Method and device for suppressing multiple waves in marine seismic data

Abstract

The present disclosure provides a seabed seismic wave multiple wave suppression method and device, comprising: obtaining a seabed seismic wave to be processed; performing a prediction operation on the seabed seismic wave to obtain first prediction data; inputting the first prediction data into a seismic wave suppression model to obtain refined multiple wave prediction; and calculating to obtain a seabed seismic primary wave according to the refined multiple wave prediction and the seabed seismic wave. Through self-refinement of the differentiable physical forward closed-loop constraint driven network, the dependence on the real primary wave label is eliminated. The network with the physical forward closed-loop constraint is iterated to realize high-precision reconstruction of the primary wave field, and finally realize high-precision unsupervised multiple wave suppression. The technical scheme of the present application can realize direct multiple wave suppression under the condition of no label, and solve the dependence of the supervised method on the label data and the linearization limitation of the traditional method.

Description

Technical Field

[0001] This disclosure relates to the field of signal processing technology, and in particular to a method and apparatus for suppressing multiple waves of submarine seismic data. Background Technology

[0002] When seismic waves propagate underground, they are reflected when they encounter interfaces with differences in wave impedance. If they propagate downwards again and are reflected a second or multiple times by other interfaces, they are eventually received by the geophone, forming multiples. Multiples represent interference waves formed after seismic waves have undergone multiple reflections and scatterings between underground strata interfaces. They differ from the effective primary wave in propagation path and time, but are highly similar in waveform, confusing the true reflection interfaces of the strata and severely interfering with imaging and interpretation. Conventional methods struggle to completely separate them and easily damage the effective signal, making them a key and challenging noise problem in seismic exploration. Suppression refers to identifying and separating interfering multiples from seismic records through prediction, filtering, inversion, or deep learning, weakening or eliminating them while preserving as much of the effective primary wave as possible, thereby improving the accuracy of seismic imaging and making the underground geological structure clearer and more reliable.

[0003] As marine seismic exploration advances into deep-water, complex tectonic zones, free-surface correlated multiples in marine seismic data severely interfere with seismic imaging quality, making multiple suppression a core aspect of seismic data processing. Existing multiple suppression techniques suffer from the following main shortcomings:

[0004] Some existing techniques have limited accuracy in obtaining the amplitude and phase of the primary wave. The subsequent adaptive subtraction process requires fine matched filtering, the parameter selection depends on human experience, and when the dynamic characteristics of the primary wave and the multiple waves are similar at close offset, it is very easy to cause energy loss of the primary wave.

[0005] Other existing technologies require a large amount of accurate primary-to-multiple wave pairing label data for training. However, obtaining real, pure primary wave labels is extremely difficult in actual exploration, and training usually relies on synthetic data, which limits the model's generalization ability on real data.

[0006] Other existing technologies face the problem of difficulty in capturing the complex nonlinear characteristics of seismic wave propagation and low computational efficiency. Summary of the Invention

[0007] In view of this, the purpose of this disclosure is to propose a method and apparatus for suppressing multiple waves of submarine seismic data, which at least to some extent solves one of the technical problems in the related art.

[0008] To achieve the above objectives, the first aspect of the exemplary embodiments of this disclosure provides a method for suppressing multiple waves of submarine seismic waves, comprising: The system obtains submarine earthquake data to be processed; performs prediction calculations on the submarine earthquake data to obtain first predicted data; inputs the first predicted data into an initial Fourier neural operator network to obtain second predicted data; inputs the submarine earthquake data and the second predicted data into a joint loss function to obtain a score; in response to the score being not less than a preset threshold, adjusts the parameters of the initial Fourier neural operator network to obtain an intermediate Fourier neural operator network, uses the intermediate Fourier neural operator network to replace the initial Fourier neural operator network for iterative training until the score is less than the preset threshold, and outputs the second predicted data; and calculates the first wave of the submarine earthquake based on the second predicted data and the submarine earthquake data.

[0009] Based on the same inventive concept, a second aspect of the exemplary embodiments of this disclosure provides a device for suppressing multiple waves of submarine seismic data, specifically comprising: The data acquisition module is configured to acquire submarine seismic data to be processed; The estimation module is configured to perform estimation calculations on submarine earthquake data to obtain first prediction data; The neural operator network operation module is configured to input first predicted data into an initial Fourier neural operator network to obtain second predicted data; input the submarine earthquake data and the second predicted data into a joint loss function to obtain a score; in response to the score being not less than a preset threshold, adjust the parameters of the initial Fourier neural operator network to obtain an intermediate Fourier neural operator network; use the intermediate Fourier neural operator network to replace the initial Fourier neural operator network for iterative training until the score is less than the preset threshold, and then output the second predicted data; and The calculation module is configured to calculate the primary wave of a submarine earthquake based on the second prediction data and the submarine earthquake data.

[0010] Based on the same inventive concept, a third aspect of the exemplary embodiments of this disclosure provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the method as described in the first aspect.

[0011] Based on the same inventive concept, a fourth aspect of the exemplary embodiments of this disclosure provides a non-transitory computer-readable storage medium storing computer instructions for causing a computer to perform the method as described in the first aspect.

[0012] Based on the same inventive concept, a fifth aspect of the exemplary embodiments of this disclosure provides a computer program product including computer program instructions that, when run on a computer, cause the computer to perform the method as described in the first aspect.

[0013] As can be seen from the above, the method and apparatus provided in this disclosure drive the network to refine itself through differentiable physical forward modeling closed-loop constraints, completely eliminating the dependence on real primary wave labels. High-precision reconstruction of the primary wave field is achieved through network iteration with physical forward modeling closed-loop constraints, and the physical constraints ensure the feasibility of the solution, ultimately achieving high-precision unsupervised multiple wave suppression. The technical solution of this application can realize direct multiple wave suppression under label-free conditions, solving the dependence of supervised methods on label data and the linearization limitations of traditional methods, and can provide a purer primary wave field for migration imaging. Attached Figure Description

[0014] To more clearly illustrate the technical solutions in this disclosure or related technologies, the accompanying drawings used in the description of the embodiments or related technologies will be briefly introduced below. Obviously, the accompanying drawings described below are only embodiments of this disclosure. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0015] Figure 1A The prior art prediction-subtraction method based on SRME is disclosed; Figure 1B The prediction-subtraction method based on supervised deep learning in the prior art is disclosed; Figure 1C A prediction-subtraction method based on traditional physical constraints in the prior art is disclosed. Figure 2 The illustration shows a flowchart of a method for suppressing multiple waves of submarine seismic data according to an embodiment of the present disclosure; Figure 3 A schematic diagram illustrating the flow of a multi-scale iterative algorithm based on Fourier transform according to an embodiment of the present disclosure is shown. Figure 4 A schematic diagram illustrating the forward closed-loop process according to an embodiment of the present disclosure is shown. Figure 5 This schematic diagram illustrates the structure of a submarine seismic data multiple wave suppression device according to an embodiment of the present disclosure; Figure 6 The diagram illustrates the structure of a submarine seismic data multiple wave suppression device according to an embodiment of the present disclosure. Detailed Implementation

[0016] It is understood that before using the technical solutions disclosed in the various embodiments of this application, users should be informed of the types, scope of use, and usage scenarios of the personal information involved in this application in an appropriate manner in accordance with relevant laws and regulations, and user authorization should be obtained.

[0017] For example, upon receiving a user's active request, a prompt message is sent to the user to explicitly inform them that the requested operation will require the acquisition and use of the user's personal information. This allows the user to independently choose whether to provide personal information to the software or hardware, such as the electronic device, application, server, or storage medium performing the operations of this application's technical solution, based on the prompt message.

[0018] As an optional but non-limiting implementation, in response to a user's active request, sending a prompt message to the user can be done via a pop-up window, where the prompt message can be presented in text format. Furthermore, the pop-up window can also include a selection control allowing the user to choose "agree" or "disagree" to provide personal information to the electronic device.

[0019] It is understood that the above notification and user authorization process are merely illustrative and do not constitute a limitation on the implementation of this application. Other methods that comply with relevant laws and regulations may also be applied to the implementation of this application.

[0020] It is understood that the data involved in this technical solution (including but not limited to the data itself, the acquisition or use of the data) shall comply with the requirements of relevant laws, regulations and related provisions.

[0021] To make the objectives, technical solutions, and advantages of this disclosure clearer, the principles and spirit of this disclosure will be described below with reference to several exemplary embodiments. It should be understood that these embodiments are provided merely to enable those skilled in the art to better understand and implement this disclosure, and are not intended to limit the scope of this disclosure in any way. Rather, these embodiments are provided to make this disclosure more thorough and complete, and to fully convey the scope of this disclosure to those skilled in the art.

[0022] In this article, it is important to understand that any number of elements in the accompanying figures is for illustrative purposes only and not for limitation, and any naming is for distinction only and has no limiting meaning.

[0023] It should be noted that, unless otherwise defined, the technical or scientific terms used in the embodiments of this disclosure should have the ordinary meaning understood by one of ordinary skill in the art to which this disclosure pertains. The terms "first," "second," and similar terms used in the embodiments of this disclosure do not indicate any order, quantity, or importance, but are merely used to distinguish different components. Terms such as "comprising" or "including" mean that the element or object preceding the word covers the element or object listed after the word and its equivalents, without excluding other elements or objects. Terms such as "connected" or "linked" are not limited to physical or mechanical connections, but can include electrical connections, whether direct or indirect. Terms such as "upper," "lower," "left," and "right" are used only to indicate relative positional relationships; when the absolute position of the described object changes, the relative positional relationship may also change accordingly. The article "a" or "an" preceding an element does not exclude the existence of multiple such elements.

[0024] The principles and spirit of this disclosure will be explained in detail below with reference to several representative embodiments. Invention Overview As mentioned earlier, when seismic waves propagate underground, they are reflected when they encounter interfaces with differences in wave impedance. If they propagate downwards again and are reflected a second or multiple times by other interfaces, they are eventually received by the geophone, forming multiple waves. Multiple waves represent interference waves formed after seismic waves have undergone multiple reflections and scatterings between underground strata interfaces. While their propagation paths and times differ from those of the effective primary waves, they are highly similar in waveform, confusing the true reflecting interfaces of the strata and severely interfering with imaging and interpretation. Conventional methods struggle to completely separate them and easily damage the effective signal, making them a critical and challenging noise problem in seismic exploration.

[0026] The suppression method described in this disclosure uses techniques such as prediction, filtering, inversion, or deep learning to identify and separate interfering multiples from seismic records, weakening or eliminating them while preserving as many effective primarys as possible, thereby improving the accuracy of seismic imaging and making underground geological structures clearer and more reliable.

[0027] As marine seismic exploration advances into deep-water complex tectonic zones, the correlation multiples on the free surface of the seabed severely interfere with the quality of seismic imaging, making multiple suppression a core aspect of seismic data processing.

[0028] specifically refer to Figure 1A , Figure 1A A prior art prediction-subtraction method based on SRME is disclosed. According to... Figure 1AIn step S1, the original seismic data is input, and the SRME data-driven convolution method is used to coarsely predict the multiple wave model. In step S2, adaptive matched filtering is used to finely match the predicted multiple waves, with parameters selected based on human experience. In step S3, the original data and the matched multiple wave model are subtracted to achieve multiple wave suppression. SRME predicts the multiple wave model through data-driven convolution, but its prediction results are only coarse estimates with limited amplitude and phase accuracy. The subsequent adaptive subtraction process requires fine matched filtering, and parameter selection relies on human experience. Furthermore, when the dynamic characteristics of the primary and multiple waves are similar at close offsets, primary wave energy loss is easily caused.

[0029] specifically refer to Figure 1B , Figure 1B This paper discloses a prior art method based on supervised deep learning. In step S10, a labeled dataset containing a large amount of accurate primary-multiple wave pairings is constructed and input for training the supervised deep learning model. In step S20, the model completes training using synthetic data. And in step S30, the trained model is applied to real seismic data to achieve separation and prediction of primary and multiple waves. This type of method requires a large amount of accurate primary-multiple wave pairing labeled data for training. However, obtaining accurate and pure primary wave labels in actual exploration is extremely difficult, and training usually relies on synthetic data, which limits the model's generalization ability on real data.

[0030] specifically refer to Figure 1C , Figure 1C This paper discloses an existing inversion method based on traditional physical constraints. In step S100, a seismic wave inversion model is established based on physical laws, and prior constraints such as sparsity are introduced. In step S200, a linearization approximation is used to simplify the complex seismic wave field propagation process. And in step S300, numerical inversion calculations are performed to separate the primary and secondary waves. For example, inversion methods based on the sparsity assumption, although physically meaningful, usually use linearization approximations, making it difficult to capture the complex nonlinear characteristics of seismic wave propagation, and have low computational efficiency.

[0031] To address the aforementioned issues, this disclosure proposes an unsupervised iterative model framework for obtaining primary waves from submarine seismic data, utilizing a forward modeling closed loop to validate a Fourier neural operator network. Specifically, this disclosure provides a method and apparatus for suppressing multiple waves from submarine seismic data. The method includes: acquiring submarine seismic data to be processed; performing prediction operations on the submarine seismic data to obtain first predicted data; inputting the first predicted data into an initial Fourier neural operator network to obtain second predicted data; inputting the submarine seismic data and the second predicted data into a joint loss function to obtain a score; adjusting the parameters of the initial Fourier neural operator network to obtain an intermediate Fourier neural operator network in response to the score being not less than a preset threshold; using the intermediate Fourier neural operator network to replace the initial Fourier neural operator network for iterative training until the score is less than the preset threshold, and outputting the second predicted data; and calculating the primary wave of the submarine seismic data based on the second predicted data and the submarine seismic data. This technical solution drives network self-refinement through differentiable physics forward modeling closed-loop constraints, completely eliminating the dependence on actual primary wave labels. High-precision reconstruction of the primary wavefield was achieved through network iteration with physical forward modeling closed-loop constraints. The physical constraints ensured the feasibility of the solution, ultimately realizing high-precision unsupervised multiple wave suppression. The technical solution of this application can realize direct multiple wave suppression under label-free conditions, solving the dependence of supervised methods on label data and the linearization limitations of traditional methods, and can provide a purer primary wavefield for migration imaging.

[0032] After introducing the basic principles of this disclosure, various non-limiting embodiments of this disclosure will be described in detail below.

[0033] Exemplary methods In some embodiments, specifically refer to Figure 2 , Figure 2 This illustration schematically depicts a method for suppressing multiple waves of submarine seismic data according to an embodiment of the present invention. Figure 2 : In step 201, the submarine seismic data to be processed is obtained. The submarine seismic data to be processed can be obtained by methods such as towed cable, OBC (Ocean Bottom Cable Seismic System), OBS (Ocean Bottom Seismometer), pre-stack shot gathers / common shot points / common receiver points / common offset gathers acquired by submarine nodes or submarine cables.

[0034] Specifically, the submarine seismic data to be processed obtained in step 201 Within the time domain, this includes primary waves, which serve as effective signals and carry information about subsurface structures. As interference signals, multiple waves generated by reflection from free surfaces and as random noise Submarine earthquake data The expression is shown in formula (1):

[0035] The independent variable The distance between the earthquake source and the detection equipment is represented by the independent variable. Indicates time. The submarine seismic data can be processed before proceeding with further processing steps. Filtering is performed to remove random noise. In a possible implementation, the filtering step can be performed using frequency-wavenumber domain filtering (f... Methods such as k-filtering, FX-domain prediction filtering (FX denoising), or median / mean filtering are used to achieve this, which will not be elaborated upon here. The purpose is to transform submarine seismic data into... .

[0036] In step 201, the raw submarine seismic observation data can also undergo preprocessing steps such as frequency band limitation, amplitude normalization, and gather regularization.

[0037] In step 202, the submarine earthquake data is used to perform prediction calculations to obtain the first prediction data.

[0038] Specifically, the first predicted data can be used to refer to the coarse multiple prediction value of the multiple component in submarine seismic data. This step can generate the first predicted data by performing frequency domain convolution / wavenumber domain cross-correlation on the original seismic data based on surface-related multiple prediction using the SRME (Surface-related multiple elimination) algorithm. SRME is a data-driven surface-related multiple prediction theory. For the specific formula for calculating coarse multiples using SRME, please refer to formula (2):

[0039] in The horizontal wavenumber describes the lateral variation characteristics of the wave field in the spatial domain and corresponds to the spatial frequency. Angular frequency describes how fast the wave field vibrates in the time domain, corresponding to the time frequency; For the predicted multiple wave spectrum function, The original data spectrum function, This is the free surface reflectance, which is approximately -1 at the sea surface. The scattered wavenumber variable characterizes the incident wavenumber component generated by the reflection of the source wavefield through the subsurface; for Summation is used to achieve cross-correlation of wave fields in the wavenumber domain, thus completing the frequency-wavenumber domain modeling of SRME multiple waves.

[0040] After obtaining the first predicted data, the reference wavefield is obtained by subtracting it from the submarine earthquake data. The construction formula of the reference wavefield is shown in formula (3):

[0041] in For reference wave field, For submarine earthquake data, This serves as the first predicted data. The reference wavefield contains key characteristic information of the primary wave and is used as a benchmark for the forward consistency verification of the subsequent first scoring function, providing directional guidance for subsequent unsupervised model iterations. Specific methods are detailed below.

[0042] Step 203: Input the first predicted data into the initial Fourier neural operator network to obtain the second predicted data. Specifically, this step can be achieved by inputting the first predicted data generated by the SRME algorithm into a two-dimensional Fourier neural operator network with parameter-initialized residual connections, and outputting the second predicted data.

[0043] specifically refer to Figure 3 , Figure 3 The iterative algorithm flow based on Fourier transform is shown. This figure illustrates the iterative solution process for the seismic / wave field, including time-domain processing, frequency-domain processing, and iterative update processing. When converting from the time domain to the frequency domain, the time-domain wave field... Perform a Fast Fourier Transform to obtain the frequency domain components. Frequency domain filtering uses operators Only the low-frequency / low-wavenumber components in the upper left corner are retained, while high frequencies are filtered out to achieve downscaling. When transforming back from the frequency domain to the time domain, a fast inverse Fourier transform is performed on the filtered components to restore the low-frequency, large-scale field in the time domain. The low-frequency components in the frequency domain are then subjected to a local linear transformation. And superimposed residuals The compensation details are locally corrected. This is achieved by fusing the low-frequency field and the correction component, followed by nonlinear constraints. get This completes one iteration. The overall formula for the iterative solution process can be expressed as formula (4):

[0044] Specifically, in the technical solution disclosed herein, the initial Fourier neural operator network learns multiple operators in the Fourier space, and the single-layer mapping of this network can be expressed as formula (5):

[0045] in, For the first Layer input features, For the first Layer output features, The activation function is GELU. This is a local linear transformation (1×1 convolution in the spatial domain). and These represent the Fourier transform and its inverse transform, respectively. For frequency domain learnable complex weights (integral operators). This represents the residual term for this layer.

[0046] Furthermore, at least one two-dimensional Fourier convolutional layer in the two-dimensional Fourier neural operator network may include the following processing steps: For input features (e.g., defined as) The input features are subjected to a two-dimensional fast Fourier transform along the time sampling dimension and the gather space dimension to obtain two-dimensional frequency domain features; a preset number of low-frequency modes are retained in the two-dimensional frequency domain features, and high-frequency modes are discarded or set to zero; the retained low-frequency modes are multiplied by the frequency domain learnable complex weight tensor to obtain the frequency domain mapping result; a two-dimensional fast inverse Fourier transform is performed on the frequency domain mapping result to obtain the frequency domain convolution output; at the same time, the input features are subjected to a spatial domain 1×1 convolution to obtain a local linear output; and the frequency domain convolution output, the local linear output and the residual term of the input features are added together and subjected to an activation function to obtain the next layer output features. The activation function can be the GELU activation function. The above processing formula can be expressed as formula (6) and formula (7):

[0047]

[0048] in, and These represent the number of low-frequency modes retained in the time and spatial dimensions, respectively. This represents learnable complex weights in the frequency domain. Represents the 1×1 convolution weights in the spatial domain. The terminator represents the frequency domain zero-padding operation, and Trunc represents the low-frequency mode truncation operation. and These represent the two-dimensional Fourier transform and the two-dimensional inverse Fourier transform, respectively.

[0049] In Fourier neural operator networks, residual connection mechanisms can also be introduced into each Fourier convolutional layer. Specifically, the introduced residuals include frequency domain convolution operators, whose output at that layer is represented as... The spatial domain convolution operator, whose output at this layer is represented as ; and the output of this layer The three values ​​are then summed after parallel computation to obtain the final output of the layer. Specifically, the residual connection mechanism is referenced in formula (8):

[0050] Residual connections effectively alleviate the gradient vanishing problem in deep networks and improve iterative stability.

[0051] In step 203, the technical solution of this disclosure may further include parameter initialization of the initial Fourier neural operator network. Specifically, since the parameters of the Fourier neural operator network in the multiple wave suppression model need to change continuously during the iteration of the multiple wave suppression model, the initial parameters of the initial Fourier neural operator network can be randomly initialized or pre-trained parameter initialization, or preset initial values ​​set by the user or other users based on relevant knowledge of certain seismic images. This method is understandable to those skilled in the art upon reading the above content, and will not be elaborated further in this disclosure.

[0052] In step 204, the submarine seismic data and the second predicted data are input into a joint loss function to obtain a score. The specific methods included in step 204 are detailed below.

[0053] In step 205, in response to a score not being less than a preset threshold, the parameters of the initial Fourier neural operator network are adjusted to obtain an intermediate Fourier neural operator network. This intermediate Fourier neural operator network is then used to replace the initial Fourier neural operator network for iterative training until the score is less than the preset threshold, at which point the second prediction data is output. Specifically, this step designs a scoring criterion for iterative iteration of the multiple wave suppression model using a joint loss function. The multiple wave suppression model continuously adjusts its parameters based on the score obtained from this scoring criterion, resulting in increasingly smaller errors in the multiple wave suppression model's results.

[0054] In one possible implementation, if the score remains above a preset threshold after multiple iterations, second prediction data can be output based on other preset conditions. These other preset conditions may include being based on the maximum epoch, early stopping of validation loss, etc. It is worth noting that these other preset conditions are not limited to the above explanations; those skilled in the art can adopt feasible preset conditions according to their actual needs, as long as the final output second prediction data meets the accuracy requirements. The term "multiple times" in this paragraph is a limitation that can be understood by those skilled in the art and can be fully explained using their knowledge; it will not be elaborated upon here.

[0055] The above embodiments may include the model output field being mapped by a forward modeling operator to maintain a high degree of consistency with the reference wave field described above in terms of numerical and waveform characteristics, reflecting the data fidelity of the reconstruction result; the reconstructed signal having sparse characteristics in the spatiotemporal domain, with concentrated energy distribution and effective suppression of background components, presenting sparse and clean characteristics; and the signal changing continuously and smoothly in the time and space dimensions, without drastic jumps, pseudo-oscillations and non-physical abrupt changes, ensuring the structural continuity of the field.

[0056] This method also includes step 206, which calculates the first wave of the submarine earthquake based on the second predicted data and the submarine earthquake data. Specifically, this calculation method may be to subtract the second predicted data from the submarine earthquake data to obtain the first wave of the submarine earthquake. The submarine earthquake data may be the original submarine earthquake data, or the submarine earthquake data after removing random noise through the filtering method described above, or the submarine earthquake data after any one or more preprocessing methods described in step 201.

[0057] In one possible implementation, the joint loss function includes a first back-calculation function and a first scoring function. Inputting the submarine seismic data and the second predicted data into the joint loss function to obtain a score includes: The submarine earthquake data and the second prediction data are input into the first inverse function (also known by those skilled in the art as the forward modeling closed-loop function, the specific principle of which will be discussed later). Figure 4 Steps 402 and 403) are used to obtain error data for the reference wavefield and synthetic seismic data. This error data is used to accurately evaluate each training iteration, thereby precisely updating the parameters in the Fourier neural operator network. The specific implementation method for this step is detailed below.

[0058] The reference wavefield of the submarine seismic data and the error data of the synthetic seismic data are input into the first scoring function to obtain a score. This first scoring function uses multiple mechanisms to score the model output for each training iteration. The specific implementation method for this step is detailed below.

[0059] specifically refer to Figure 4 In one possible implementation, inputting the submarine earthquake data and the second predicted data into a first back-calculation function to obtain error data for the submarine earthquake data and the synthetic earthquake data includes: In step 401, the second predicted data is subtracted from the submarine earthquake data to obtain the estimated first wave data.

[0060] Specifically, step 401 can be expressed as formula (9):

[0061] in, Represents submarine earthquake data, This represents the second predicted data. This represents the estimated value of the first wave of an undersea earthquake.

[0062] In step 402, the estimated primary wave data and the preset wavelet data are deconvolved to obtain the reflection coefficient data.

[0063] Specifically, due to the physical meaning, the estimated value of the first wave of a submarine earthquake... It can be regarded as a wavelet With reflection coefficient The convolution of is specifically represented by formula (10):

[0064] The expression for the wavelet is given by formula (11):

[0065] in Main frequency, For time.

[0066] As can be seen from the above, the estimation value of the first wave of a submarine earthquake can be obtained. Perform deconvolution calculations to obtain the reflection coefficient function model. .

[0067] In step 403, the reflection coefficient data is convolved with the preset wavelet data to obtain synthetic seismic wave data.

[0068] The reflection coefficient function model is convolved with a preset wavelet model to obtain the synthetic seismic wave. Specifically, the formula for the synthetic seismic wave is expressed by formula (12):

[0069] in To synthesize seismic waves, For the reflection coefficient function model, This is a wavelet. The combination of steps 402 and 403 above realizes the first inverse function.

[0070] And in step 404, error data is calculated based on the reference wavefield and the synthetic seismic wave data. Specifically, step 404 can be implemented by subtracting the reference wavefield from the synthetic seismic wave data. Those skilled in the art know that since both submarine seismic data and synthetic seismic wave data are vectors, other methods can also be used, such as the average of the differences in each dimension of the vector, the root mean square, etc.

[0071] In one possible implementation, the first scoring function includes an error loss term and a physical constraint term; the error loss term is used to obtain the first score from the error data of the submarine seismic data and the synthetic seismic waves.

[0072] Specifically, the error loss term This is used to ensure that the reference wavefield can be reconstructed after physical forward modeling of the estimated first-order wave. Error loss term. This can be expressed as formula (13):

[0073] in, The normalization coefficient for the total data volume. Represents the number of earthquake traces, This refers to the number of sampling points per channel. This is an estimate of the first wave of an undersea earthquake. For the forward propagation operator, This indicates the first wave of a submarine earthquake after forward calculation. Dao Di The estimated value of each sampling point As a pseudo-first wave reference wavefield, formula (11) calculates the squared error between the estimated value and the true value for all channels and all sampling points, and then sums and averages the results to obtain the error loss term. The smaller this loss term is, the closer the model output is to the actual observation data after passing through the forward operator, and the higher the reconstruction accuracy. This is used to evaluate the main scoring criteria mentioned above.

[0074] Furthermore, the physical constraints include sparsity constraints, total variation constraints, and wave equation constraints; wherein the sparsity constraints are used to obtain a second score by describing the sparsity of at least one of the estimated primary wave data, reflection coefficient data, or intermediate physical quantities derived from the estimated primary wave data. The total variation constraints are used to obtain a third score by describing the continuity and smoothness of at least one of the estimated primary wave data, reflection coefficient data, or intermediate physical quantities derived from the estimated primary wave data. The wave equation constraints are used to obtain a fourth score by describing the physical realizability of at least one of the estimated primary wave data, reflection coefficient data, or intermediate physical quantities derived from the estimated primary wave data.

[0075] The rating includes a first rating, a second rating, a third rating, and a fourth rating.

[0076] In one possible implementation, the reference wavefield is calculated based on the first predicted data and the submarine earthquake data; the reference wavefield is used as a reference target for calculating the error of the synthetic seismic wave data as a first scoring function when iterating the Fourier neural operator network.

[0077] In one possible implementation, a filtering step is included before performing prediction calculations on the submarine seismic data to obtain first prediction data. The filtering step is used to remove random noise from the submarine seismic data.

[0078] It should be noted that the method of this disclosure embodiment can be executed by a single device, such as a computer or server. The method of this embodiment can also be applied to a distributed scenario, where multiple devices cooperate to complete the task. In such a distributed scenario, one of these devices may execute only one or more steps of the method of this disclosure embodiment, and the multiple devices will interact with each other to complete the method described.

[0079] It should be noted that the above description describes some embodiments of this disclosure. Other embodiments are within the scope of the appended claims. In some cases, the actions or steps recorded in the claims can be performed in a different order than that shown in the above embodiments and still achieve the desired result. Furthermore, the processes depicted in the drawings do not necessarily require a specific or sequential order to achieve the desired result. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0080] Exemplary device Based on the same inventive concept, corresponding to any of the above embodiments, this disclosure also provides a device for suppressing multiple waves of submarine seismic data.

[0081] refer to Figure 5 The submarine seismic data multiple wave suppression device includes: The data acquisition module is configured to acquire submarine seismic data to be processed; The estimation module is configured to perform estimation calculations on submarine earthquake data to obtain first prediction data; The neural operator network operation module is configured to input the first predicted data into the initial Fourier neural operator network to obtain the second predicted data, input the submarine earthquake data and the second predicted data into the joint loss function to obtain a score, and adjust the parameters of the initial Fourier neural operator network to obtain an intermediate Fourier neural operator network in response to the score being not less than a preset threshold. The intermediate Fourier neural operator network is used to replace the initial Fourier neural operator network for iterative training until the score is less than the preset threshold, and then the second predicted data is output. And a calculation module, configured to calculate the first wave of a submarine earthquake based on the second prediction data and the submarine earthquake data.

[0082] For ease of description, the above apparatus is described in terms of its functions, divided into various modules. Of course, in implementing this disclosure, the functions of each module can be implemented in one or more software and / or hardware.

[0083] The apparatus described above is used to implement the corresponding submarine seismic data multiple wave suppression method in any of the foregoing embodiments, and has the beneficial effects of the corresponding method embodiments, which will not be repeated here.

[0084] Based on the same inventive concept, corresponding to the methods of any of the above embodiments, this disclosure also provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the submarine seismic data multiple wave suppression method described in any of the above embodiments.

[0085] Figure 6 This embodiment illustrates a more specific hardware structure of an electronic device, which may include a processor 1010, a memory 1020, an input / output interface 1030, a communication interface 1040, and a bus 1050. The processor 1010, memory 1020, input / output interface 1030, and communication interface 1040 are interconnected internally via the bus 1050.

[0086] The processor 1010 can be implemented using a general-purpose CPU (Central Processing Unit), microprocessor, application-specific integrated circuit (ASIC), or one or more integrated circuits, and is used to execute relevant programs to implement the technical solutions provided in the embodiments of this specification.

[0087] The memory 1020 can be implemented in the form of ROM (Read-Only Memory), RAM (Random Access Memory), static storage device, dynamic storage device, etc. The memory 1020 can store the operating system and other applications. When the technical solutions provided in the embodiments of this specification are implemented by software or firmware, the relevant program code is stored in the memory 1020 and is called and executed by the processor 1010.

[0088] The input / output interface 1030 is used to connect input / output modules to realize information input and output. The input / output modules can be configured as components in the device (not shown in the figure) or externally connected to the device to provide corresponding functions. Input devices may include keyboards, mice, touch screens, microphones, various sensors, etc., and output devices may include displays, speakers, vibrators, indicator lights, etc.

[0089] The communication interface 1040 is used to connect a communication module (not shown in the figure) to enable communication between this device and other devices. The communication module can communicate via wired means (such as USB, Ethernet cable, etc.) or wireless means (such as mobile network, WIFI, Bluetooth, etc.).

[0090] Bus 1050 includes a pathway for transmitting information between various components of the device, such as processor 1010, memory 1020, input / output interface 1030, and communication interface 1040.

[0091] It should be noted that although the above-described device only shows the processor 1010, memory 1020, input / output interface 1030, communication interface 1040, and bus 1050, in specific implementations, the device may also include other components necessary for normal operation. Furthermore, those skilled in the art will understand that the above-described device may only include the components necessary for implementing the embodiments of this specification, and not necessarily all the components shown in the figures.

[0092] The electronic devices described above are used to implement the corresponding submarine seismic data multiple wave suppression method in any of the foregoing embodiments, and have the beneficial effects of the corresponding method embodiments, which will not be repeated here.

[0093] The memory 1020 stores machine-readable instructions executable by the processor 1010. When the electronic device is running, the processor 1010 communicates with the memory 1020 via the bus 1050, causing the processor 1010 to execute the following instructions during operation: Obtain the submarine seismic data to be processed; The submarine earthquake data is used to perform estimation calculations to obtain the first prediction data; The first prediction data is input into the initial Fourier neural operator network to obtain the second prediction data; Input the submarine earthquake data and the second prediction data into the joint loss function to obtain a score; In response to a score not being less than a preset threshold, the parameters of the initial Fourier neural operator network are adjusted to obtain an intermediate Fourier neural operator network. This intermediate network is then used to replace the initial network for iterative training until the score is less than the preset threshold, at which point second prediction data is output. The first wave of the submarine earthquake is calculated based on the second prediction data and the submarine earthquake data.

[0094] Through the above method, the electronic device enables the self-refinement of the differential physical forward modeling closed-loop constraint-driven network, completely eliminating the dependence on real primary wave labels. High-precision reconstruction of the primary wave field is achieved through network iteration with physical forward modeling closed-loop constraints. The physical constraints guarantee the feasibility of the solution, ultimately achieving high-precision unsupervised multiple wave suppression. The technical solution of this application can realize direct multiple wave suppression under label-free conditions, solving the dependence of supervised methods on label data and the linearization limitations of traditional methods, and providing a purer primary wave field for migration imaging.

[0095] Exemplary program product Based on the same inventive concept, corresponding to the methods of any of the above embodiments, this disclosure also provides a non-transitory computer-readable storage medium storing computer instructions for causing the computer to execute the submarine seismic data multiple wave suppression method as described in any of the above embodiments.

[0096] The computer-readable medium of this embodiment includes permanent and non-permanent, removable and non-removable media, and information storage can be implemented by any method or technology. Information can be computer-readable instructions, data structures, program modules, or other data. Examples of computer storage media include, but are not limited to, phase-change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technologies, CD-ROM, digital versatile optical disc (DVD) or other optical storage, magnetic tape, magnetic disk storage or other magnetic storage devices, or any other non-transfer medium that can be used to store information accessible by a computing device.

[0097] The aforementioned non-transitory computer-readable storage media can be any available medium or data storage device that a computer can access, including but not limited to magnetic storage (e.g., floppy disks, hard disks, magnetic tapes, magneto-optical disks (MOs), etc.), optical storage (e.g., CDs, DVDs, BDs, HVDs, etc.), and semiconductor storage (e.g., ROMs, EPROMs, EEPROMs, non-volatile memory (NAND flash), solid-state drives (SSDs)).

[0098] The computer instructions stored in the storage medium of the above embodiments are used to cause the computer to execute the submarine seismic data multiple wave suppression method as described in any of the embodiments in the exemplary method section above, and have the beneficial effects of the corresponding method embodiments, which will not be repeated here.

[0099] Based on the same inventive concept, corresponding to the submarine seismic data multiple suppression method described in any of the above embodiments, this disclosure also provides a computer program product, which includes computer program instructions. In some embodiments, the computer program instructions can be executed by one or more processors of a computer to cause the computer and / or the processor to perform the submarine seismic data multiple suppression method. Corresponding to the execution entity for each step in each embodiment of the submarine seismic data multiple suppression method, the processor executing the corresponding step can belong to the corresponding execution entity.

[0100] The computer program product of the above embodiments is used to cause the computer and / or the processor to execute the submarine seismic data multiple wave suppression method as described in any of the above embodiments, and has the beneficial effects of the corresponding method embodiments, which will not be repeated here.

[0101] Those skilled in the art will recognize that embodiments of this disclosure can be implemented as a system, method, or computer program product. Therefore, this disclosure can be implemented as entirely hardware, entirely software (including firmware, resident software, microcode, etc.), or a combination of hardware and software, generally referred to herein as a "circuit," "module," or "system." Furthermore, in some embodiments, this disclosure can also be implemented as a computer program product contained in one or more computer-readable media, which includes computer-readable program code.

[0102] Any combination of one or more computer-readable media may be used. A computer-readable medium can be a computer-readable signal medium or a computer-readable storage medium. A computer-readable storage medium can be, for example,, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination thereof. More specific examples (not exhaustive) of a computer-readable storage medium may include: an electrical connection having one or more wires, a portable computer disk, a hard disk, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fiber, portable compact disk read-only memory (CD-ROM), optical storage device, magnetic storage device, or any suitable combination thereof. In this document, a computer-readable storage medium can be any tangible medium that contains or stores a program that can be used by or in connection with an instruction execution system, apparatus, or device.

[0103] Computer-readable signal media may include data signals propagated in baseband or as part of a carrier wave, carrying computer-readable program code. Such propagated data signals may take various forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination thereof. Computer-readable signal media may also be any computer-readable medium other than computer-readable storage media, capable of sending, propagating, or transmitting programs for use by or in connection with an instruction execution system, apparatus, or device.

[0104] Program code contained on a computer-readable medium may be transmitted using any suitable medium, including but not limited to wireless, wire, optical fiber, RF, etc., or any suitable combination thereof.

[0105] Computer program code for performing the operations of this disclosure can be written in one or more programming languages ​​or a combination thereof, including object-oriented programming languages ​​such as Java, Smalltalk, and C++, and conventional procedural programming languages ​​such as the "C" language or similar programming languages. The program code can be executed entirely on the user's computer, partially on the user's computer, as a standalone software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In cases involving remote computers, the remote computer can be connected to the user's computer via any type of network, including a local area network (LAN) or a wide area network (WAN), or it can be connected to an external computer (e.g., via the Internet using an Internet service provider).

[0106] It should be understood that each block of a flowchart and / or block diagram, and every combination of blocks in a flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, or other programmable data processing device to produce a machine that, when executed by a computer or other programmable data processing device, creates means for implementing the functions / operations specified in the blocks of the flowchart and / or block diagram.

[0107] These computer program instructions may also be stored in a computer-readable medium that enables a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer-readable medium produce a product comprising an instruction apparatus that implements the functions / operations specified in the boxes of a flowchart and / or block diagram.

[0108] Computer program instructions may also be loaded onto a computer, other programmable data processing apparatus, or other device to cause a series of operational steps to be performed on the computer, other programmable data processing apparatus, or other device to produce a computer-implemented process, such that the instructions that execute on the computer or other programmable apparatus can provide a process for implementing the functions / operations specified in the boxes of a flowchart and / or block diagram.

[0109] Furthermore, although the operations of the methods of this disclosure are described in a specific order in the accompanying drawings, this does not require or imply that these operations must be performed in that specific order, or that all of the operations shown must be performed to achieve the desired result. Rather, the steps depicted in the flowcharts may be executed in a different order. Additionally or alternatively, certain steps may be omitted, multiple steps may be combined into one step, and / or one step may be broken down into multiple steps.

[0110] The flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of systems, methods, and computer program products according to various embodiments of this application. Each block in a flowchart or block diagram may represent a module, segment, or portion of code, which contains one or more executable instructions for implementing a specified logical function. It should also be noted that in some alternative implementations, the functions indicated in the blocks may occur in a different order than those indicated in the drawings. For example, two consecutively indicated blocks may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved. It should also be noted that each block in a block diagram or flowchart, and combinations of blocks in a block diagram or flowchart, can be implemented using a dedicated hardware-based system that performs the specified function or operation, or using a combination of dedicated hardware and computer instructions.

[0111] It should be noted that although several modules or units for the device used to perform actions have been mentioned in the detailed description above, this division is not mandatory. In fact, according to the embodiments of this application, the features and functions of two or more modules or units described above can be embodied in one module or unit. Conversely, the features and functions of one module or unit described above can be further divided and embodied by multiple modules or units.

[0112] Those skilled in the art should understand that the discussion of any of the above embodiments is merely exemplary and is not intended to imply that the scope of this application (including the claims) is limited to these examples; within the framework of this application, the technical features of the above embodiments or different embodiments can also be combined, the steps can be implemented in any order, and there are many other variations of different aspects of the embodiments of this application as described above, which are not provided in the details for the sake of brevity.

[0113] Additionally, to simplify the description and discussion, and to avoid obscuring the embodiments of this application, the well-known power / ground connections to integrated circuit (IC) chips and other components may or may not be shown in the provided drawings. Furthermore, the apparatus may be shown in block diagram form to avoid obscuring the embodiments of this application, and this also takes into account the fact that the details of the implementation of these block diagram apparatuses are highly dependent on the platform on which the embodiments of this application will be implemented (i.e., these details should be fully understood by those skilled in the art). While specific details (e.g., circuits) have been set forth to describe exemplary embodiments of this application, it will be apparent to those skilled in the art that the embodiments of this application can be implemented without these specific details or with variations thereof. Therefore, these descriptions should be considered illustrative rather than restrictive.

[0114] Although this application has been described in conjunction with specific embodiments thereof, many substitutions, modifications, and variations of these embodiments will be apparent to those skilled in the art from the foregoing description. For example, other memory architectures (e.g., dynamic RAM (DRAM)) may be used with the embodiments discussed.

[0115] The embodiments of this application are intended to cover all such substitutions, modifications, and variations that fall within the broad scope of the appended claims. Therefore, any omissions, modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the embodiments of this application should be included within the protection scope of this application.

[0116] While the spirit and principles of this disclosure have been described with reference to several specific embodiments, it should be understood that this disclosure is not limited to the disclosed specific embodiments, and the division of aspects does not imply that features in these aspects cannot be combined for benefit; such division is merely for convenience of expression. This disclosure is intended to cover various modifications and equivalent arrangements included within the spirit and scope of the appended claims. The scope of the appended claims is to be interpreted in the broadest sense, thereby encompassing all such modifications and equivalent structures and functions.

Claims

1. A method for suppressing multiples of submarine seismic waves, comprising: Obtain the submarine seismic data to be processed; The submarine earthquake data is used to perform a prediction calculation to obtain the first prediction data; The first prediction data is input into the initial Fourier neural operator network to obtain the second prediction data; The submarine earthquake data and the second predicted data are input into the joint loss function to obtain the score; In response to the score being not less than a preset threshold, the parameters of the initial Fourier neural operator network are adjusted to obtain an intermediate Fourier neural operator network. The intermediate Fourier neural operator network is used to replace the initial Fourier neural operator network for iterative training until the score is less than the preset threshold, and then the second prediction data is output. as well as The first wave of the submarine earthquake is calculated based on the second predicted data and the submarine earthquake data.

2. The method of claim 1, wherein, The joint loss function includes a first back-inference function and a first scoring function; The submarine seismic data and the second predicted data are input into the joint loss function to obtain the score, including: The submarine earthquake data and the second predicted data are input into the first back-inference function to obtain the error data of the submarine earthquake data and the synthetic earthquake data; as well as The reference wavefield based on the submarine seismic data and the error data of the synthetic seismic data are input into the first scoring function to obtain the score.

3. The method according to claim 2, wherein, The error data of the submarine earthquake data and the second predicted data are input into the first back-calculation function to obtain the error data of the submarine earthquake data and the synthetic earthquake data, including: Subtract the second predicted data from the submarine earthquake data to obtain the estimated first wave data; The estimated primary wave data and the preset wavelet data are deconvolved to obtain the reflection coefficient data; The reflection coefficient data is convolved with the preset wavelet data to obtain synthetic seismic wave data; and The error data is calculated based on the reference wavefield and the synthetic seismic wave data.

4. The method according to claim 3, wherein, The first scoring function includes an error loss term and a physical constraint term; the error loss term is used to obtain a first score by describing the error data of the submarine seismic data and the synthetic seismic wave; The physical constraints include sparsity constraints, total variation constraints, and wave equation constraints. The sparsity constraint term is used to obtain a second score by describing the sparsity of at least one of the estimated first-wave data, the reflection coefficient data, or an intermediate physical quantity derived from the estimated first-wave data; the total variation constraint term is used to obtain a third score by describing the continuity and smoothness of at least one of the estimated first-wave data, the reflection coefficient data, or an intermediate physical quantity derived from the estimated first-wave data; and the wave equation constraint term is used to obtain a fourth score by describing the physical realizability of at least one of the estimated first-wave data, the reflection coefficient data, or an intermediate physical quantity derived from the estimated first-wave data. The rating includes the first rating, the second rating, the third rating, and the fourth rating.

5. The method according to claim 4, wherein, The reference wavefield is calculated based on the first predicted data and the submarine earthquake data; the reference wavefield is used as a reference target for calculating the error of the synthetic seismic wave data by the first scoring function when iterating the Fourier neural operator network.

6. The method according to claim 1, wherein, Before performing prediction calculations on the submarine earthquake data to obtain the first prediction data, a filtering step is included, which is used to remove random noise from the submarine earthquake data.

7. A device for suppressing multiple waves of submarine seismic data, comprising: The data acquisition module is configured to acquire submarine seismic data to be processed; The estimation module is configured to perform estimation calculations on the submarine earthquake data to obtain first prediction data; The neural operator network operation module is configured to input the first predicted data into an initial Fourier neural operator network to obtain second predicted data, input the submarine earthquake data and the second predicted data into the joint loss function to obtain the score, adjust the parameters of the initial Fourier neural operator network to obtain an intermediate Fourier neural operator network in response to the score being not less than a preset threshold, use the intermediate Fourier neural operator network to replace the initial Fourier neural operator network for iterative training until the score is less than the preset threshold, and output the second predicted data. as well as The calculation module is configured to calculate the first wave of a submarine earthquake based on the second predicted data and the submarine earthquake data.

8. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor, when executing the program, implements the method as claimed in any one of claims 1 to 6.

9. A non-transitory computer-readable storage medium storing computer instructions for causing a computer to perform the method of any one of claims 1 to 6.

10. A computer program product comprising computer program instructions that, when executed on a computer, cause the computer to perform the method as described in any one of claims 1 to 6.

Images