Seismic data processing method, device, electronic device and storage medium

Abstract

The present invention discloses a method, device, electronic device, and storage medium for processing seismic data, wherein the method comprises: acquiring seismic data to be processed; processing the seismic data to be processed based on a pre-trained target seismic data processing model, and determining a recognition rate and seismic data to be applied corresponding to the seismic data to be processed; wherein the target seismic data processing model is a pre-trained neural network model; and determining target seismic data based on the recognition rate, the seismic data to be applied, and the seismic data to be processed. Based on the above technical solution, the collected seismic data is processed by the target seismic data processing model to obtain a corresponding recognition rate and seismic data to be applied, thereby determining the target seismic data, thereby improving the processing efficiency of seismic data.

Description

Technical Field

[0001] The present invention relates to the field of data processing technology, and in particular to a method, device, electronic equipment and storage medium for processing seismic data. Background Art

[0002] During the data acquisition process of seismic exploration, the seismic signal is inevitably interfered by noise due to multiple factors such as the acquisition instrument, ground environment, and surface structure. Therefore, it is necessary to perform noise reduction processing on the acquired seismic signal according to the type of noise.

[0003] However, in actual seismic data acquisition and processing, the types of noise are complex and diverse, and some noise is not common typical noise. Conventional noise processing methods require the design of appropriate noise reduction algorithms for different noise types, which reduces the efficiency of signal processing. Summary of the Invention

[0004] The present invention provides a method, device, electronic device and storage medium for processing seismic data to solve the problem that existing seismic data processing methods require corresponding noise reduction algorithms for different types of seismic data, thereby resulting in low data processing efficiency.

[0005] According to one aspect of the present invention, a method for processing seismic data is provided, the method comprising:

[0006] Obtaining seismic data to be processed;

[0007] Processing the seismic data to be processed based on a pre-trained target seismic data processing model to determine a recognition rate and seismic data to be applied corresponding to the seismic data to be processed; wherein the target seismic data processing model is a pre-trained neural network model;

[0008] Target seismic data is determined based on the recognition rate, the seismic data to be applied, and the seismic data to be processed.

[0009] According to another aspect of the present invention, there is provided a device for processing seismic data, the device comprising:

[0010] A data acquisition module, used for acquiring seismic data to be processed;

[0011] a data processing module, configured to process the seismic data to be processed based on a pre-trained target seismic data processing model, and determine a recognition rate and seismic data to be applied corresponding to the seismic data to be processed; wherein the target seismic data processing model is a pre-trained neural network model;

[0012] A target data determination module is used to determine target seismic data based on the recognition rate, the seismic data to be applied and the seismic data to be processed.

[0013] According to another aspect of the present invention, an electronic device is provided, comprising:

[0014] at least one processor; and

[0015] a memory communicatively connected to the at least one processor; wherein,

[0016] The memory stores a computer program executable by the at least one processor. The computer program is executed by the at least one processor to enable the at least one processor to perform the seismic data processing method described in any embodiment of the present invention.

[0017] According to another aspect of the present invention, a computer-readable storage medium is provided, wherein the computer-readable storage medium stores computer instructions, and the computer instructions are used to enable a processor to implement the seismic data processing method according to any embodiment of the present invention when executed.

[0018] The technical solution of an embodiment of the present invention obtains seismic data to be processed and processes the seismic data to be processed based on a pre-trained target seismic data processing model to determine a recognition rate and seismic data to be applied corresponding to the seismic data to be processed, and then determines target seismic data based on the recognition rate, the seismic data to be applied, and the seismic data to be processed. Based on the above technical solution, the acquired seismic data is processed using the target seismic data processing model to obtain a corresponding recognition rate and seismic data to be applied, thereby determining the target seismic data, thereby improving the processing efficiency of seismic data.

[0019] It should be understood that the content described in this section is not intended to identify the key or important features of the embodiments of the present invention, nor is it intended to limit the scope of the present invention. Other features of the present invention will become readily understood through the following description. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0021] Figure 1 is a flow chart of a method for processing seismic data provided by an embodiment of the present invention;

[0022] Figure 2is a schematic structural diagram of a target seismic data processing model provided by an embodiment of the present invention;

[0023] Figure 3 is a flow chart of a method for processing seismic data provided by an embodiment of the present invention;

[0024] Figure 4 This is a structural block diagram of a seismic data processing device provided by an embodiment of the present invention;

[0025] Figure 5 It is a structural diagram of an electronic device provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0026] In order to enable those skilled in the art to better understand the solutions of the present invention, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the embodiments described are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts should fall within the scope of protection of the present invention.

[0027] It should be noted that the terms "first", "second", etc. in the description and claims of the present invention and the above-mentioned drawings are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that the numbers used in this way can be interchanged where appropriate, so that the embodiments of the present invention described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions. For example, a process, method, system, product or device that includes a series of steps or units is not necessarily limited to those steps or units clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0028] Example 1

[0029] Figure 1 This is a flow chart of a method for processing seismic data provided by an embodiment of the present invention. This embodiment is applicable to the case where the collected seismic data is processed based on a target seismic data processing model to obtain target seismic data. The method can be executed by a seismic data processing device, which can be implemented in the form of hardware and / or software. The seismic data processing device can be configured in an electronic device, which can be a server, a terminal device, etc.

[0030] like Figure 1 As shown, the method includes:

[0031] S110: Obtain seismic data to be processed.

[0032] The seismic data to be processed may be collected seismic data, for example, seismic waveform data.

[0033] Specifically, the seismic data to be processed can be obtained from the database. For example, seismic data associated with the target area can be obtained and used as the seismic data to be processed. It can be understood that in order to achieve rapid indexing of seismic data, a relationship mapping table between seismic data and each exploration area can be established after the seismic data acquisition is completed. Then, when the seismic data to be processed is obtained, the seismic data corresponding to the target area can be determined based on the relationship mapping table and used as the seismic data to be processed.

[0034] On the basis of the above technical solution, before obtaining the seismic data to be processed, it also includes: obtaining original data, and processing the original data to determine a sample data set; training the seismic data processing model to be trained based on the sample data set to obtain the target seismic data processing model.

[0035] The raw data may be unprocessed seismic data, i.e., raw seismic data. The sample dataset may be understood as a dataset used for model training. The seismic data processing model to be trained may be a pre-set, untrained model, such as a multi-task model. Accordingly, the target seismic data processing model may be a model obtained by training the seismic data processing model to be trained.

[0036] Specifically, before processing the seismic data to be processed, it is necessary to obtain the original data, process the original data to determine the sample data set, and then train the seismic data processing model to be trained based on the sample data set to obtain the target seismic data processing model.

[0037] Based on the above technical solution, the processing of the original data to determine the sample data set includes: determining the identification label corresponding to each of the original data, and determining at least one random square wave noise based on preset noise characteristic data; determining the sample data set based on the identification label, the random square wave noise and the original data.

[0038] The identification tag can be understood as the data type of the raw data, such as normal seismic data and noisy seismic data. The preset noise characteristic data can be a pre-set noise construction data interval, such as the peak value interval, the trough value interval, and the starting point value interval. Random square wave noise can be understood as square wave noise constructed based on the preset noise characteristic data.

[0039] Specifically, an identification label corresponding to each of the original data is determined, and at least one random square wave noise is determined based on preset noise characteristic data. The sample data set is determined based on the identification label, the random square wave noise and the original data. For example, based on given seismic data, the seismic trace without square wave noise is marked as 0, and the seismic trace containing square wave noise is marked as 1; the label data corresponding to the seismic trace marked as 0 is set to [1,0], and the label data corresponding to the seismic trace marked as 1 is set to [0,1]. When constructing random square wave noise based on the preset noise characteristic data, it should be noted that the square wave noise contains three elements: peak amplitude v1, trough amplitude v2 and starting point t0; and then according to the given range [v 11 , v 12 ]、[v 21 ,v 22 ] and [t 01 ,t 02 ], random settings: amplitude v1, amplitude v2, starting point t0, after determining the characteristic data of the noise, based on the given seismic trace length n, construct a square waveform noise such as:

[0040] On the basis of the above technical solution, the sample data set is determined based on the identification label, the random square wave noise and the original data, including: obtaining the identification label of the original data, and taking the original data with the identification label as normal data as positive sample data; superimposing the random square wave noise with the positive sample data to determine the negative sample data, and modifying the identification label of the negative sample data to noise data.

[0041] The identification tag may be a tag used to determine whether the current sample data is normal data. Positive sample data may be understood as seismic data that does not contain square wave noise. It should be noted that, in order to determine whether noise suppression is required for the sample data, the sample data includes an identification tag and a suppression tag. The suppression tag may be a tag used to determine whether suppression is required for the current sample data. It is possible that the suppression tag for the positive sample data is the original data, and the suppression tag for the negative sample data is the random square wave noise.

[0042] Specifically, the identification label of the original data is obtained, the original data with the identification label as normal data is used as positive sample data, and the random square wave noise is superimposed on the positive sample data to determine the negative sample data, and the identification label of the negative sample data is modified to the noise data. It should be noted that a sample data contains earthquake data x, identification label y1 and suppression label data y2, that is, a set of sample data is {x i ,y 1i ,y 2i}, and the data samples are divided into positive and negative samples; the ratio of the number of positive and negative samples is (including but not limited to) 1:1; the positive sample is constructed as follows: the input data x is a normal seismic trace, the identification label data y1 is [1,0], and the suppressed label data is the same as the input data, that is, y2=x; the negative sample is constructed as follows: the input data x is a normal seismic trace d randomly selected and superimposed with random square waveform noise n, that is, x=d+n; the identification label data y1 is [0,1]; the suppressed label data is the above-mentioned square waveform noise n, that is, y2=n; and then the obtained positive and negative samples are divided to obtain a validation set and a training set, and the ratio of the training set to the validation set can be 8:2.

[0043] On the basis of the above technical solution, the seismic data processing model to be trained is trained based on the sample data set to obtain the target seismic processing model, including: obtaining the current number of iterations of the multi-task loss function; if the current number of iterations does not meet the preset iteration threshold, adjusting the network parameters of the seismic data processing model to be trained through the back propagation algorithm, and continuing the iterative processing based on the multi-task loss function; if the current number of iterations meets the preset iteration threshold, the seismic data processing model to be trained corresponding to the current number of iterations is used as the target seismic processing model.

[0044] The multi-task loss function may be a pre-built loss function corresponding to the seismic data processing model. The current number of iterations may be understood as the number of calculations of the current loss function. The preset iteration threshold may be the maximum number of iterations of the loss function.

[0045] Specifically, the current number of iterations of the multi-task loss function is obtained, and whether the training is completed is determined based on the current number of iterations. If the current number of iterations does not meet the preset iteration threshold, the network parameters of the seismic data processing model to be trained are adjusted through the back propagation algorithm, and the iterative processing is continued based on the multi-task loss function. If the current number of iterations meets the preset iteration threshold, the seismic data processing model to be trained corresponding to the current number of iterations is used as the target seismic processing model. The obtained sample data set can be input into the seismic data processing model to be trained, and the output of the model is obtained. The seismic data processing model to be trained is trained based on the output value and the multi-task loss function, and the target seismic data processing model is obtained. Multi-task loss function Among them, the suppression loss function is L1(Θ)=||yf Θ (x)|| 2 , the recognition loss function is L2(Θ)=-logsoftmax(f Θ (x)), x is the input data, y is the label data, Θ represents the network parameters that the network needs to train and learn; σ1, σ2 are adjustment coefficients that need to be trained.

[0046] S120 , processing the seismic data to be processed based on the pre-trained target seismic data processing model, and determining a recognition rate and seismic data to be applied corresponding to the seismic data to be processed.

[0047] The target seismic data processing model is a pre-trained neural network model. The recognition rate can be understood as the probability that the current seismic data is normal seismic data. The seismic data to be applied can be the output data obtained after suppressing the seismic data.

[0048] Specifically, the seismic data to be processed is processed based on the pre-trained target seismic data processing model, and the recognition rate and the seismic data to be applied corresponding to the seismic data to be processed are determined.

[0049] Based on the above technical solution, the seismic data processing model includes a shared network, a first task network and a second task network; the seismic data processing model obtained based on pre-training processes the seismic data to be processed and determines the recognition rate and seismic data to be applied corresponding to the seismic data to be processed, including: processing the seismic data to be processed based on the shared network to obtain data features corresponding to the seismic data to be processed; determining the recognition rate corresponding to the seismic data to be processed based on the first task network and the data features; and determining the seismic data to be applied corresponding to the seismic data to be processed based on the second task network and the data features.

[0050] The shared network can be a network for feature extraction from seismic data. The first task network can be a network for class recognition based on the extracted data features. The second task network can be a network for noise suppression in seismic data. Data features can be understood as characteristic information corresponding to seismic data.

[0051] Specifically, the seismic data processing model includes a shared network, a first task network and a second task network, and then the seismic data to be processed is processed based on the shared network to obtain data features corresponding to the seismic data to be processed, and the recognition rate corresponding to the seismic data to be processed is determined based on the first task network and the data features, and the seismic data to be applied corresponding to the seismic data to be processed is determined based on the second task network and the data features.

[0052] It should be noted that the technical solution provided by the embodiment of the present invention can be that after obtaining the characteristic data corresponding to the current seismic data through a shared network, the first task network and the second task network simultaneously process the characteristic data to obtain the recognition rate corresponding to the current seismic data and the seismic data to be applied. It can also be that the first task network first determines the recognition rate corresponding to the current seismic data, and then determines whether the second task network needs to suppress noise on the current seismic data based on the recognition rate to obtain the seismic data to be applied.

[0053] Based on the above technical solution, the shared network includes at least four groups of first convolutional layers; two adjacent first convolutional layers in the shared network are connected by downsampling; the first task network includes at least three groups of second convolutional layers and at least two groups of fully connected layers; the second task network includes at least four groups of first convolutional layers.

[0054] The first convolutional layer includes a ReLU activation function. The fully connected layer is located after the second convolutional layer. Two adjacent first convolutional layers in the second task network are connected via upsampling. The first convolutional layers of the same scale in the first task network and the shared network are connected via a skip connection.

[0055] Specifically, such as Figure 2 As shown, the target seismic data processing model in the embodiment of the present invention consists of three parts, namely a shared network, a task 1 network and a task 2 network. The shared network consists of 4 groups of convolutional layers with 64 channels and a filter size of 3. Each group of convolutional layers contains a ReLU activation function, and each group of convolutional layers is connected by downsampling; the task 1 network consists of 3 groups of convolutional layers with 64 channels and a filter size of 3 and two groups of fully connected layers. The input channels of the fully connected layers are 32 and 2 respectively; the task 2 network consists of 4 groups of convolutional layers with 64 channels and a filter size of 3. Each group of convolutional layers contains a ReLU activation function, and each group of convolutional layers is connected by upsampling; in addition, in order to avoid feature loss, skip connections are added between feature maps of the same scale in the shared network and the task 2 network.

[0056] S130 . Determine target seismic data based on the recognition rate, the seismic data to be applied, and the seismic data to be processed.

[0057] The target seismic data can be understood as the output data obtained after processing the seismic data based on the model.

[0058] Specifically, the target seismic data is determined based on the recognition rate, the seismic data to be applied and the seismic data to be processed. For example, whether the seismic data to be applied needs to be output can be determined based on the recognition rate. If the recognition rate meets the preset conditions, the original seismic data, that is, the seismic data to be processed, is output. If the recognition rate does not meet the preset conditions, the seismic data to be applied is output.

[0059] On the basis of the above technical solution, the target seismic data is determined based on the recognition rate, the seismic data to be applied and the seismic data to be processed, including: if the recognition rate is less than the preset recognition threshold, the seismic data to be applied is used as the target seismic data; if the recognition rate is greater than the preset recognition threshold, the seismic data to be processed is used as the target seismic data.

[0060] The preset recognition threshold may be a preset recognition rate value, for example, 95%.

[0061] Specifically, if the recognition rate is less than a preset recognition threshold, the seismic data to be applied is used as the target seismic data; if the recognition rate is greater than the preset recognition threshold, the seismic data to be processed is used as the target seismic data. It is understood that if the recognition rate is greater than the preset recognition threshold, it means that the current seismic data is likely to be normal data, and there is no need to suppress noise on the current seismic data. The seismic data to be processed can be output as the target seismic data. If the recognition rate is less than the preset recognition threshold, the suppressed seismic data to be applied needs to be output as the target seismic data.

[0062] The technical solution of an embodiment of the present invention obtains seismic data to be processed and processes the seismic data to be processed based on a pre-trained target seismic data processing model to determine a recognition rate and seismic data to be applied corresponding to the seismic data to be processed, and then determines target seismic data based on the recognition rate, the seismic data to be applied, and the seismic data to be processed. Based on the above technical solution, the acquired seismic data is processed using the target seismic data processing model to obtain a corresponding recognition rate and seismic data to be applied, thereby determining the target seismic data, thereby improving the processing efficiency of seismic data.

[0063] Example 2

[0064] Figure 3 This is a flowchart of a method for processing seismic data provided by an embodiment of the present invention. This embodiment further optimizes the aforementioned method based on the aforementioned embodiment. For detailed implementation details, please refer to the technical solution of this embodiment. Technical terms that are identical or corresponding to those in the aforementioned embodiment are not further described here.

[0065] like Figure 3 The method of the embodiment of the present invention includes:

[0066] Construct a sample data set: Specifically, construct a square waveform noise identification label: for example, based on a given seismic trace, mark the seismic trace without square waveform noise as 0, and mark the seismic trace with square waveform noise as 1; the label data corresponding to the seismic trace marked as 0 is set to [1,0], and the label data corresponding to the seismic trace marked as 1 is set to [0,1].

[0067] Constructing the sample data set also includes constructing the noise elimination label: square waveform noise contains three elements: amplitude v1, amplitude v2, and starting point t0. The specific implementation steps are as follows: ① According to the given range [v 11 , v 12 ]、[v 21 ,v 22 ] and [t 01 ,t 02 ], random settings: amplitude v1, amplitude v2, starting point t0; ② According to the given seismic trace length n, construct square waveform noise:

[0068] Then, a sample data set is constructed based on the noise identification label and the noise elimination label: It should be noted that a sample data contains earthquake data x, identification label y1 and suppression label data y2, that is, a set of sample data is {x i ,y 1i ,y 2i}, and the data samples are divided into positive and negative samples; the ratio of the number of positive and negative samples is (including but not limited to) 1:1; the positive sample is constructed as follows: the input data x is a normal seismic trace, the identification label data y1 is [1,0], and the suppressed label data is the same as the input data, that is, y2=x; the negative sample is constructed as follows: the input data x is a normal seismic trace d randomly selected and superimposed with random square waveform noise n, that is, x=d+n; the identification label data y1 is [0,1]; the suppressed label data is the above-mentioned square waveform noise n, that is, y2=n; and then the obtained positive and negative samples are divided to obtain a validation set and a training set, and the ratio of the training set to the validation set can be 8:2.

[0069] Determine the target seismic data processing model: Specifically, the target seismic data processing model provided by the embodiment of the present invention consists of three parts, namely a shared network, a task 1 network, and a task 2 network. The shared network consists of 4 groups of convolutional layers with 64 channels and a filter size of 3. Each group of convolutional layers contains a ReLU activation function, and each group of convolutional layers is connected by downsampling; the task 1 network consists of 3 groups of convolutional layers with 64 channels and a filter size of 3 and two groups of fully connected layers. The input channels of the fully connected layers are 32 and 2 respectively; the task 2 network consists of 4 groups of convolutional layers with 64 channels and a filter size of 3. Each group of convolutional layers contains a ReLU activation function, and each group of convolutional layers is connected by upsampling; in addition, in order to avoid feature loss, skip connections are added between feature maps of the same scale in the shared network and the task 2 network.

[0070] It should be noted that the construction method of the seismic data processing model to be trained is the same as that of the target seismic data processing model, and then the seismic data processing model to be trained is trained based on the sample data. The seismic data processing model to be trained can be randomly initialized to obtain an initial model, and then the sample data is input into the initial model to obtain the data of the output layer, calculate the loss function, and judge whether the current number of iterations meets the maximum number of training times (the default value is 50); if not, the network parameters of the current network are adjusted by the back propagation algorithm until the maximum number of training times is reached, and the network model that reaches the training times is determined to be the same as the target seismic data processing model. The back propagation algorithm is the process of obtaining the optimal network parameters by minimizing the loss function. The minimization process of the objective function can be achieved through the Adam optimization algorithm. The loss function is:

[0071] Among them, the suppression loss function is L1(Θ)=||yf Θ (x)|| 2 , the recognition loss function is L2(Θ)=-logsoftmax(f Θ (x)), x is the input data, y is the label data, Θ represents the network parameters that the network needs to train and learn; σ1, σ2 are adjustment coefficients that need to be trained.

[0072] Seismic data processing: Specifically, the seismic data is input into the target seismic data processing model, and Task 1 identifies the seismic data. If the probability that it is a normal seismic trace is greater than a given value (the default value is 0.95), the output value of this trace is the original input data; otherwise, the seismic data after suppressing the square wave noise obtained by Task 2 is output.

[0073] The technical solution of an embodiment of the present invention obtains seismic data to be processed and processes the seismic data to be processed based on a pre-trained target seismic data processing model to determine a recognition rate and seismic data to be applied corresponding to the seismic data to be processed, and then determines target seismic data based on the recognition rate, the seismic data to be applied, and the seismic data to be processed. Based on the above technical solution, the acquired seismic data is processed using the target seismic data processing model to obtain a corresponding recognition rate and seismic data to be applied, thereby determining the target seismic data, thereby improving the processing efficiency of seismic data.

[0074] Example 3

[0075] Figure 4 This is a structural block diagram of a seismic data processing device provided by an embodiment of the present invention. Figure 4 As shown, the device includes: a data acquisition module 410, a data processing module 420 and a target data determination module 430.

[0076] The data acquisition module 410 is used to acquire seismic data to be processed;

[0077] a data processing module 420 configured to process the seismic data to be processed based on a pre-trained target seismic data processing model, and determine a recognition rate and seismic data to be applied corresponding to the seismic data to be processed; wherein the target seismic data processing model is a pre-trained neural network model;

[0078] The target data determination module 430 is configured to determine target seismic data based on the recognition rate, the seismic data to be applied, and the seismic data to be processed.

[0079] Based on the above technical solution, the seismic data processing model includes a shared network, a first task network and a second task network; the data processing module is used to process the seismic data to be processed based on the shared network to obtain data features corresponding to the seismic data to be processed; determine the recognition rate corresponding to the seismic data to be processed based on the first task network and the data features; and determine the seismic data to be applied corresponding to the seismic data to be processed based on the second task network and the data features.

[0080] Based on the above technical solution, the shared network includes at least four groups of first convolutional layers; wherein the first convolutional layer includes a ReLU activation function; two adjacent first convolutional layers in the shared network are connected by downsampling; the first task network includes at least three groups of second convolutional layers and at least two groups of fully connected layers; wherein the fully connected layer is located after the second convolutional layer; the second task network includes at least four groups of first convolutional layers; wherein two adjacent first convolutional layers in the second task network are connected by upsampling; the first convolutional layers of the same scale in the first task network and the shared network are connected by skip connections.

[0081] Based on the above technical solution, the device also includes: a target model determination module, which is used to obtain original data and process the original data to determine a sample data set; train the seismic data processing model to be trained based on the sample data set to obtain the target seismic processing model.

[0082] Based on the above technical solution, the target model determination module is used to determine the identification label corresponding to each of the original data, and determine at least one random square wave noise based on preset noise characteristic data; and determine the sample data set based on the identification label, the random square wave noise and the original data.

[0083] On the basis of the above technical solution, the target model determination module is used to obtain the identification label of the original data, and take the original data whose identification label is normal data as positive sample data; wherein, the sample data includes an identification label and a suppression label; the suppression label of the positive sample data is the original data; the random square wave noise is superimposed on the positive sample data to determine the negative sample data, and the identification label of the negative sample data is modified to noise data; wherein, the suppression label of the negative sample data is the random square wave noise.

[0084] Based on the above technical solution, the target model determination module is used to obtain the current number of iterations of the multi-task loss function; if the current number of iterations does not meet the preset iteration threshold, the network parameters of the seismic data processing model to be trained are adjusted through the back propagation algorithm, and the iterative processing is continued based on the multi-task loss function; if the current number of iterations meets the preset iteration threshold, the seismic data processing model to be trained corresponding to the current number of iterations is used as the target seismic processing model.

[0085] The technical solution of an embodiment of the present invention obtains seismic data to be processed and processes the seismic data to be processed based on a pre-trained target seismic data processing model to determine a recognition rate and seismic data to be applied corresponding to the seismic data to be processed, and then determines target seismic data based on the recognition rate, the seismic data to be applied, and the seismic data to be processed. Based on the above technical solution, the acquired seismic data is processed using the target seismic data processing model to obtain a corresponding recognition rate and seismic data to be applied, thereby determining the target seismic data, thereby improving the processing efficiency of seismic data.

[0086] The seismic data processing device provided in the embodiment of the present invention can execute the seismic data processing method provided in any embodiment of the present invention, and has the corresponding functional modules and beneficial effects of the execution method.

[0087] Example 4

[0088] Figure 5 A schematic diagram of the structure of an electronic device 10 that can be used to implement an embodiment of the present invention is shown. The electronic device is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device can also represent various forms of mobile devices, such as personal digital processing, cellular phones, smart phones, wearable devices (such as helmets, glasses, watches, etc.) and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely examples and are not intended to limit the implementation of the present invention described and / or claimed herein.

[0089] like Figure 5 As shown, the electronic device 10 includes at least one processor 11, and a memory connected to the at least one processor 11, such as a read-only memory (ROM) 12, a random access memory (RAM) 13, etc., wherein the memory stores a computer program that can be executed by the at least one processor, and the processor 11 can perform various appropriate actions and processes according to the computer program stored in the read-only memory (ROM) 12 or the computer program loaded from the storage unit 18 to the random access memory (RAM) 13. Various programs and data required for the operation of the electronic device 10 can also be stored in the RAM 13. The processor 11, ROM 12 and RAM 13 are connected to each other via a bus 14. An input / output (I / O) interface 15 is also connected to the bus 14.

[0090] Multiple components in the electronic device 10 are connected to the I / O interface 15, including an input unit 16, such as a keyboard, a mouse, etc.; an output unit 17, such as various types of displays, speakers, etc.; a storage unit 18, such as a magnetic disk, an optical disk, etc.; and a communication unit 19, such as a network card, a modem, a wireless communication transceiver, etc. The communication unit 19 allows the electronic device 10 to exchange information / data with other devices via a computer network such as the Internet and / or various telecommunication networks.

[0091] The processor 11 can be any general-purpose and / or specialized processing component with processing and computing capabilities. Some examples of the processor 11 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various specialized artificial intelligence (AI) computing chips, various processors running machine learning model algorithms, a digital signal processor (DSP), and any suitable processor, controller, microcontroller, etc. The processor 11 executes the various methods and processes described above, such as the method for processing seismic data.

[0092] In some embodiments, the method for processing seismic data can be implemented as a computer program that is tangibly contained in a computer-readable storage medium, such as the storage unit 18. In some embodiments, part or all of the computer program can be loaded and / or installed on the electronic device 10 via the ROM 12 and / or the communication unit 19. When the computer program is loaded into the RAM 13 and executed by the processor 11, one or more steps of the method for processing seismic data described above can be performed. Alternatively, in other embodiments, the processor 11 can be configured to perform the method for processing seismic data in any other appropriate manner (e.g., by means of firmware).

[0093] Various embodiments of the systems and techniques described herein can be implemented in digital electronic circuit systems, integrated circuit systems, field programmable gate arrays (FPGAs), application specific integrated circuits (ASICs), application specific standard products (ASSPs), system-on-chip systems (SOCs), programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various embodiments can include being implemented in one or more computer programs that are executable and / or interpreted on a programmable system that includes at least one programmable processor, which can be a special purpose or general purpose programmable processor that can receive data and instructions from a storage system, at least one input device, and at least one output device, and transmit data and instructions to the storage system, the at least one input device, and the at least one output device.

[0094] Computer programs for implementing the methods of the present invention may be written in any combination of one or more programming languages. These computer programs may be provided to a processor of a general-purpose computer, a special-purpose computer, or other programmable data processing device, such that when the computer program is executed by the processor, the functions / operations specified in the flowcharts and / or block diagrams are implemented. The computer program may be executed entirely on the machine, partially on the machine, as a stand-alone software package, partially on the machine and partially on a remote machine, or entirely on a remote machine or server.

[0095] In the context of the present invention, computer-readable storage media can be tangible media that can contain or store a computer program for use with an instruction execution system, device or equipment or used in combination with an instruction execution system, device or equipment. Computer-readable storage media can include but are not limited to electronic, magnetic, optical, electromagnetic, infrared or semiconductor systems, devices or equipment, or any suitable combination of the foregoing. Alternatively, computer-readable storage media can be machine-readable signal media. More specific examples of machine-readable storage media can include electrical connections based on one or more lines, portable computer disks, hard disks, random access memories (RAM), read-only memories (ROM), erasable programmable read-only memories (EPROM or flash memory), optical fibers, portable compact disk read-only memories (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination of the foregoing.

[0096] To provide interaction with a user, the systems and techniques described herein can be implemented on an electronic device having: a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to the user; and a keyboard and pointing device (e.g., a mouse or trackball) through which the user can provide input to the electronic device. Other types of devices can also be used to provide interaction with the user; for example, the feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form (including acoustic input, voice input, or tactile input).

[0097] The systems and techniques described herein can be implemented in a computing system that includes back-end components (e.g., as a data server), or a computing system that includes middleware components (e.g., an application server), or a computing system that includes front-end components (e.g., a user computer with a graphical user interface or web browser through which a user can interact with implementations of the systems and techniques described herein), or a computing system that includes any combination of such back-end components, middleware components, or front-end components. The components of the system can be interconnected by any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include: a local area network (LAN), a wide area network (WAN), a blockchain network, and the Internet.

[0098] A computing system may include clients and servers. The clients and servers are typically remote from each other and typically interact via a communication network. This client-server relationship arises through computer programs running on the respective computers, creating a client-server relationship. The server may be a cloud server, also known as a cloud computing server or cloud host. This server is a hosting product within the cloud computing service ecosystem that addresses the management difficulties and limited scalability of traditional physical hosting and VPS services.

[0099] It should be understood that the various forms of the processes shown above can be used to reorder, add, or delete steps. For example, the steps described in the present invention can be performed in parallel, sequentially, or in a different order, as long as the desired results of the technical solution of the present invention can be achieved. This is not limited herein.

[0100] The above specific embodiments do not limit the scope of protection of the present invention. Those skilled in the art will appreciate that various modifications, combinations, sub-combinations, and substitutions may be made based on design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention are intended to be included within the scope of protection of the present invention.

Claims

1. A method for processing seismic data, characterized in that: include: Obtaining seismic data to be processed; Processing the seismic data to be processed based on a pre-trained target seismic data processing model to determine a recognition rate and seismic data to be applied corresponding to the seismic data to be processed; wherein the target seismic data processing model is a pre-trained neural network model; determining target seismic data based on the recognition rate, the seismic data to be applied, and the seismic data to be processed; The target seismic data processing model includes a shared network, a first task network and a second task network; The process of processing the seismic data to be processed based on the pre-trained target seismic data processing model and determining the recognition rate and seismic data to be applied corresponding to the seismic data to be processed includes: Processing the seismic data to be processed based on the shared network to obtain data features corresponding to the seismic data to be processed; determining a recognition rate corresponding to the seismic data to be processed based on the first task network and the data features; determining, based on the second task network and the data features, seismic data to be applied corresponding to the seismic data to be processed; The shared network includes at least four groups of first convolutional layers; wherein the first convolutional layers include a ReLU activation function; and two adjacent first convolutional layers in the shared network are connected by downsampling; The first task network includes at least three groups of second convolutional layers and at least two groups of fully connected layers; wherein the fully connected layers are located after the second convolutional layers; The second task network includes at least four groups of first convolutional layers; wherein, two adjacent first convolutional layers in the second task network are connected by upsampling; and the second task network and the first convolutional layers of the same scale in the shared network are connected by a skip connection; Before obtaining the seismic data to be processed, the method further includes: Acquiring original data, and processing the original data to determine a sample data set; Training the seismic data processing model to be trained based on the sample data set to obtain the target seismic processing model; Processing the original data to determine a sample data set includes: Determining an identification tag corresponding to each of the raw data, and determining at least one random square wave noise based on preset noise characteristic data; determining the sample data set based on the identification tag, the random square wave noise, and the original data; The determining the sample data set based on the identification tag, the random square wave noise, and the original data includes: Obtaining an identification label of the original data, and taking the original data whose identification label is normal data as positive sample data; wherein the sample data includes an identification label and a suppression label; the suppression label of the positive sample data is the original data; The random square wave noise is superimposed on the positive sample data to determine negative sample data, and the identification label of the negative sample data is modified to noise data; wherein the suppressed label of the negative sample data is the random square wave noise.

2. The method according to claim 1, characterized in that The training of the seismic data processing model to be trained based on the sample data set to obtain the target seismic data processing model includes: Get the current iteration number of the multi-task loss function; If the current number of iterations does not meet a preset iteration threshold, adjusting the network parameters of the seismic data processing model to be trained by a back propagation algorithm, and continuing the iterative processing based on the multi-task loss function; If the current number of iterations meets a preset iteration threshold, the seismic data processing model to be trained corresponding to the current number of iterations is used as the target seismic processing model.

3. A seismic data processing device, characterized in that: include: A data acquisition module, used for acquiring seismic data to be processed; a data processing module, configured to process the seismic data to be processed based on a pre-trained target seismic data processing model, and determine a recognition rate and seismic data to be applied corresponding to the seismic data to be processed; wherein the target seismic data processing model is a pre-trained neural network model; a target data determination module, configured to determine target seismic data based on the recognition rate, the seismic data to be applied, and the seismic data to be processed; The target seismic data processing model includes a shared network, a first task network, and a second task network; the data processing module is used to process the seismic data to be processed based on the shared network to obtain data features corresponding to the seismic data to be processed; determine a recognition rate corresponding to the seismic data to be processed based on the first task network and the data features; and determine seismic data to be applied corresponding to the seismic data to be processed based on the second task network and the data features; The shared network includes at least four groups of first convolutional layers; wherein the first convolutional layers include a ReLU activation function; two adjacent first convolutional layers in the shared network are connected by downsampling; the first task network includes at least three groups of second convolutional layers and at least two groups of fully connected layers; wherein the fully connected layers are located after the second convolutional layers; the second task network includes at least four groups of first convolutional layers; wherein two adjacent first convolutional layers in the second task network are connected by upsampling; the second task network and the first convolutional layers of the same scale in the shared network are connected by a skip connection; The device further includes: a target model determination module, configured to obtain raw data, process the raw data to determine a sample data set; train a seismic data processing model to be trained based on the sample data set to obtain the target seismic processing model; The target model determination module is configured to determine an identification label corresponding to each of the raw data, and determine at least one random square wave noise based on preset noise characteristic data; and determine the sample data set based on the identification label, the random square wave noise, and the raw data; The target model determination module is used to obtain the identification label of the original data, and use the original data whose identification label is normal data as positive sample data; wherein, the sample data includes an identification label and a suppression label; the suppression label of the positive sample data is the original data; the random square wave noise is superimposed on the positive sample data to determine the negative sample data, and the identification label of the negative sample data is modified to noise data; wherein, the suppression label of the negative sample data is the random square wave noise.

4. An electronic device, characterized in that: The electronic device comprises: at least one processor; and a memory communicatively connected to the at least one processor; wherein, The memory stores a computer program executable by the at least one processor. The computer program is executed by the at least one processor to enable the at least one processor to perform the seismic data processing method according to any one of claims 1 to 2.

5. A computer-readable storage medium, characterized in that The computer-readable storage medium stores computer instructions, and the computer instructions are used to enable a processor to implement the seismic data processing method according to any one of claims 1 to 2 when executed.

Images