Three-dimensional pore pressure prediction method and device

Abstract

According to the method, a deep learning network model and a seismic inversion result are introduced into three-dimensional pore pressure, elastic parameter data such as logging density, longitudinal and transverse wave velocity and the like and three-dimensional elastic attribute data obtained through seismic inversion are collected as a data set, and the network model is optimized and trained by using an optimization algorithm; the trained model is applied to a test area, a three-dimensional pore pressure prediction result is calculated and output, more comprehensive geologic feature attributes are provided by using multi-source data such as logging and earthquakes, deep features can be learned from the multi-source data, and pore pressure prediction can be better carried out.

Description

Technical Field

[0001] The invention relates to a three-dimensional pore pressure prediction method and device, belonging to the technical field of oil and gas field development and coal field mining and exploration. Background Art

[0002] Formation pore pressure is a key factor in assessing reservoir stability, designing drilling plans, and predicting production performance. Pore pressure refers to the force exerted on the fluid in the pores of the rock formation. It is usually related to the depth of the formation and expressed in static water pressure, such as megapascals or pounds per square inch. Abnormal pore pressure may manifest as overpressure and underpressure. Overpressure is caused by additional pressure sources that cause the pore pressure to exceed the normal level, which may cause dangerous events such as blowouts; while underpressure means that the pore pressure is lower than the normal level, which may cause problems such as well leakage and circulation loss, increasing drilling costs.

[0003] Current pore pressure prediction methods can be divided into direct methods and indirect methods. Direct methods include leak tests, rock mechanics experiments, and seismic data acquisition, which aim to directly measure formation pressure. However, these methods can usually only provide measurements at a specific depth, it is difficult to obtain a continuous pore pressure profile, and the process of collecting downhole data is costly and risky. Indirect methods usually rely on theoretical models to derive pore pressure. The more common of these methods are the use of undercompaction theory and effective stress method.

[0004] In recent years, the rapid development of machine learning technology, especially the rise of deep learning, has provided new possibilities for the prediction of pore pressure. Compared with traditional mechanical models, data-driven models provide a simpler establishment process and strong adaptability, and can handle a variety of diverse, large and complex data. These models are good at capturing the intrinsic patterns of data and have strong generalization capabilities. Many studies have explored the application of machine learning in pore pressure prediction. Although machine learning models have been successful in some applications, many existing models still face the challenge of capturing the intrinsic temporal dynamics of the data. Existing models perform poorly when processing time series data, resulting in reduced prediction accuracy.

[0005] In order to address the shortcomings of traditional methods, the present invention proposes a new deep learning network model framework to predict pore pressure. The model combines the efficient parallel computing capability of convolutional neural networks and the sequential processing advantages of recurrent neural networks, effectively identifies abnormal pressure events across multiple layers, and improves computing efficiency through the convolution acceleration mechanism. It is suitable for real-time prediction of large-scale three-dimensional data. At the same time, by combining deep learning technology with high-precision seismic inversion, it can not only achieve high-precision three-dimensional pore pressure prediction, but also reveal lateral changes in pore pressure that are difficult to detect by traditional methods. This provides strong technical support for the precise management and efficient development of oil and gas fields. Summary of the invention

[0006] In view of the deficiencies in the prior art, the present invention provides a method and device for predicting pore pressure based on a deep learning network model to overcome the deficiencies in the prior art.

[0007] In a first aspect, the present invention provides a three-dimensional pore pressure prediction method based on a deep learning network model, comprising the following steps:

[0008] S1: Collect well logging data and seismic data for preprocessing and construct a data set. The data set includes a training set and a validation set. The training set includes labeled training data and unlabeled training data. The labeled training data includes relevant curve data in the well logging data, such as density, longitudinal wave velocity P, and shear wave velocity S. The unlabeled training data is the three-dimensional elastic parameter information obtained by prestack inversion of seismic data;

[0009] Furthermore, the relevant curve data in the logging data, such as density, P-wave velocity, and S-wave velocity, are extracted as labeled training data, and the three-dimensional elastic parameter information is obtained through prestack inversion of seismic data as unlabeled training data;

[0010] S2: Define the initial parameters and initial model of the model, establish the deep learning network model structure according to the discretized state space model framework, and use the collected logging and seismic data for iterative training to ensure that the model learns the nonlinear relationship in the input data;

[0011] S3: Randomly select training set data and validation set data in a certain ratio to train the model, and continuously iterate the model by adjusting the state size and sequence length hyperparameters until the value of the loss function falls within the preset threshold range or the number of iterations reaches the preset value;

[0012] S4: Input the three-dimensional density, longitudinal wave velocity P, and shear wave velocity S of the three-dimensional seismic data into the trained model, output the prediction result of the model, and obtain the prediction result of the three-dimensional pore pressure.

[0013] In a second aspect, the present invention also provides a three-dimensional pore pressure prediction device based on a deep learning network model, comprising:

[0014] The data acquisition module collects logging data and seismic data and constructs a data set. The data set includes a training set, a validation set, and a test set. The training set includes labeled training data and unlabeled training data.

[0015] The model building module builds the deep learning network model structure based on the discretized state space model framework and uses the collected logging and seismic data for iterative training to ensure that the model learns the nonlinear relationship in the input data;

[0016] The model training module randomly selects training set data and validation set data to train the model until the value of the loss function falls within the preset threshold range or the number of iterations reaches the preset value;

[0017] Specifically, the training set data and the validation set data can be randomly selected in a ratio of 8:2 for model training. The training process includes two parts: forward propagation and back propagation parameter update. The model is iterated continuously by adjusting the hyperparameters until the value of the loss function falls within the preset threshold range or the number of iterations reaches the preset value.

[0018] The prediction module inputs the three-dimensional density, longitudinal wave velocity P, and shear wave velocity S of the three-dimensional seismic data into the trained model, outputs the prediction results of the model, and obtains the prediction results of the three-dimensional pore pressure.

[0019] The beneficial effects of the present invention are as follows: existing machine learning methods often have difficulty in capturing the intrinsic temporal dynamics of geological data, such as well logging and seismic data, resulting in less than ideal prediction accuracy; most existing pore pressure prediction technologies focus on one-dimensional prediction and are not adapted to three-dimensional applications, while the method proposed in this article successfully achieves three-dimensional pore pressure prediction, which highlights the shortcomings of other methods in this regard. The present invention combines a deep learning network model with seismic inversion technology to jointly mine the pore pressure characteristics in multi-source data such as well logging and seismic data. The deep learning network model uses convolution operations instead of recursive inference to reduce the computational complexity of processing long sequence data. Compared with traditional recursive operations, convolution operations can be processed in parallel, thereby significantly improving computational efficiency. It has obvious advantages in parallel processing, especially in large-scale data sets, which can significantly reduce training time. At the same time, the deep learning network model can effectively capture long-term dependencies and complex nonlinear trends in different data without matching scales, thereby improving the accuracy and reliability of predictions, improving the generalization of the model and the accuracy of pore pressure prediction, and being able to accurately predict the distribution of lithofacies in unknown wells or well-free areas. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] In order to more clearly illustrate the specific implementation methods of the present invention or the technical solutions in the prior art, the drawings required for use in the specific implementation methods or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are some implementation methods of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0021] Figure 1 It is a schematic flow chart of a method for predicting three-dimensional pore pressure based on a deep learning network model according to an embodiment of the present invention.

[0022] Figure 2 It is the input attributes and label data of the labeled training data input in one embodiment of the present invention.

[0023] Figure 3 It is an input attribute of the unlabeled seismic inversion elastic parameter training data input in an embodiment of the present invention.

[0024] Figure 4 It is the input data and prediction result of an embodiment of the present invention.

[0025] Figure 5 It is the three-dimensional pore pressure prediction result in one embodiment of the present invention. DETAILED DESCRIPTION

[0026] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions 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.

[0027] like Figure 1 As shown, in a first aspect, the present invention provides a three-dimensional pore pressure prediction method, comprising the following steps:

[0028] S1: Collect well logging data and seismic data for preprocessing and construct a data set, which includes a training set and a validation set. The training set includes labeled training data and unlabeled training data;

[0029] Further, such as Figure 2-3 The labeled training data shown includes relevant curve data in the well logging data, such as density, longitudinal wave velocity P, and shear wave velocity S, and the unlabeled training data is the three-dimensional elastic parameter information obtained by pre-stack inversion of seismic data;

[0030] Furthermore, the preprocessing includes normalizing the elastic parameters obtained by seismic inversion after feature extraction, and mapping them to a specific interval to speed up the training and convergence of the model. The normalization formula is as follows:

[0031]

[0032] Where d is the characteristic value of the elastic parameter data, dmin and dmax represent the minimum and maximum values ​​in the data set, respectively, and d` is the normalized value in the range of [0,1].

[0033] S2: Define the initial parameters and initial model of the model, establish the deep learning network model structure according to the discretized state space model framework, and use the collected logging and seismic data for iterative training to ensure that the model learns the nonlinear relationship in the input data; the initial architecture of the deep learning network model is as follows:

[0034] Among them, h(t) represents the hidden state, describing the current internal state of the system, h'(t) represents the spatial state of the next step; x(t) represents the external input, which affects the behavior of the system; y(t) is the output of the system; where A(t)∈R N×N ,B(t)∈R N×M ,C(t)∈R M×N ,They are also often called state matrices, input matrices, and output matrices.

[0035] Define the initial parameters of the model, including the state matrix A, input matrix, output matrix, time step and convolution kernel size;

[0036] Specifically, determine the initial state matrix A as shown below:

[0037]

[0038] Among them, A nk Represents the element in the nth row and kth column of the matrix; B n represents the th element of the matrix, which together describe the evolution of the state space and the approximation process of the function.

[0039] You can also choose a diagonal matrix or a more complex structured matrix, such as a DPLR matrix. Set the initial values ​​of the input matrix B and the output matrix C, and choose an appropriate time step for the discretized model.

[0040] Construct discrete time model: Convert the continuous time model into a discrete time model and obtain the discrete time state equation as shown below;

[0041]

[0042] Among them, x k is the state vector at discrete time point k, where k represents the number of discrete time steps. As k increases, the system state x k Will update to the state x of the next time step k+1 ;f k is the input vector.

[0043] In order to use it for pore pressure prediction, the initial architecture of deep learning needs to be discretized as u = (u 0 ,u 1 ,u 2 , ...), so that deep learning can effectively process input data from well logging or seismic, and the discretized state transfer equation is:

[0044]

[0045] y k=Ch k y k refers to the discretized output; h k represents the hidden state at time k-1, h k-1 Represents the hidden state at the previous moment;

[0046] in, The discretized matrix can be expressed as:

[0047]

[0048] Where exp refers to the natural exponential function; (Δ, A, B) are continuous parameters, is a discrete parameter, and the transformation is given by the formula and Definition, where (f A ,f B ) is the discretization rule that ensures that the model can maintain stable performance and accuracy when processing well log and seismic time series.

[0049] According to the discretized state equation, define the convolution kernel The size of the convolution kernel is determined according to the resolution of the input data to ensure that the convolution kernel can capture long-range dependencies, so that the output y can be expressed as the input x and the convolution kernel The convolution can be expressed as:

[0050]

[0051] in The derivation is to assume that the initial state is x -1 = 0. Since the discrete recursion is linear, we expand it analytically to obtain the equivalent convolutional form of the model.

[0052]

[0053] In other words, for input x, the output y has a simple closed form:

[0054]

[0055] This can then be vectorized into a single convolution via the explicit formula for the kernel:

[0056]

[0057] This equation is a single convolution, referred to as a state-space kernel.

[0058] The convolution operation is computed using the following state-space kernel, processing the input sequence in parallel to increase computation speed:

[0059]

[0060] Among them, h i The weight of the state-space kernel, u k-i is the input sequence. The deep learning network model enhances the flexibility and adaptability of the model by associating weight parameters with input data. Specifically, the parameters of the model can be dynamically adjusted according to the characteristics of the input, thereby achieving optimized prediction for different data.

[0061] Adopting a selective state space model, the model parameters can be dynamically adjusted according to the characteristics of the input data.

[0062] B=S B (x)

[0063] C=S C (x)

[0064] Δ=τA(Parameter+S A (x))(8)

[0065] In the above formula: S B (x), S C (x) represents a matrix dynamically generated according to the input x, allowing the model to adjust its behavior according to different characteristics of the input; Δ represents a learnable parameter used to estimate the discrete interval of the input samples, enhancing the handling of time dependencies.

[0066] S3: Randomly select training set data and validation set data in a certain ratio to train the network model, and continuously iterate the model by adjusting the hyperparameters until the value of the loss function falls within the preset threshold range or the number of iterations reaches the preset value;

[0067] Specifically, the model training method in step S3 specifically includes the following contents: considering the influence of different input attributes on the prediction results, inputting seismic attributes and logging parameters into the network model, obtaining the prediction results, calculating the error between the output value of the output layer and the actual value, adjusting the model hyperparameters according to the loss error, such as the learning rate of the model, the number of network layers, the number of neurons, the state size and the sequence length, repeatedly performing the above steps, and optimizing the deep learning network model multiple times until the value of the loss function falls within the preset threshold range, or the number of iterations reaches the preset value;

[0068] Specifically, the training set data and the validation set data are randomly selected in a ratio of 8:2 for model training. The training process of the neural network includes two parts: forward propagation and back propagation parameter update. In the forward propagation process, the output signal of each layer will be passed to the next layer as the input signal. These input signals pass through the network model and the additional hidden layer in turn, and finally reach the output layer. The nonlinear transformation is completed through the activation function, and the prediction result is output;

[0069] Loss function calculation:

[0070] Among them, L sz is the cross entropy loss function, y i It is a soft tag, P i is the predicted probability distribution of pore pressure by the trained model, n is the number of parameter categories, T is the temperature factor, The adjustment factor of the cross entropy loss function, is the sum of absolute errors, which represents the average deviation between the predicted probability distribution of the model and the soft labels produced by the adjusted network model. max(y) is the maximum value of the soft label, which represents the maximum probability value output by the adjusted network model in all categories. min(y) is the minimum value of the soft label, which represents the minimum probability value output by the adjusted processing network model in all categories.

[0071] Specifically, the network model architecture in one embodiment of the present invention is a plurality of convolutional layers, a linear projection layer and a dynamic state space module. In this model, the convolution unit is composed of two convolutional layers, wherein the first convolutional layer contains 32 3x3 convolutional kernels, the step length is 1, and the SiLU activation function is adopted; the second convolutional layer contains 64 3x3 convolutional kernels, the step length is also 1, and the SiLU activation function is also adopted; in the dynamic state space module, the number of states is set to 64, and the sequence length is set to 100;

[0072] S4: Input the three-dimensional density, longitudinal wave velocity P, and shear wave velocity S of the three-dimensional seismic data into the trained deep learning network model, output the prediction results of the model, and obtain the prediction results of the three-dimensional pore pressure.

[0073] like Figure 4 FIG. 1 is a schematic diagram of input data and prediction results according to an embodiment of the present invention. Figure 4 a to Figure 4 c is the predicted result of P wave velocity, S wave velocity and density, Figure 4 d is the actual logging pore pressure (solid line) and the predicted result (dashed line).

[0074] like Figure 5 As shown, in one embodiment of the present invention, the model is applied to the three-dimensional pore pressure prediction result of the three-dimensional seismic data area to be predicted, all wells are included in the training process, and data such as density, P-wave and S-wave velocity are input into the optimized deep learning network model to obtain the three-dimensional pore pressure prediction result.

[0075] The present invention introduces deep learning network models and seismic inversion results into three-dimensional pore pressure, and uses multi-source data such as well logging and seismic data to provide more comprehensive geological feature attributes, which can jointly learn deep layer features from multi-source data and better predict pore pressure. Due to the above technical solution, the present invention has the following advantages:

[0076] 1. Extracting features such as density and wave velocity from well logging data and elastic property features from seismic data provides rich input features for pore pressure prediction, which helps to improve the prediction ability of the model;

[0077] 2. The present invention uses a deep learning network model and comprehensively utilizes labeled logging data and unlabeled seismic data for training, which provides rich input features for pore pressure prediction and helps to improve the prediction ability of the model;

[0078] 3. It breaks through the one-dimensional limitation of traditional methods and realizes high-precision prediction of three-dimensional pore pressure. The prediction results are highly matched with the actual measured values ​​and the error is smaller.

[0079] 4. The deep learning network model combines the efficient parallel computing capabilities of convolutional neural networks and the sequential processing advantages of recurrent neural networks, improves computing efficiency, and can process large-scale three-dimensional data in real time.

[0080] In a second aspect, the present invention also provides a three-dimensional pore pressure prediction device based on a deep learning network model, comprising:

[0081] The data acquisition module collects well logging data and seismic data for preprocessing and constructs a data set. The data set includes a training set and a validation set. The training set includes labeled training data and unlabeled training data. The labeled training data includes relevant curve data in the well logging data, such as density, longitudinal wave velocity P, and shear wave velocity S. The unlabeled training data is the three-dimensional elastic parameter information obtained by prestack inversion of seismic data.

[0082] The model building module defines the initial parameters and initial model of the model, establishes the deep learning network model structure according to the discretized state space model framework, and uses the collected logging and seismic data for iterative training to ensure that the model learns the nonlinear relationship in the input data;

[0083] The model training module randomly selects training set data and validation set data in a certain ratio to train and validate the model, and continuously iterates the model by adjusting the state size and sequence length hyperparameters until the value of the loss function falls within the preset threshold range or the number of iterations reaches the preset value;

[0084] The prediction module inputs the three-dimensional density, longitudinal wave velocity P, and shear wave velocity S of the three-dimensional seismic data into the trained deep learning network model, outputs the prediction results of the model, and obtains the prediction results of the three-dimensional pore pressure.

[0085] The present invention provides a computer storage medium, wherein the computer storage medium stores at least one instruction, wherein the instruction is suitable for being loaded by a processor and executing the method steps of one or more embodiments of the present specification.

[0086] The present invention provides an electronic device, which may include: a processor and a memory; wherein the memory stores a computer program, and the computer program is suitable for being loaded by the processor and executing the method steps of one or more embodiments of the present specification.

[0087] The above embodiments only express several implementation methods of the present invention, and the descriptions thereof are relatively specific and detailed, but they cannot be understood as limiting the scope of the present invention. It should be pointed out that, for those of ordinary skill in the art, several modifications and improvements can be made without departing from the concept of the present invention, and these all belong to the protection scope of the present invention. Therefore, the protection scope of the present invention shall be subject to the attached claims.

Claims

1. A three-dimensional pore pressure prediction method based on a deep learning network model, characterized in that: The following steps are involved: S1: Collect well logging data and seismic data for preprocessing and construct a data set. The data set includes a training set and a validation set. The training set includes labeled training data and unlabeled training data. The labeled training data includes relevant curve data in the well logging data, such as density, longitudinal wave velocity P, and shear wave velocity S. The unlabeled training data is the three-dimensional elastic parameter information obtained by prestack inversion of seismic data; S2: Define the initial parameters and initial model of the model, establish the deep learning network model structure according to the discretized state space model framework, and use the collected logging and seismic data for iterative training to ensure that the model learns the nonlinear relationship in the input data; S3: Randomly select training set data and validation set data in a certain ratio to train and validate the model, and continuously iterate the model by adjusting the state size and sequence length hyperparameters until the value of the loss function falls within the preset threshold range or the number of iterations reaches the preset value; S4: Input the three-dimensional density, longitudinal wave velocity P, and shear wave velocity S of the three-dimensional seismic data into the trained deep learning network model, output the prediction results of the model, and obtain the prediction results of the three-dimensional pore pressure.

2. The three-dimensional pore pressure prediction method based on the deep learning network model according to claim 1, characterized in that: Preprocessing includes normalizing the elastic parameters obtained by seismic inversion after feature extraction, and mapping them to a specific interval to speed up the training and convergence of the model. The normalization formula is as follows: Where d is the characteristic value of the elastic parameter data, dmin and dmax represent the minimum and maximum values ​​in the data set, respectively, and d` is the normalized value in the range of [0,1].

3. The three-dimensional pore pressure prediction method based on the deep learning network model according to claim 1, characterized in that: Step S1 is specifically to randomly divide the data into a training set and a validation set in a ratio of 8:

2. The training process includes two parts: forward propagation and back propagation parameter update.

4. The three-dimensional pore pressure prediction method based on the deep learning network model according to claim 3, characterized in that: Back propagation and parameter updating include applying the trained model to the test well and inputting the corresponding attributes, calculating the error between the output value of the output layer and the actual value, and adjusting the model hyperparameters such as the learning rate, number of network layers, and number of neurons of the model according to the loss error; Loss function calculation: Among them, L sz is the cross entropy loss function, y i It is a soft tag, P i is the predicted probability distribution of pore pressure by the trained model, n is the number of parameter categories, T is the temperature factor, The adjustment factor of the cross entropy loss function, is the sum of absolute errors, which represents the average deviation between the predicted probability distribution of the model and the soft labels produced by the adjusted network model. max(y) is the maximum value of the soft label, which represents the maximum probability value output by the adjusted network model in all categories. min(y) is the minimum value of the soft label, which represents the minimum probability value output by the adjusted processing network model in all categories.

5. The three-dimensional pore pressure prediction method based on the deep learning network model according to claim 1, characterized in that: Step S2 also includes: defining the initial parameters of the model including the state matrix A, the input matrix B, the output matrix C, the time step and the convolution kernel size, and the initial architecture of the deep learning network model, as follows: h'(t) = A(t)x(t) + B(t)u(t) y(t)=C(t)x(t)+D(t)u(t) Among them, h(t) represents the hidden state, describing the current internal state of the system, h'(t) represents the spatial state of the next step; x(t) represents the external input, which affects the behavior of the system; y(t) is the output of the system; where A(t)∈R N×N ,B(t)∈R N×M ,C(t)∈R M×N ; Specifically, determine the initial state matrix A as shown below: Among them, A nk Represents the element in the nth row and kth column of the matrix; B n represents the th element of the matrix; Construct discrete time model: Convert the continuous time model into a discrete time model and obtain the discrete time state equation as shown below; Among them, x k is the state vector at discrete time point k, where k represents the number of discrete time steps. As k increases, the system state x k Will be updated to the state x at the next time step k+1 ;f k is the input vector.

6. The three-dimensional pore pressure prediction method based on the deep learning network model according to claim 5, characterized in that: The initial architecture of deep learning is discretized into u = (u0, u1, u2, ...), so that deep learning can effectively process input data from well logging or seismic. The state transfer equation after discretization is: y t =Ch(t) y t refers to the discretized output, t specifically refers to the discretized time point; h(t) represents the hidden state at time t, and h(t-1) represents the hidden state at the previous moment; The following formula is used to calculate the discrete matrix during the discretization process: in, Represents the discretized matrix, which converts continuous parameters (Δ, A, B) into discrete parameters This transformation is given by the formula and Definition, where (f A ,f B ) is the discretization rule; According to the discretized state equation, define the convolution kernel The size of the convolution kernel is determined according to the resolution of the input data to ensure that the convolution kernel can capture long-range dependencies, so that the output y can be expressed as the input x and the convolution kernel The convolution can be expressed as: Assume that the initial state is x -1 = 0. Since the discrete recursion is linear, we expand it analytically to obtain the equivalent convolutional form of the model. In other words, for input x, the output y has a simple closed form: This can then be vectorized into a single convolution via the explicit formula for the kernel: This equation is a single convolution, referred to as a state-space kernel.

7. The three-dimensional pore pressure prediction method based on the deep learning network model according to claim 6, characterized in that: Adopting a selective state space model, the model parameters can be dynamically adjusted according to the characteristics of the input data; B=S B (x) C=S C (x) Δ=τA(Parameter+S A (x)) In the above formula: S B (x), S C (x) represents a matrix dynamically generated based on the input x, allowing the model to adjust its behavior based on different characteristics of the input; Δ represents a learnable parameter that is used to estimate the discrete interval of input samples and enhances the handling of time dependencies.

8. The three-dimensional pore pressure prediction method based on the deep learning network model according to claim 7, characterized in that: The convolution operation is computed using the following state-space kernel, processing the input sequence in parallel to increase computation speed: Among them, h i The weight of the state-space kernel, u k-i is the input sequence.

9. The three-dimensional pore pressure prediction method based on the deep learning network model according to claim 8, characterized in that: The network model architecture consists of multiple convolutional layers, linear projection layers and dynamic state space modules. In this model, the convolution unit consists of two convolutional layers. The first convolutional layer contains 32 3x3 convolution kernels, the step size is 1, and the SiLU activation function is used; the second convolutional layer contains 64 3x3 convolution kernels, the step size is also 1, and the SiLU activation function is also used; in the dynamic state space module, the number of states is set to 64 and the sequence length is set to 100.

10. A three-dimensional pore pressure prediction device based on a deep learning network model, comprising: The data acquisition module collects well logging data and seismic data for preprocessing and constructs a data set. The data set includes a training set and a validation set. The training set includes labeled training data and unlabeled training data. The labeled training data includes relevant curve data in the well logging data, such as density, longitudinal wave velocity P, and shear wave velocity S. The unlabeled training data is the three-dimensional elastic parameter information obtained by prestack inversion of seismic data. The model building module defines the initial parameters and initial model of the model, establishes the deep learning network model structure according to the discretized state space model framework, and uses the collected logging and seismic data for iterative training to ensure that the model learns the nonlinear relationship in the input data; The model training module randomly selects training set data and validation set data in a certain ratio to train and validate the model, and continuously iterates the model by adjusting the state size and sequence length hyperparameters until the value of the loss function falls within the preset threshold range or the number of iterations reaches the preset value; The prediction module inputs the three-dimensional density, longitudinal wave velocity P, and shear wave velocity S of the three-dimensional seismic data into the trained deep learning network model, outputs the prediction results of the model, and obtains the prediction results of the three-dimensional pore pressure.

Images