoising results are generated.
It achieves high-precision surface wave noise suppression without training and optimization in cross-regional applications, meeting production needs and improving the efficiency and quality of seismic data processing.
Smart Images
Figure CN122283918A_ABST
Abstract
Description
TECHNICAL FIELD
[0001] The present application relates to the technical field of geophysical exploration, in particular to an intelligent surface wave noise suppression method and device. BACKGROUND
[0002] As the main noise type in land seismic data, how to effectively suppress surface wave noise has always been the focus of research in the seismic processing flow. At present, the surface wave suppression technology based on conventional algorithms has been widely applied in commercial software, but the processing effect and efficiency are largely dependent on the experience of the processing personnel, and excessive manual dependence directly affects the progress and cycle of the processing project.
[0003] With the continuous development of artificial intelligence technology, surface wave noise suppression technology based on deep learning has also been rapidly researched and developed. However, in general, most of the intelligent surface wave noise suppression technologies are based on pure data driving or by introducing the constraint of surface wave physical characteristics in the loss function to improve the denoising accuracy. Such methods have good processing effect in the data of the work area involved in training, but there are still problems of poor processing effect in cross-work area direct application. Generally, a certain number of sample labels need to be prepared in the target processing work area for network optimization before inference, but this method generally cannot be accepted by production application because it needs to obtain a batch of good denoising results through preprocessing of the current work area.
[0004] Therefore, how to solve the above problems is a problem that needs to be solved at present. SUMMARY
[0005] The present application provides an intelligent surface wave noise suppression method and device, aiming to improve the above problems. In a first aspect, the present application provides an intelligent surface wave noise suppression method, which comprises:
[0006] inputting the noise data to be processed into a pre-trained denoising network model, and outputting predicted surface wave noise;
[0007] constraining the predicted surface wave noise according to the surface wave frequency of the target work area to generate first noise data after constraint;
[0008] constraining the noise data after constraint according to the velocity distribution of the surface wave of the target work area to generate second noise data;
[0009] generating target data after denoising according to the second noise data and the noise data to be processed.
[0010] In a possible embodiment, the training process of the denoising network model is as follows:
[0011] constructing a scale sample label library;
[0012] establishing a noise suppression deep learning network;
[0013] constructing a loss function, and training the noise suppression deep learning network by using sample labels in the scale sample label library to obtain a denoising network model.
[0014] In a possible embodiment, the loss function satisfies:
[0015] Q = Q1 + Q2 = ||Y label -Y pred || 2 +a·ssim(Y label -Y pred );
[0016] wherein Q1 is a two-norm constraint, and satisfies:
[0017] Q1 = ||Y label -Y pred || 2 ;
[0018] wherein Y label is label data, Y pred is the predicted noise of the surface wave, Q2 is a structural similarity constraint, and a represents a weight factor.
[0019] In a possible embodiment,
[0020] Q2 = a·ssim(Y label -Y pred );
[0021] wherein,
[0022] wherein μ Y represents a mean value of data Y, σ Y represents a standard deviation of data Y, and represents a covariance standard deviation of two data, and C1 and C2 are constants.
[0023] In a possible embodiment, the C1 satisfies:
[0024] C1 = K1L;
[0025] wherein K1 = 0.01, and L = max(Y label )-min(Y label ).
[0026] In a possible embodiment, the C2 satisfies:
[0027] C2 = K2L;
[0028] wherein K2 = 0.03.
[0029] In one possible embodiment, the first noise data satisfies:
[0030]
[0031] in, The first noise data is represented by f, which represents the high cutoff frequency of the surface wave in the target work area, and Freq_Filter represents the low-pass filter function.
[0032] In one possible embodiment, the second noise data satisfies:
[0033]
[0034] in, The second noise data is represented by Vel_MuteTaper, the cutoff function is represented by v, the surface wave velocity is represented by t, and the initial distribution time of the surface wave at zero offset is represented by t.
[0035] In one possible embodiment, the denoised target data satisfies:
[0036]
[0037] In the formula, X represents the noise data to be processed, M represents the matching operator of the adaptive least squares subtraction algorithm, and S represents the denoised target data.
[0038] Secondly, this application provides an intelligent surface wave noise suppression device, the device comprising:
[0039] The processing unit is used to input the noise data to be processed into the pre-trained denoising network model and output the predicted surface wave noise.
[0040] The first constraint unit is used to constrain the surface wave frequency of the target work area on the predicted surface wave noise and generate constrained first noise data.
[0041] The second constraint unit is used to constrain the constrained noise data according to the velocity distribution of the surface wave in the target work area, and generate the second noise data.
[0042] A denoising unit is configured to generate denoised target data based on the second noise data and the noise data to be processed.
[0043] The intelligent surface wave noise suppression method and apparatus provided in this application involve inputting the noise data to be processed into a pre-trained denoising network model, which outputs predicted surface wave noise; constraining the predicted surface wave noise to the surface wave frequency of the target work area, generating constrained first noise data; constraining the constrained noise data according to the velocity distribution of the surface waves in the target work area, generating second noise data; and generating denoised target data based on the second noise data and the noise data to be processed. This utilizes the trained denoising network model to process the noise data to be processed, and introduces the apparent velocity and frequency constraints of the surface wave noise in the current work area to achieve accurate prediction of surface waves, ultimately obtaining a high-precision denoising result. This facilitates better service for surface wave noise suppression in seismic data processing, providing technical support for the practical application of intelligent noise suppression. Attached Figure Description
[0044] To more clearly illustrate the technical solutions of the embodiments of this application, the accompanying drawings used in the embodiments will be briefly introduced below. It should be understood that the following drawings only show some embodiments of this application and should not be regarded as a limitation of the scope. For those skilled in the art, other related drawings can be obtained based on these drawings without creative effort.
[0045] Figure 1 This is a schematic diagram of the structure of an electronic device provided in the first embodiment of this application;
[0046] Figure 2 A flowchart illustrating an intelligent surface wave noise suppression method provided in the second embodiment of this application;
[0047] Figure 3 for Figure 2 The diagram shows the network architecture of a denoising network model in an intelligent surface wave noise suppression method.
[0048] Figure 4 This is a schematic diagram of the apparent velocity constraint for surface waves;
[0049] Figure 5 To adopt Figure 2 A schematic diagram showing the comparison between a smart surface wave noise suppression method and existing technologies after noise reduction;
[0050] Figure 6 This is a functional module diagram of an intelligent surface wave noise suppression device provided in the third embodiment of this application. Detailed Implementation
[0051] To make the objectives, technical solutions, and advantages of the embodiments of this application clearer, the technical solutions of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.
[0052] First embodiment:
[0053] Figure 1 This is a schematic diagram of the structure of an electronic device provided in an embodiment of this application. In this application, it can be... Figure 1 The schematic diagram illustrates an example electronic device 100 for implementing the intelligent surface wave noise suppression method and apparatus of the embodiments of this application. The electronic device 100 is used to perform the following steps:
[0054] The noise data to be processed is input into a pre-trained denoising network model, which outputs predicted surface wave noise. The predicted surface wave noise is constrained by the surface wave frequency of the target work area to generate constrained first noise data. The constrained noise data is then constrained according to the velocity distribution of the surface wave in the target work area to generate second noise data. The denoised target data is then generated based on the second noise data and the noise data to be processed.
[0055] like Figure 1 The diagram shows the structure of an electronic device 100. The electronic device 100 includes one or more processors 102, one or more storage devices 104, input devices 106, and output devices 108. These components are interconnected via a bus system and / or other forms of connection mechanisms (not shown). It should be noted that... Figure 1 The components and structure of the electronic device 100 shown are merely exemplary and not limiting; the electronic device may have, as needed. Figure 1 The components shown may also have Figure 1 Other components and structures not shown.
[0056] The processor 102 may be a central processing unit (CPU) or other processing unit with intelligent surface wave noise suppression capability and / or instruction execution capability, and may control other components in the electronic device 100 to perform the desired function.
[0057] It should be understood that the processor 102 in the embodiments of this application may be a central processing unit (CPU), or it may be other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor may be a microprocessor or any conventional processor, etc.
[0058] The storage device 104 may include one or more computer program products, which may include various forms of computer-readable storage media.
[0059] It should be understood that the storage device 104 in the embodiments of this application may be volatile memory or non-volatile memory, or may include both volatile and non-volatile memory. The non-volatile memory may be read-only memory (ROM), programmable read-only memory (PROM), erasable programmable read-only memory (EPROM), electrically erasable programmable read-only memory (EEPROM), or flash memory. The volatile memory may be random access memory (RAM), which is used as an external cache. By way of example, but not limitation, many forms of random access memory (RAM) are available, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate synchronous DRAM (DDR SDRAM), enhanced synchronous DRAM (ESDRAM), synchronous linked DRAM (SLDRAM), and direct rambus RAM (DR RAM).
[0060] The computer-readable storage medium may store one or more computer program instructions, which the processor 102 may execute to implement the client functions (implemented by the processor) in the embodiments of this application described below, and / or other desired functions. Various applications and various data may also be stored in the computer-readable storage medium, such as various data used and / or generated by the applications.
[0061] The input device 106 may be a device used by a user to input commands, and may include one or more of the following: keyboard, mouse, microphone, and touch screen.
[0062] Second embodiment:
[0063] Reference Figure 2 The flowchart shown illustrates an intelligent surface wave noise suppression method, which specifically includes the following steps:
[0064] Step S201: Input the noise data to be processed into the pre-trained denoising network model and output the predicted surface wave noise.
[0065] The training process of the denoising network model is as follows:
[0066] Build a large-scale sample label library;
[0067] Establish a noise suppression deep learning network;
[0068] A loss function is constructed, and the noise suppression deep learning network is trained using sample labels from the aforementioned sample label library to obtain a denoising network model.
[0069] The construction of the large-scale sample label library is as follows:
[0070] The generalization ability of data-driven intelligent surface wave noise suppression technology largely depends on the quantity and quality of sample labels. To improve the generalization ability of the denoising network, it is necessary to collect and prepare a certain scale of sample labels. However, it is unrealistic to achieve surface wave suppression for all types based on the same denoising network. Therefore, this invention mainly focuses on establishing basin-level samples for low-frequency, low-speed characteristic surface waves. By collecting data before and after denoising from multiple actual work areas, the sample labels are first screened after evaluation of the denoising effect by industry experts, and then further expanded through data augmentation. Data augmentation is mainly performed in the following two ways: (a) data mirroring and inversion; (b) data segmentation of each common shot set's single-arranged data along the channel containing the smallest absolute value of the offset distance to obtain two single-sided data points.
[0071] Optionally, the network adopts a fully convolutional UNET network structure, with the input data being the data before denoising and the output data being surface wave noise. A schematic diagram of the network structure is shown below. Figure 3As shown.
[0072] Optionally, the loss function satisfies:
[0073] Q = Q1 + Q2 = ||Y label -Y pred || 2 +a·ssim(Y label -Y pred );
[0074] Where Q1 is a 2-norm constraint, satisfying:
[0075] Q1 = ||Y label -Y pred || 2 ;
[0076] Among them, Y label For label data, Y pred Let Q2 be the predicted surface wave noise, Q2 be the structural similarity constraint, and α be the weighting factor.
[0077] Optionally, α is 0.15.
[0078] Alternatively, Q2 = a·ssim(Y) label -Y pred );
[0079] in,
[0080] Where, μ Y σ represents the mean of the data Y, while σ represents the mean of the data Y. Y This represents the standard deviation of data Y, while C1 and C2 represent the costandard deviations of the two data points, where C1 and C2 are constants.
[0081] Optionally, C1 satisfies: C1 = K1L; where K1 = 0.01, L = max(Y label )-min(Y label ).
[0082] Optionally, C2 satisfies: C2 = K2L; where K2 = 0.03.
[0083] Step S202: Apply target work area surface wave frequency constraints to the predicted surface wave noise to generate constrained first noise data.
[0084] Optionally, the first noise data satisfies:
[0085]
[0086] in, The first noise data is represented by f, which represents the high cutoff frequency of the surface wave in the target work area, and Freq_Filter represents the low-pass filter function.
[0087] Step S203: Constrain the constrained noise data according to the velocity distribution of the surface wave in the target work area to generate second noise data.
[0088] Optionally, the second noise data satisfies:
[0089]
[0090] in, This represents the second noise data, where Vel_MuteTaper represents the shearing function, v represents the surface wave velocity, and t represents the zero-offset surface wave initiation distribution time. For example... Figure 4 As shown, the user determines the surface wave distribution based on the velocity of the surface wave in the current work area and the zero offset time.
[0091] Step S204: Generate denoised target data based on the second noise data and the noise data to be processed.
[0092] Optionally, the denoised target data satisfies:
[0093]
[0094] In the formula, X represents the noise data to be processed, M represents the matching operator of the adaptive least squares subtraction algorithm, and S represents the denoised target data.
[0095] In summary, such as Figure 5 As shown, the intelligent surface wave noise suppression method provided in this embodiment was used to perform cross-work area application inference on actual data from an untrained location in eastern China. The processing results were compared with those of conventional commercial software. Figure 5 (b) Data after denoising using conventional algorithms. Figure 5 (c) represents the denoising result of cross-work area reasoning in this invention. Figure 5 (d) Noise data suppressed by conventional algorithms. Figure 5 In Figure (e), the noise data suppressed by the method of this invention is shown. As can be seen from the figure, the denoising effect obtained by this invention is significantly comparable to that of commercial software, verifying that the technical process proposed in this invention has good denoising accuracy in cross-work area denoising applications and can meet production needs.
[0096] Third embodiment:
[0097] Based on the same inventive concept, this application also provides an intelligent surface wave noise suppression device, such as... Figure 6As shown, the intelligent surface wave noise suppression device includes: a processing unit 510, a first constraint unit 520, a second constraint unit 530, and a noise reduction unit 540. The specific functions of each unit are as follows:
[0098] The processing unit 510 is used to input the noise data to be processed into a pre-trained denoising network model and output the predicted surface wave noise.
[0099] The first constraint unit 520 is used to constrain the surface wave frequency of the target work area on the predicted surface wave noise and generate constrained first noise data.
[0100] The second constraint unit 530 is used to constrain the constrained noise data according to the velocity distribution of the surface wave in the target work area, and generate second noise data.
[0101] The denoising unit 540 is used to generate denoised target data based on the second noise data and the noise data to be processed.
[0102] In one possible embodiment, the training process of the denoising network model is as follows:
[0103] Build a large-scale sample label library;
[0104] Establish a noise suppression deep learning network;
[0105] A loss function is constructed, and the noise suppression deep learning network is trained using sample labels from the aforementioned sample label library to obtain a denoising network model.
[0106] In one possible embodiment, the loss function satisfies:
[0107] Q = Q1 + Q2 = ||Y label -Y pred || 2 +a·ssim(Y label -Y pred );
[0108] Where Q1 is a 2-norm constraint, satisfying:
[0109] Q1 = ||Y label -Y pred || 2 ;
[0110] Among them, Y label For label data, Y pred Let Q2 be the predicted surface wave noise, Q2 be the structural similarity constraint, and α be the weighting factor.
[0111] In one possible embodiment,
[0112] Q2=a·ssim(Y label -Y pred );
[0113] in,
[0114] Where, μ Y σ represents the mean of the data Y, while σ represents the mean of the data Y. Y This represents the standard deviation of data Y, while C1 and C2 represent the costandard deviations of the two data points, where C1 and C2 are constants.
[0115] In one possible embodiment, C1 satisfies:
[0116] C1 = K1L;
[0117] Where K1 = 0.01, L = max(Y label )-min(Y label ).
[0118] In one possible embodiment, C2 satisfies:
[0119] C2 = K2L;
[0120] Where K2 = 0.03.
[0121] In one possible embodiment, the first noise data satisfies:
[0122]
[0123] in, The first noise data is represented by f, which represents the high cutoff frequency of the surface wave in the target work area, and Freq_Filter represents the low-pass filter function.
[0124] In one possible embodiment, the second noise data satisfies:
[0125]
[0126] in, The second noise data is represented by Vel_MuteTaper, the cutoff function is represented by v, the surface wave velocity is represented by t, and the initial distribution time of the surface wave at zero offset is represented by t.
[0127] In one possible embodiment, the denoised target data satisfies:
[0128]
[0129] In the formula, X represents the noise data to be processed, M represents the matching operator of the adaptive least squares subtraction algorithm, and S represents the denoised target data.
[0130] Furthermore, this embodiment also provides a computer-readable storage medium storing a computer program, which, when run by a processing device, executes the steps of any of the intelligent surface wave noise suppression methods provided in Embodiment 2 above.
[0131] The computer program product of the intelligent surface wave noise suppression method and apparatus provided in this application includes a computer-readable storage medium storing program code. The instructions included in the program code can be used to execute the methods described in the preceding method embodiments. For specific implementation, please refer to the method embodiments, which will not be repeated here.
[0132] In summary, the intelligent surface wave noise suppression method and apparatus provided in this application enhances the network's ability to identify noise by constructing a large-scale sample label library and introducing structural information constraints through a loss function. Furthermore, during cross-work area inference, the frequency and apparent velocity constraints of surface waves in the target work area are added, and an adaptive subtraction algorithm is combined to improve the network's denoising accuracy. Thus, intelligent processing of seismic surface wave noise suppression in cross-work area applications without the need for training and optimization is realized.
[0133] It should be noted that the above embodiments can be implemented, in whole or in part, by software, hardware (such as circuits), firmware, or any other combination thereof. When implemented using software, the above embodiments can be implemented, in whole or in part, in the form of a computer program product. The computer program product includes one or more computer instructions or computer programs. When the computer instructions or computer programs are loaded or executed on a computer, all or part of the processes or functions described in the embodiments of this application are generated. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, the computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via wired (e.g., infrared, wireless, microwave, etc.) means. The computer-readable storage medium can be any available medium that a computer can access or a data storage device such as a server or data center that includes one or more sets of available media. The available medium can be a magnetic medium (e.g., floppy disk, hard disk, magnetic tape), an optical medium (e.g., DVD), or a semiconductor medium. A semiconductor medium can be a solid-state drive.
[0134] It should be understood that the term "and / or" in this article is merely a description of the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A alone, A and B simultaneously, or B alone. A and B can be singular or plural. Additionally, the character " / " in this article generally indicates an "or" relationship between the preceding and following related objects, but it can also represent an "and / or" relationship. Please refer to the context for a more accurate understanding.
[0135] In this application, "at least one" means one or more, and "more than one" means two or more. "At least one of the following" or similar expressions refer to any combination of these items, including any combination of single or multiple items. For example, at least one of a, b, or c can mean: a, b, c, a-b, a-c, b-c, or a-b-c, where a, b, and c can be single or multiple.
[0136] It should be understood that in the various embodiments of this application, the order of the above-mentioned processes does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of this application.
[0137] Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.
[0138] Those skilled in the art will understand that, for the sake of convenience and brevity, the specific working processes of the systems, devices, and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.
[0139] In the several embodiments provided in this application, it should be understood that the disclosed systems, apparatuses, and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between apparatuses or units may be electrical, mechanical, or other forms.
[0140] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.
[0141] In addition, the functional units in the various embodiments of this application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit.
[0142] The above description is merely a preferred embodiment of this application and is not intended to limit this application. Various modifications and variations can be made to this application by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the protection scope of this application. It should be noted that similar reference numerals and letters in the following figures indicate similar items; therefore, once an item is defined in one figure, it does not need to be further defined and explained in subsequent figures.
Claims
1. An intelligent surface wave noise suppression method, characterized in that, The method includes: The noise data to be processed is input into a pre-trained denoising network model, which outputs predicted surface wave noise. The predicted surface wave noise is constrained by the surface wave frequency of the target work area to generate the first noise data after constraint; The constrained noise data is constrained based on the velocity distribution of the surface wave in the target work area to generate second noise data. Based on the second noise data and the noise data to be processed, the denoised target data is generated.
2. The method according to claim 1, characterized in that, The training process of the denoising network model is as follows: Build a large-scale sample label library; Establish a deep learning network for noise suppression; A loss function is constructed, and the noise suppression deep learning network is trained using sample labels from the aforementioned sample label library to obtain a denoising network model.
3. The method according to claim 2, characterized in that, The loss function satisfies: Q=Q1+Q2=||Y label -THE pred || 2 +a·max(Y label -THE pred ): Where Q1 is a 2-norm constraint, satisfying: Q1=||Y label -Y pred || 2 ; Among them, Y label For label data, Y pred Let Q2 be the predicted surface wave noise, Q2 be the structural similarity constraint, and α be the weighting factor.
4. The method according to claim 3, characterized in that, Q2=a·ssim(Y label -THE pred ): in, Where, μ Y σ represents the mean of the data Y, while σ represents the mean of the data Y. Y This represents the standard deviation of data Y, while C1 and C2 represent the costandard deviations of the two data points, where C1 and C2 are constants.
5. The method according to claim 4, characterized in that, C1 satisfies: C1 = K1L; Where K1 = 0.01, L = max(Y label )-min(Y label ).
6. The method according to claim 5, characterized in that, C2 satisfies: C2 = K2L; Where K2 = 0.
03.
7. The method according to any one of claims 3-6, characterized in that, The first noise data satisfies: in, The first noise data is represented by f, which represents the high cutoff frequency of the surface wave in the target work area, and Freq_Filter represents the low-pass filter function.
8. The method according to claim 7, characterized in that, The second noise data satisfies: in, The second noise data is represented by Vel_MuteTaper, the cutoff function is represented by v, the surface wave velocity is represented by t, and the initial distribution time of the surface wave at zero offset is represented by t.
9. The method according to claim 8, characterized in that, The denoised target data satisfies: In the formula, X represents the noise data to be processed, M represents the matching operator of the adaptive least squares subtraction algorithm, and S represents the denoised target data.
10. An intelligent surface wave noise suppression device, characterized in that, The device includes: The processing unit is used to input the noise data to be processed into the pre-trained denoising network model and output the predicted surface wave noise. The first constraint unit is used to constrain the surface wave frequency of the target work area on the predicted surface wave noise and generate constrained first noise data. The second constraint unit is used to constrain the constrained noise data according to the velocity distribution of the surface wave in the target work area, and generate the second noise data. The denoising unit is used to generate denoised target data based on the second noise data and the noise data to be processed.