Underwater high-pressure air gun bubble characteristic rapid prediction method, system and equipment based on machine learning, and medium

Abstract

The invention discloses an underwater high-pressure air gun bubble characteristic rapid prediction method, system and equipment based on machine learning, and a medium, and belongs to the field of deep sea resource exploration, and solves the problem that a traditional air gun bubble pressure wavelet prediction method depends on an empirical formula and a physical model. And the precision and efficiency of pressure wavelet calculation are influenced by factors such as unit division, boundary conditions and load working condition simulation authenticity. The method comprises the steps that a pressure wavelet data set of an air gun is collected and preprocessed; a training set and a test set are obtained, a machine learning model is adopted to learn the training set of the single air gun, and a forward modeling machine learning model of the single air gun is obtained; combining the air gun bubble kinetic equation into a single air gun forward modeling machine learning model to obtain a predicted complete pressure wavelet curve; and training and optimizing the pressure wavelet data set of the multiple air guns by adopting a machine learning model to obtain a multi-gun combination forward modeling machine learning model, and realizing prediction of the pressure wavelets. The method is suitable for air gun body design and manufacturing scenes.

Description

Technical Field

[0001] The invention belongs to the field of deep-sea resource exploration, and in particular relates to a technology for predicting bubble characteristics of an underwater high-pressure air gun. Background Art

[0002] Marine seismic exploration is an important means of finding and evaluating seabed oil and gas resources. As an artificial seismic wave source, the quality of the seismic wave signal generated by the airgun directly affects the exploration effect. The pressure wavelet generated by the airgun bubble is one of the key factors affecting the signal quality. The traditional airgun bubble pressure wavelet prediction method relies on empirical formulas and physical models, and the accuracy and efficiency of the pressure wavelet calculation need to be improved. For example, when using the finite element method to establish the airgun bubble model, it is necessary to define more parameters, and the calculation accuracy will be affected by factors such as unit division, boundary conditions, and the authenticity of load condition simulation. It takes too much time to calculate a large amount of data. In order to solve the limitations of traditional methods in predicting airgun bubble pressure waves, machine learning technology provides a new solution.

[0003] As an advanced data analysis technology, machine learning can learn and discover potential patterns from large amounts of data, providing a new perspective for the prediction of airgun bubble pressure subwaves. A neural network is a computational model inspired by the structure of neurons in the human brain that can recognize patterns or predict outcomes by learning from large amounts of data. A neural network consists of several layers, including an input layer, multiple hidden layers, and an output layer. The nodes of each layer are connected to the nodes of the adjacent layers through weights, and the input signals from the previous layer are processed through activation functions, which are then passed to the next layer. The process of training a neural network usually includes forward propagation, loss function calculation, and backpropagation algorithms to adjust the weights and biases in the network.

[0004] However, in the process of machine learning to realize airgun bubble pressure wavelet prediction, obtaining the pressure wavelet dataset is a difficult point to achieve this goal. The current existing technology does not have any commercial software that can realize airgun bubble pressure wavelet calculation, and the traditional calculation method is inefficient and has poor accuracy. Summary of the invention

[0005] The present invention proposes a method, system, equipment and medium for rapid prediction of underwater high-pressure air gun bubble characteristics based on machine learning, with the purpose of solving the problem that the traditional air gun bubble pressure sub-wave prediction method relies on empirical formulas and physical models, and the accuracy and efficiency of pressure sub-wave calculation are affected by factors such as unit division, boundary conditions, and load condition simulation authenticity.

[0006] The method for rapid prediction of bubble characteristics of underwater high-pressure air guns based on machine learning proposed in the present invention includes:

[0007] S1: Collect pressure wavelet data sets of a single airgun and a combination of multiple airguns and perform preprocessing;

[0008] S2: Divide the processed pressure wavelet data set into a training set and a test set, use a machine learning model to learn the training set of a single airgun, optimize the hyperparameters of the machine learning model, and obtain a forward machine learning model of a single airgun;

[0009] S3: Incorporating the airgun bubble dynamics equation into the single airgun forward machine learning model, and using the label of the single airgun pressure wavelet as a neural network input to obtain a predicted airgun bubble complete pressure wavelet curve;

[0010] S4: Using a machine learning model to train and optimize the pressure wavelet data sets of multiple air guns to obtain a multi-gun combination forward machine learning model, and using the multi-gun combination forward machine learning model to achieve prediction of the pressure wavelet.

[0011] Furthermore, a preferred solution is provided: the pressure wavelet data sets of the single airgun and the combination of multiple airguns are obtained through the mathematical and physical model of airgun bubbles.

[0012] Furthermore, a preferred solution is provided: the preprocessing includes: data feature scaling and numerical feature normalization.

[0013] Furthermore, a preferred solution is provided: the label of the single airgun pressure sub-wave comprises: a main pulse peak value and a bubble pulse peak value.

[0014] Furthermore, a preferred solution is provided: S2 includes:

[0015] Establishing a single airgun pressure wavelet forward model and initializing hyperparameters of the forward model;

[0016] The pressure wavelet data set of a single airgun is divided into a training set and a test set using the cross-validation method.

[0017] The single airgun pressure wavelet forward model is trained and optimized using a single airgun training set combined with hyperparameter optimization;

[0018] The prediction error of the model was tested using a single airgun test set and cost function;

[0019] The hyperparameters of the forward model are adjusted until the prediction error of the model is minimized, thereby obtaining a single airgun forward machine learning model.

[0020] Furthermore, a preferred solution is provided: the hyperparameters of the forward model include: the number of hidden layers, the number of neurons in the hidden layers, the activation function, the batch size and the number of iterations.

[0021] Furthermore, a preferred solution is provided: the S3 includes: integrating the airgun bubble dynamics equation into the cost function to achieve the prediction of the complete pressure sub-wave curve of the airgun bubble.

[0022] The present invention also proposes a computer device, comprising a memory and a processor, wherein the memory stores a computer program. When the processor runs the computer program stored in the memory, the processor executes the method for rapid prediction of bubble characteristics of underwater high-pressure air guns based on machine learning according to any one of the above-mentioned combinations.

[0023] The present invention also provides a computer-readable storage medium, which is used to store a computer program, and the computer program executes the steps of the method for rapid prediction of underwater high-pressure air gun bubble characteristics based on machine learning described in any combination of the above-mentioned solutions.

[0024] The present invention also proposes a system for rapid prediction of bubble characteristics of underwater high-pressure airguns based on machine learning, the system is implemented based on the method for rapid prediction of bubble characteristics of underwater high-pressure airguns based on machine learning described in any one of the above scheme combinations, and the system comprises:

[0025] Data processing module: used to collect pressure wavelet data sets of a single airgun and a combination of multiple airguns, and perform preprocessing;

[0026] Model building module: used to divide the processed pressure wavelet data set into training set and test set, use the machine learning model to learn the training set of a single airgun, optimize the hyperparameters of the machine learning model, and obtain the forward machine learning model of a single airgun;

[0027] Curve prediction module: used to combine the airgun bubble dynamics equation into the single airgun forward machine learning model, and use the label of the single airgun pressure wavelet as a neural network input to obtain the predicted airgun bubble complete pressure wavelet curve;

[0028] Model extension module: used to train and optimize the pressure wavelet data sets of multiple air guns using a machine learning model to obtain a multi-gun combination forward machine learning model, and use the multi-gun combination forward machine learning model to realize the prediction of the pressure wavelet.

[0029] Compared with the prior art, the advantages of the present invention are:

[0030] The present invention adopts the mathematical and physical model of airgun bubbles to obtain the airgun bubble pressure wavelet data set, and uses the cross-validation method to combine them into different training sets and test sets. The deep neural network is trained based on the training set, and the hyperparameter optimization algorithm is used to continuously adjust the hyperparameters of the neural network, such as the number of hidden layers, activation function, batch size, etc. The model is continuously adjusted according to the test set to achieve fast and accurate prediction of the airgun bubble pressure wavelet.

[0031] The present invention combines machine learning with airgun bubble pressure wavelet prediction, which can provide a reference for the design and manufacture of airgun gun bodies in engineering practice. Based on the airgun bubble pressure wavelet data set, a machine learning method is used to establish a model to predict the parametric target of the airgun bubble pressure wave and the complete pressure wave curve. Based on the single airgun pressure wavelet prediction model, a multi-gun combined pressure wavelet calculation forward model is established.

[0032] The method described in the present invention uses a self-developed mathematical and physical model of airgun bubbles to obtain a data set to train the model. Through the application of machine learning technology, the method can not only improve the accuracy of prediction, but also achieve more efficient data processing and interpretation in seismic exploration.

[0033] The present invention provides a new technical means for marine seismic exploration, aims to provide a reference for the design of airgun body parameters, and is expected to promote technological progress in this field.

[0034] The present invention is suitable for application scenarios such as air gun body design and manufacturing. BRIEF DESCRIPTION OF THE DRAWINGS

[0035] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0036] Figure 1 This is a flow chart of the single airgun bubble pressure sub-wave prediction based on machine learning as described in the first specific implementation mode of the present invention.

[0037] Figure 2 This is a schematic diagram of using a deep neural network to predict the pressure subwave characteristics of a single airgun as described in a specific embodiment of the present invention. The figure includes an input layer, a hidden layer and an output layer. The input layer is parameters such as the airgun chamber volume Vg0, excitation pressure Pg0, sinking depth Hg and airgun opening area Amax, and the output layer is curves such as the airgun bubble pressure subwave and spectrum.

[0038] Figure 3This is the expected effect diagram of underwater high-pressure airgun bubble pressure wavelet prediction using machine learning method as described in the first embodiment of the present invention. The blue solid line is the numerical calculation result of the mathematical and physical model of airgun bubbles, and the orange dotted line is the expected prediction result of the machine learning model. The neural network model needs to be continuously optimized in the future to achieve better prediction results. DETAILED DESCRIPTION

[0039] In the following description, specific details such as specific system structures and technologies are provided for the purpose of illustration rather than limitation, so as to provide a thorough understanding of the embodiments of the present application. However, it should be clear to those skilled in the art that the present application may also be implemented in other embodiments without these specific details. In other cases, detailed descriptions of well-known systems, devices, circuits, and methods are omitted to prevent unnecessary details from obstructing the description of the present application.

[0040] It should be understood that when used in this specification and the appended claims, the term "comprising" indicates the presence of described features, integers, steps, operations, elements and / or components, but does not exclude the presence or addition of one or more other features, integers, steps, operations, elements, components and / or groups thereof.

[0041] It should also be understood that the terms used in this application specification are only for the purpose of describing specific embodiments and are not intended to limit the present application. As used in this application specification and the appended claims, the singular forms "a", "an" and "the" are intended to include plural forms unless the context clearly indicates otherwise.

[0042] The following is a clear and complete description of the technical solutions in the embodiments of the present application in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.

[0043] In the following description, many specific details are set forth to facilitate a full understanding of the present application, but the present application may also be implemented in other ways different from those described herein, and those skilled in the art may make similar generalizations without violating the connotation of the present application. Therefore, the present application is not limited to the specific implementation methods disclosed below.

[0044] Implementation method 1:

[0045] Reference Figure 1 , Figure 2 , Figure 3 This embodiment will be described.

[0046] This embodiment provides a method for quickly predicting bubble characteristics of underwater high-pressure air guns based on machine learning. This embodiment provides the following technical solutions:

[0047] A method for quickly predicting bubble characteristics of underwater high-pressure air guns based on machine learning, comprising the following steps:

[0048] Step 1: Use the mathematical and physical model of airgun bubbles to obtain the pressure wavelet data sets of a single airgun and a combination of multiple airguns, and process the data sets.

[0049] Specifically, for the single airgun pressure wavelet prediction problem, before machine learning, data sets under different parameter settings are obtained based on the mathematical and physical model of the airgun bubble.

[0050] This implementation briefly introduces the acquisition of data sets by taking the multiple input variables and pressure wavelet curve output of a single airgun as an example. The airgun bubble far-field pressure wavelet curve and characteristic parameters under different parameters can be obtained through the mathematical and physical model of the airgun bubble. The input data includes parameters such as airgun volume, excitation pressure, sinking depth and environment, and the output data is the pressure wavelet curve.

[0051] 800 pressure points were extracted at a time interval of 0.5 ms. Based on the curve, characteristic parameters such as the main pulse peak value, ghost wave peak value, and initial bubble ratio can also be extracted.

[0052] When more parameters are involved, data needs to be preprocessed to improve computational efficiency. For example, feature engineering is needed to scale data features to ensure that the algorithm learns more efficiently and accurately. By unifying the scales of different features, the convergence speed and generalization ability of the model can be improved.

[0053] For example, the sinking depth of a certain type of air gun is generally 5 to 12 meters, and the volume is 20 to 250 in. 3 , the working pressure is 1000~2000psi. When processing the input data, in order to ensure that the neural network does not favor certain features due to the scale differences of the features, all numerical features can be normalized to the range of 0 to 1.

[0054] Step 2: Divide the data set into a training set and a test set, and use a machine learning model to learn the single airgun training data set, such as gradient descent, neural network, etc. The model structure and hyperparameters are continuously optimized to obtain a single airgun forward machine learning model.

[0055] Specifically, this step requires combining the data set obtained in step 1 into different training sets and test sets using a cross-validation method to facilitate evaluation and verification of the model.

[0056] A forward model is established, and the model's hyperparameters such as the number of hidden layers, activation function, batch size, and number of iterations are initialized. The model can approximately simulate or predict the mapping relationship from a multivariate input to output (characteristic parameters such as initial bubble ratio and bubble pulsation period) by learning the complex relationship between input and output.

[0057] The model is continuously trained and optimized through the training data set, and the test data set is used to evaluate and optimize the accuracy of the model. For example, the polynomial regression algorithm can be used to predict the airgun parameter target (such as the initial bubble ratio). Formula (1) is the regularized cost function J(w,b) formula, which is mainly used to evaluate the difference between the predicted value and the actual value of the model, in order to quantify the performance of the model. Formulas (2) and (3) are gradient descent formulas. Gradient descent is an optimization algorithm commonly used in cost functions in machine learning. It iteratively adjusts the parameters of the model to reduce the prediction error.

[0058]

[0059] In the formula, m represents the number of training samples, x and y represent the input variables and output variables of the data respectively, and f w,b (x (i) ) represents the model prediction value, y (i) Represents the true value, i represents the i-th training data, λ represents the regularization parameter, which is used to control the strength of the regularization term; w and b represent the weight and bias of the model respectively.

[0060] Step 3: Integrate the airgun bubble dynamics equation into the forward model of a single airgun, and use the main pulse peak value, bubble pulse peak value and other labels in the pressure wavelet as neural network inputs to predict the complete pressure wavelet curve of the airgun bubble.

[0061] This step combines the deep neural network with the bubble dynamics equation to achieve the prediction of the complete airgun bubble pressure wavelet curve.

[0062] Specifically, based on the single airgun forward model, the airgun bubble dynamics equation is incorporated into the cost function, which can more accurately predict the complete pressure wavelet curve generated by the airgun bubble.

[0063] This process involves an in-depth understanding of the dynamics of airgun bubbles, including the generation, expansion, contraction and interaction of bubbles with the surrounding fluid, and taking into account multiple factors such as multiple bubbles and free liquid surface. It combines thermodynamic models to consider thermodynamic factors such as airgun gas release and bubble heat transfer to achieve the prediction of airgun bubble pressure waves.

[0064] Formula (4) and Formula (5) are the flow field velocity u(r) and pressure p(r) at any point r in the flow field.

[0065]

[0066] Where R is the bubble radius; is the bubble pulsation velocity; H is the enthalpy difference from the flow field at the bubble wall to infinity; p a is the flow field pressure at infinity; v is the relative speed between the bubble migration speed and the flow field speed.

[0067] In addition, characteristic parameters in the pressure wavelet, such as main pulse peak, ghost wave peak, bubble pulse peak and other labels, can also be extracted and used as neural network input to achieve accurate simulation of the complete pressure wavelet curve. This is also a feature engineering strategy that can specifically improve the prediction accuracy of the model in key areas, thereby improving the overall prediction performance and the generalization ability of the model.

[0068] Step 4: Expand the model to more complex situations, i.e., multiple airgun combinations, to improve the generalization ability of the model. The machine learning model is trained and optimized on the pressure wavelet dataset of multiple airgun combinations to obtain the forward machine learning model of the multi-gun combination.

[0069] This step expands the model to the case of multi-gun combinations in order to improve the generalization ability of the model so that it can adapt to more complex practical application scenarios. By training and optimizing the machine learning model for the multi-gun combination pressure wavelet data set, a forward machine learning model suitable for the multi-gun combination can be obtained. In this process, factors that need to be considered include the influence of the mutual influence between airguns, the airgun sinking depth, and the airgun asynchronous excitation factors on the pressure wavelet. Through literature research and trial calculations, a pressure wavelet data set with good quality can be obtained. On this basis, a multi-gun pressure wavelet forward model is established to achieve the prediction of the pressure wavelet.

[0070] This implementation method uses a mathematical and physical model of airgun bubbles to obtain an airgun bubble pressure wavelet data set, which is combined into different training sets and test sets using a cross-validation method. A deep neural network is trained based on the training set, and a hyperparameter optimization algorithm is used to continuously adjust the number of hidden layers, activation function, batch size and other hyperparameters of the neural network. The model is continuously adjusted according to the test set to achieve rapid and accurate prediction of the airgun bubble pressure wavelet.

[0071] Implementation Method 2

[0072] A system for rapid prediction of bubble characteristics of an underwater high-pressure air gun based on machine learning is provided. The system is implemented based on the method for rapid prediction of bubble characteristics of an underwater high-pressure air gun based on machine learning described in Implementation Example 1.

[0073] This implementation provides the following technical solutions:

[0074] The rapid prediction system of underwater high-pressure airgun bubble characteristics based on machine learning includes:

[0075] Data processing module: used to collect pressure wavelet data sets of a single airgun and a combination of multiple airguns, and perform preprocessing;

[0076] Model building module: used to divide the processed pressure wavelet data set into training set and test set, use the machine learning model to learn the training set of a single airgun, optimize the hyperparameters of the machine learning model, and obtain the forward machine learning model of a single airgun;

[0077] Curve prediction module: used to combine the airgun bubble dynamics equation into the single airgun forward machine learning model, and use the label of the single airgun pressure wavelet as a neural network input to obtain the predicted airgun bubble complete pressure wavelet curve;

[0078] Model extension module: used to train and optimize the pressure wavelet data sets of multiple air guns using a machine learning model to obtain a multi-gun combination forward machine learning model, and use the multi-gun combination forward machine learning model to realize the prediction of the pressure wavelet.

[0079] Those skilled in the art will appreciate that the above are only preferred embodiments of the present invention, and the various embodiments of the present disclosure and / or the features described in the claims may be combined or combined in various ways, even if such combinations or combinations are not explicitly described in the present disclosure. It is not intended to limit the present invention. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art may still modify the technical solutions described in the aforementioned embodiments, or perform equivalent substitutions on some of the technical features therein. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention shall be included in the protection scope of the present invention.

[0080] Although the preferred embodiments of the present invention have been described, those skilled in the art may make other changes and modifications to these embodiments once they have learned the basic creative concept. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments and all changes and modifications that fall within the scope of the present invention.

[0081] Obviously, those skilled in the art can make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if these modifications and variations of the present invention fall within the scope of the claims of the present invention and their equivalents, the present invention is also intended to include these modifications and variations.

[0082] It will be appreciated by those skilled in the art that the embodiments of the present disclosure may be provided as methods, systems or computer program products. Therefore, the present disclosure may be provided in the form of a complete hardware embodiment, a complete software embodiment or an embodiment combining software and hardware. Furthermore, the present disclosure may be provided in the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0083] The present disclosure is described with reference to flowcharts and / or block diagrams of methods, devices (systems) and computer program products according to embodiments of the present disclosure. It should be understood that each process and / or block in the flowchart and / or block diagram and the combination of processes and / or blocks in the 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, an embedded processor or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 These computer program instructions can also be stored in a computer-readable memory that can guide a computer or other programmable data processing device to work in a specific way, so that the instructions stored in the computer-readable memory produce a product including an instruction device, which implements the functions specified in the process. Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0084] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0085] Finally, it should be noted that the above implementation modes are only used to illustrate the technical solutions of the present disclosure rather than to limit its protection scope. Although the present disclosure has been described in detail with reference to the above implementation modes, ordinary technicians in the relevant field should understand that after reading the present disclosure, those skilled in the art can still make various changes, modifications or equivalent substitutions to the specific implementation modes of the invention, but these changes, modifications or equivalent substitutions are all within the protection scope of the disclosed claims to be approved.

Claims

1. A method for rapid prediction of bubble characteristics of underwater high-pressure airguns based on machine learning, characterized in that: The method comprises: S1: Collect pressure wavelet data sets of a single airgun and a combination of multiple airguns and perform preprocessing; S2: Divide the processed pressure wavelet data set into a training set and a test set, use a machine learning model to learn the training set of a single airgun, optimize the hyperparameters of the machine learning model, and obtain a forward machine learning model of a single airgun; S3: Incorporating the airgun bubble dynamics equation into the single airgun forward machine learning model, and using the label of the single airgun pressure wavelet as a neural network input to obtain a predicted airgun bubble complete pressure wavelet curve; S4: Using a machine learning model to train and optimize the pressure wavelet data sets of multiple air guns to obtain a multi-gun combination forward machine learning model, and using the multi-gun combination forward machine learning model to achieve prediction of the pressure wavelet.

2. The method for rapid prediction of bubble characteristics of underwater high-pressure airguns based on machine learning according to claim 1 is characterized in that: The pressure wavelet data sets of the single airgun and the combination of multiple airguns are obtained through the mathematical and physical model of airgun bubbles.

3. The method for rapid prediction of bubble characteristics of underwater high-pressure air guns based on machine learning according to claim 1 is characterized in that: The preprocessing includes: data feature scaling and numerical feature normalization.

4. The method for rapid prediction of bubble characteristics of underwater high-pressure airguns based on machine learning according to claim 1 is characterized in that: The labels of the single airgun pressure sub-waves include: main pulse peak value and bubble pulse peak value.

5. The method for rapid prediction of bubble characteristics of underwater high-pressure airguns based on machine learning according to claim 1 is characterized in that: The S2 includes: Establishing a single airgun pressure wavelet forward model and initializing hyperparameters of the forward model; The pressure wavelet data set of a single airgun is divided into a training set and a test set using the cross-validation method. The single airgun pressure wavelet forward model is trained using a single airgun training set combined with hyperparameter optimization; The prediction error of the model was tested using a single airgun test set and cost function; The hyperparameters of the forward model are adjusted until the prediction error of the model is minimized, thereby obtaining a single airgun forward machine learning model.

6. The method for rapid prediction of bubble characteristics of underwater high-pressure airguns based on machine learning according to claim 5 is characterized in that: The hyperparameters of the forward model include: the number of hidden layers, the number of neurons in the hidden layers, the activation function, the batch size and the number of iterations.

7. The method for rapid prediction of bubble characteristics of underwater high-pressure airguns based on machine learning according to claim 5 is characterized in that: The S3 includes: integrating the airgun bubble dynamics equation into the cost function to achieve the prediction of the complete pressure sub-wave curve of the airgun bubble.

8. Computer device, characterized in that It includes a memory and a processor, wherein a computer program is stored in the memory. When the processor runs the computer program stored in the memory, the processor executes the method for rapid prediction of bubble characteristics of underwater high-pressure air guns based on machine learning according to any one of claims 1-7.

9. A computer-readable storage medium, characterized in that The computer-readable storage medium is used to store a computer program, and the computer program executes the steps of the method for rapid prediction of bubble characteristics of underwater high-pressure air guns based on machine learning as described in any one of claims 1-7.

10. The underwater high-pressure air gun bubble characteristics rapid prediction system based on machine learning is characterized by: The system is implemented based on the method for rapid prediction of bubble characteristics of underwater high-pressure air guns based on machine learning as described in any one of claims 1 to 7, and the system includes: Data processing module: used to collect pressure wavelet data sets of a single airgun and a combination of multiple airguns, and perform preprocessing; Model building module: used to divide the processed pressure wavelet data set into training set and test set, use the machine learning model to learn the training set of a single airgun, optimize the hyperparameters of the machine learning model, and obtain the forward machine learning model of a single airgun; Curve prediction module: used to combine the airgun bubble dynamics equation into the single airgun forward machine learning model, and use the label of the single airgun pressure wavelet as a neural network input to obtain the predicted airgun bubble complete pressure wavelet curve; Model extension module: used to train and optimize the pressure wavelet data sets of multiple air guns using a machine learning model to obtain a multi-gun combination forward machine learning model, and use the multi-gun combination forward machine learning model to realize the prediction of the pressure wavelet.

Images