A low-speed mudstone intelligent prediction method, system, device and medium

Abstract

The present invention discloses a method, system, device and medium for intelligent prediction of low-velocity mudstone, including: calibrating the target layer by combining well logging data, and dividing the layer into background mudstone, low-velocity mudstone, gas-bearing sandstone and dry-layer sandstone. Based on the division results, a large number of datasets for neural network training are generated through geological body evolution. A convolutional neural network (CNN) is constructed, and the input data is normalized to prevent gradient explosion. By training the model, an optimized neural network model is obtained. Using the AVO prestack inversion technology, the longitudinal and transverse wave velocities and density parameters are extracted from the prestack seismic data to assist in the identification of low-velocity mudstone. The processed data is input into the trained model to achieve intelligent prediction of low-velocity mudstone. The present invention can effectively improve the identification accuracy of medium-deep reservoirs, avoid misjudgment caused by the "bright spot trap" of low-velocity mudstone, and provide reliable technical support for oil and gas exploration.

Description

Technical Field

[0001] The present invention relates to the technical field of seismic data processing and interpretation in oil and gas exploration, and particularly relates to a method, system, device and medium for intelligent prediction of low-velocity mudstone based on AVO prestack inversion and geological body evolution. Background Technique

[0002] The Yinggehai Basin in the South China Sea is a Cenozoic sedimentary basin. After years of exploration and development, multiple medium and large-sized gas fields have been discovered, becoming an important offshore natural gas production area in China. However, exploration and development practices have shown that although the bright spot identification technology has achieved a high success rate in the identification of shallow reservoirs in the Yinggehai Basin, it has repeatedly encountered failures in the identification of medium and deep reservoirs. Through in-depth exploration and analysis, it is found that the widely developed low-velocity mudstone from the Huangliu Formation to the Meishan Formation in the Yinggehai Basin is the main cause of this problem.

[0003] The low-velocity mudstone is a lower-velocity mudstone layer developed under the background mudstone, which appears as a "bright spot" on the seismic section and is very similar to the seismic response of gas-bearing sandstone, easily forming a "bright spot trap" of low-velocity mudstone. The existence of this low-velocity mudstone interferes with the accurate identification of gas-bearing reservoirs by traditional bright spot identification technology, making it difficult to distinguish gas-bearing sandstone from low-velocity mudstone on the seismic section, thus increasing the difficulty and uncertainty of exploration.

[0004] In order to overcome the interference of low-velocity mudstone on the identification of gas-bearing sandstone, a method and system for intelligent prediction of low-velocity mudstone based on AVO prestack inversion and geological body evolution are proposed. By combining seismic data analysis, AVO inversion technology and geological body evolution simulation, it is possible to effectively identify the developed intervals of low-velocity mudstone, thereby improving the identification accuracy of gas-bearing sandstone and providing important technical support for exploration in the Yinggehai Basin and other similar geological conditions. Summary of the Invention

[0005] The present invention aims at the defects of the prior art and provides a method, system, device and medium for intelligent prediction of low-velocity mudstone.

[0006] In order to achieve the above invention objectives, the technical solutions adopted by the present invention are as follows:

[0007] A method for intelligent prediction of low-velocity mudstone, comprising the following steps:

[0008] Step 1: Combine well logging data analysis and well logging interpretation results to divide the target layer into intervals. The division results are used as the output Labels of the neural network, and the division types include background mudstone, low-velocity mudstone, gas-bearing sandstone and dry layer sandstone.

[0009] Step 2: According to the results of formation interval division, horizontally and vertically evolve the geological bodies to generate a large amount of data for neural network training. During the evolution of the geological bodies, wavelet analysis is carried out using well logging data to generate a pre-stack angle gather. The data includes: TWT, pre-stack angle gather, Vp, Vs, Den, and Label.

[0010] Step 3: Select the ReLU function as the activation function and construct a CNN neural network. The CNN neural network includes four convolutional modules and three fully connected layers. The first three convolutional modules each include a convolutional layer and a pooling layer, and the last convolutional module only contains a convolutional layer. The convolutional kernel sizes and strides of each module are set respectively, and max pooling processing is performed.

[0011] Step 4: Normalize the sample set to prevent gradient explosion during model training.

[0012] Step 5: Input the normalized data into the CNN network, initialize the weight vector and bias, calculate the error and perform training until the error converges, and finally store the parameter weights.

[0013] Step 6: Use pre-stack seismic data for AVO pre-stack inversion to accurately invert the P-wave and S-wave velocities and density parameters above and below the interface, and assist in the identification of low-velocity mudstone.

[0014] Step 7: Low-velocity mudstone prediction: Extract the P-wave and S-wave velocities and density data obtained from the pre-stack seismic gather and AVO inversion, intercept the data within the target layer, normalize it and then input it into the trained model to obtain the prediction results of low-velocity mudstone.

[0015] Furthermore, the basis for classification is:

[0016] Background mudstone: shale content is greater than 0.9, and the P-wave and S-wave velocities are greater than or equal to those of the upper and lower sandstone formations;

[0017] Low-velocity mudstone: shale content is greater than 0.9, and the P-wave and S-wave velocities are less than those of the upper and lower gas-bearing sandstone formations;

[0018] Gas-bearing sandstone: The well logging interpretation result is gas-bearing sandstone;

[0019] Dry layer sandstone: The well logging interpretation result is dry layer sandstone.

[0020] Furthermore, in the step of generating the dataset by evolving the geological bodies, when performing wavelet analysis using well logging data, the Ricker wavelet used is the wavelet that conforms to the post-stack seismic data of the wellside trace.

[0021] Furthermore, the convolutional modules in the CNN neural network use max pooling for the pooling layer.

[0022] Furthermore, the formula for normalization is:

[0023] X ′ = X - μ

[0024] Wherein, X is the original data and μ is the mean value.

[0025] Furthermore, the angle range of the pre-stack seismic gather is from 0° to 30°, and data for each channel is calculated every 3°.

[0026] Furthermore, the extracted pre-stack seismic gather data and the P-wave and S-wave velocity and density data obtained by AVO inversion are unified to a length of 200 through interpolation.

[0027] The present invention also discloses an intelligent prediction system for low-velocity mudstone, which can be used to implement the above-mentioned intelligent prediction method for low-velocity mudstone. Specifically, it includes:

[0028] Stratigraphic calibration module: used to divide the target layer by combining well logging data analysis and well logging interpretation results. This module can divide background mudstone, low-velocity mudstone, gas-bearing sandstone, and dry-layer sandstone, and determine the type of each stratigraphic section according to parameters such as shale content and P-wave and S-wave velocities.

[0029] Geological body evolution data generation module: used to perform horizontal and vertical changes on the geological body based on the division results of the stratigraphic calibration module, and generate a large number of data sets that conform to the laws of geological evolution. This module includes functions such as wavelet analysis, reflection coefficient calculation, pre-stack angle gather generation, and time-depth relationship calculation.

[0030] CNN neural network construction module: used to construct a convolutional neural network (CNN), including setting convolutional layers, pooling layers, and fully connected layers. This module is responsible for initializing the network structure and selecting activation functions (such as the ReLU function), and configuring the convolutional kernel size, stride, and pooling method.

[0031] Data normalization module: used to perform zero-mean processing on the input sample data to prevent the occurrence of gradient explosion during model training. This module normalizes the data according to a specific formula.

[0032] Model training module: used to input the normalized data into the CNN neural network, initialize the weight vector and bias, calculate the error, and optimize the network parameters until the error converges. This module finally outputs the trained model and related weight parameters.

[0033] AVO pre-stack inversion module: used to perform AVO pre-stack inversion based on pre-stack seismic data, and accurately invert the P-wave and S-wave velocities and density parameters above and below the interface to provide auxiliary information for the identification of low-velocity mudstone.

[0034] Low-velocity mudstone prediction module: It is used to extract relevant data from pre-stack seismic data and AVO inversion results, process and normalize it, and then input it into the trained CNN model to output the prediction result of low-velocity mudstone.

[0035] The present invention also discloses a computer device, including a memory, a processor, and a computer program stored on the memory and executable on the processor. When the processor executes the program, the above-mentioned low-velocity mudstone intelligent prediction method is implemented.

[0036] The present invention also discloses a computer-readable storage medium, on which a computer program is stored. When the program is executed by a processor, the above-mentioned low-velocity mudstone intelligent prediction method is implemented.

[0037] Compared with the prior art, the advantages of the present invention are as follows:

[0038] 1. Improve recognition accuracy: By combining AVO pre-stack inversion and geological body evolution technology, it can effectively distinguish low-velocity mudstone and gas-bearing sandstone, significantly improve the recognition accuracy of middle and deep reservoirs, and avoid misjudgment caused by the "bright spot trap" of low-velocity mudstone.

[0039] 2. Strong adaptability: It can be widely applied to various sedimentary basins with similar development characteristics of low-velocity mudstone, and has broad applicability.

[0040] 3. Data-driven and intelligent: By constructing and training a convolutional neural network (CNN), it can process a large amount of complex seismic data and geological information, realize intelligent prediction, and reduce the error caused by human subjective judgment.

[0041] 4. Reduce exploration risks: Accurately identify the boundaries of low-velocity mudstone and gas-bearing reservoirs, reduce the uncertainty in exploration, reduce the exploration risks caused by misjudgment of low-velocity mudstone, and improve the economic benefits of exploration.

[0042] 5. Automated process: It has a high degree of automation. The entire process from data processing, model training to prediction result output is automatically completed, saving a large amount of time and human resources.

[0043] 6. Adapt to complex geological conditions: Through geological body evolution simulation and AVO inversion, it can adapt to complex geological conditions, such as areas with multi-level mudstone development and complex sedimentary environments, and provide more accurate geological interpretations. Description of the Drawings

[0044] Figure 1 It is a flowchart of the low-velocity mudstone intelligent prediction method according to the embodiment of the present invention. Detailed Embodiments

[0045] To make the objectives, technical solutions, and advantages of the present invention more clear and understandable, the following provides further detailed descriptions of the present invention based on the accompanying drawings and by way of examples.

[0046] As Figure 1 shown, the present invention provides a low-speed mudstone intelligent prediction method, including the following steps:

[0047] Step 1: Formation section calibration. By combining well logging data analysis and well logging interpretation results, the target formation can be divided into sections, and the division results are the outputs (Labels) of the neural network. The division types include background mudstone, low-speed mudstone, gas-bearing sandstone, and dry-layer sandstone. The division bases are as follows: ① Background mudstone: shale content (0 - 1) is greater than 0.9, and the P-wave and S-wave velocities are greater than or equal to those of the upper and lower sandstone formations; ② Low-speed mudstone: shale content (0 - 1) is greater than 0.9, and the P-wave and S-wave velocities are less than those of the upper and lower gas-bearing sandstone formations; ③ Gas-bearing sandstone: the well logging interpretation result is gas-bearing sandstone; ④ Dry-layer sandstone: the well logging interpretation result is dry layer.

[0048] Step 2: Generation of a dataset through geological body evolution. The results of the formation section division are the geological bodies. The geological bodies are horizontally and vertically changed within a certain probability range to generate a large amount of data that meets the requirements for neural network training. During the geological body evolution process, first, wavelet analysis is performed on the well logging data to obtain a Ricker wavelet that conforms to the post-stack seismic data of the wellside trace. Then, using the transformed geological body data, the well logging data corresponding to the geological body is generated, mainly including Vp, Vs, and Den. Next, the reflection coefficient corresponding to the geological body is calculated through the Zoeppritz equation in combination with the well logging data (Vp, Vs, Den). Then, the reflection coefficient matrix and the Ricker wavelet obtained from the wavelet analysis are convolved to obtain a pre-stack angle gather. The angle range of this angle gather is 0 - 30°, and one data is calculated every 3°. Finally, there are 10 data in the angle gather. Then, the TWT data of this "new" well is calculated using the time-depth relationship and Vp. The data length is set to 200. Finally, the data obtained through the geological body transformation includes TWT, pre-stack angle gather, Vp, Vs, Den, and Label.

[0049] Step 3: Construct a CNN neural network and select the ReLU function as the activation function. This CNN neural network includes a convolutional layer and a fully connected layer. There are four convolutional modules. The first three convolutional modules each include a convolutional layer and a pooling layer. Here, max pooling (MaxPooling) is selected. The last convolutional module only contains a convolutional layer. The convolutional layer of the first convolutional module has a convolutional kernel size of 3, a stride of 1, and no padding. The pooling layer has a convolutional kernel size of 3 and a stride of 2. The convolutional layer of the second convolutional module has a convolutional kernel size of 2, a stride of 2, and no padding. The pooling layer has a convolutional kernel size of 3 and a stride of 2. The convolutional layer of the third convolutional module has a convolutional kernel size of 2, a stride of 2, and a padding length of 1. The pooling layer has a convolutional kernel size of 2 and a stride of 1. The convolutional layer of the fourth convolutional module has a convolutional kernel size of 3, a stride of 2, and no padding. The fully connected layer contains three linear layers.

[0050] Step 4: Data normalization. To prevent the problem of gradient explosion during model training, zero-mean processing is performed on the sample set. The calculation formula is:

[0051] X ′ = X - μ

[0052] In the formula, X is the original data and μ is the mean value.

[0053] Step 5: Model training. Input the normalized training set data into the CNN network, initialize the weight vector w and the bias b, calculate the error between the node output and the expected output. When the error converges, the training ends and all parameter weights are stored.

[0054] Step 5: AVO prestack inversion. Use prestack seismic data for Avoid prestack inversion. This method can accurately invert the P-wave velocity, S-wave velocity, and density parameters above and below the interface, and assist in the identification of low-velocity mudstone.

[0055] Step 6: Prediction of low-velocity mudstone. First, extract the prestack seismic gather. At the same time, extract the single-trace data at the same position from the P-wave velocity, S-wave velocity, and density volumes obtained by AVO inversion, intercept the data within the target layer, interpolate it to a length of 200, and then perform normalization processing. Input the normalized data into the trained network model to obtain the final prediction result of low-velocity mudstone.

[0056] In another embodiment of the present invention, a low-velocity mudstone intelligent prediction system is provided. This system can be used to implement the above-mentioned low-velocity mudstone intelligent prediction method. Specifically, it includes:

[0057] Interval Calibration Module: It is used to divide the target layer by combining logging data analysis and logging interpretation results. This module can divide background shale, low-velocity shale, gas-bearing sandstone, and dry-layer sandstone, and determine the type of each interval according to parameters such as shale content and P-wave and S-wave velocities.

[0058] Geological Body Evolution Data Generation Module: It is used to perform horizontal and vertical changes on the geological body based on the division results of the interval calibration module, and generate a large number of data sets that conform to the geological evolution law. This module includes functions such as wavelet analysis, reflection coefficient calculation, pre-stack angle gather generation, and time-depth relationship calculation.

[0059] CNN Neural Network Construction Module: It is used to construct a convolutional neural network (CNN), including setting convolutional layers, pooling layers, and fully connected layers. This module is responsible for initializing the network structure and selecting activation functions (such as the ReLU function), and configuring the convolutional kernel size, stride, and pooling method.

[0060] Data Normalization Module: It is used to perform zero-mean processing on the input sample data to prevent the occurrence of gradient explosion during model training. This module normalizes the data according to a specific formula.

[0061] Model Training Module: It is used to input the normalized data into the CNN neural network, initialize the weight vector and bias, calculate the error and optimize the network parameters until the error converges. This module finally outputs the trained model and related weight parameters.

[0062] AVO Pre-stack Inversion Module: It is used to perform AVO pre-stack inversion based on pre-stack seismic data, and accurately invert the P-wave and S-wave velocities and density parameters above and below the interface, providing auxiliary information for the identification of low-velocity shale.

[0063] Low-velocity Shale Prediction Module: It is used to extract relevant data from pre-stack seismic data and AVO inversion results, process and normalize it, and then input it into the trained CNN model to output the prediction results of low-velocity shale.

[0064] In another embodiment of the present invention, a terminal device is provided. The terminal device includes a processor and a memory. The memory is used to store a computer program, and the computer program includes program instructions. The processor is used to execute the program instructions stored in the computer storage medium. The processor may be a Central Processing Unit (CPU), or may also be other general-purpose processors, Digital Signal Processors (DSPs), Application Specific Integrated Circuits (ASICs), Field-Programmable Gate Arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. It is the computing core and control core of the terminal, and is suitable for implementing one or more instructions. Specifically, it is suitable for loading and executing one or more instructions to implement the corresponding method flow or corresponding function. The processor described in the embodiment of the present invention can be used for the operation of the low-speed mudstone intelligent prediction method.

[0065] In another embodiment of the present invention, a storage medium is provided, specifically a computer-readable storage medium (Memory). The computer-readable storage medium is the memory device in the terminal device and is used to store programs and data. It can be understood that the computer-readable storage medium here can include both the built-in storage medium in the terminal device and, of course, the extended storage medium supported by the terminal device. The computer-readable storage medium provides a storage space that stores the operating system of the terminal. And in this storage space, one or more instructions suitable for being loaded and executed by the processor are also stored. These instructions can be one or more computer programs (including program codes). It should be noted that the computer-readable storage medium here can be a high-speed RAM memory or a non-volatile memory, such as at least one disk memory.

[0066] One or more instructions stored in the computer-readable storage medium can be loaded and executed by the processor to implement the corresponding steps of the low-speed mudstone intelligent prediction method in the above embodiment; one or more instructions in the computer-readable storage medium are loaded and executed by the processor.

[0067] Those of ordinary skill in the art will realize that the embodiments described herein are to assist the reader in understanding the implementation methods of the present invention, and it should be understood that the scope of protection of the present invention is not limited to such specific statements and embodiments. Those of ordinary skill in the art can make various other specific deformations and combinations that do not depart from the essence of the present invention based on these technical revelations disclosed in the present invention, and these deformations and combinations are still within the scope of protection of the present invention.

Claims

1. An intelligent prediction method for low-speed mudstone, characterized in that, It includes the following steps: Step 1: Combine logging data analysis and logging interpretation results to divide the target layer into sections; the division results are used as the output Labels of the neural network, and the division types include background mudstone, low-velocity mudstone, gas-bearing sandstone, and dry-layer sandstone; Step 2: According to the results of the layer section division, horizontally and vertically evolve the geological body to generate a large amount of data for neural network training; During the evolution of the geological body, use logging data for wavelet analysis to generate a pre-stack angle gather. The data includes: TWT, pre-stack angle gather, Vp, Vs, Den, and Label; Step 3: Select the ReLU function as the activation function to construct a CNN neural network. The CNN neural network includes four convolutional modules and three fully connected layers; the first three convolutional modules include a convolutional layer and a pooling layer, and the last convolutional module only contains a convolutional layer; the convolutional kernel size and stride of each module are set respectively, and max pooling processing is performed; Step 4: Normalize the sample set to prevent gradient explosion during model training; Step 5: Input the normalized data into the CNN network, initialize the weight vector and bias, calculate the error and perform training until the error converges, and finally store the parameter weights; Step 6: Use pre-stack seismic data for AVO pre-stack inversion to accurately invert the P-wave and S-wave velocities and density parameters above and below the interface to assist in the identification of low-velocity mudstone; Step 7: Low-velocity mudstone prediction: Extract the P-wave and S-wave velocities and density data obtained from the pre-stack seismic gather and AVO inversion, intercept the data within the target layer, normalize it, and then input it into the trained model to obtain the prediction results of low-velocity mudstone.

2. The intelligent prediction method for low-speed mudstone according to claim 1, characterized in that: The basis for the division types is: Background mudstone: The shale content is greater than 0.9, and the P-wave and S-wave velocities are greater than or equal to those of the upper and lower sandstone layers; Low-velocity mudstone: The shale content is greater than 0.9, and the P-wave and S-wave velocities are less than those of the upper and lower gas-bearing sandstone layers; Gas-bearing sandstone: The logging interpretation result is gas-bearing sandstone; Dry-layer sandstone: The logging interpretation result is dry-layer sandstone.

3. The intelligent prediction method for low-speed mudstone according to claim 1, wherein: When using logging data for wavelet analysis, the Ricker wavelet used is the wavelet that conforms to the post-stack seismic data of the well-side trace.

4. The intelligent prediction method for low-speed mudstone according to claim 1, wherein: The convolutional module in the CNN neural network uses max pooling for the pooling layer.

5. The intelligent prediction method for low-speed mudstone according to claim 1, wherein: The formula for normalization processing is: X ′ = X - μ In the formula, X is the original data, and μ is the mean value.

6. The intelligent prediction method for low-speed mudstone according to claim 1, wherein: The angle range of the pre-stack seismic gather is from 0° to 30°, and data is calculated every 3°.

7. The intelligent prediction method for low-speed mudstone according to claim 1, wherein: The extracted P-wave and S-wave velocities and density data obtained from the pre-stack seismic gather and AVO inversion are unified to a length of 200 through interpolation.

8. An intelligent prediction system for low-velocity mudstone, which can be used to implement an intelligent prediction method for low-velocity mudstone according to any one of claims 1 to 7. Specifically, it includes: Layer calibration module: used to combine logging data analysis and logging interpretation results to divide the target layer into sections; this module divides background mudstone, low-velocity mudstone, gas-bearing sandstone, and dry-layer sandstone, and determines the type of each layer section according to the parameters of shale content and P-wave and S-wave velocities; Geological body evolution data generation module: It is used to generate a large number of data sets that conform to the geological evolution law by changing the geological body horizontally and vertically based on the division result of the stratigraphic calibration module; this module includes functions such as wavelet analysis, reflection coefficient calculation, pre-stack angle gather generation, and time-depth relationship calculation; CNN neural network construction module: It is used to construct a convolutional neural network CNN, including setting convolutional layers, pooling layers, and fully connected layers; this module is responsible for initializing the network structure, selecting activation functions, and configuring the convolutional kernel size, stride, and pooling method; Data normalization module: It is used to perform zero-mean processing on the input sample data to prevent the phenomenon of gradient explosion during model training; this module normalizes the data according to a specific formula; Model training module: It is used to input the normalized data into the CNN neural network, initialize the weight vector and bias, calculate the error, and optimize the network parameters until the error converges; This module finally outputs the trained model and related weight parameters; AVO pre-stack inversion module: It is used to perform AVO pre-stack inversion based on pre-stack seismic data, and accurately invert the P-wave and S-wave velocities and density parameters above and below the interface, providing auxiliary information for the identification of low-velocity mudstone; Low-velocity mudstone prediction module: It is used to extract relevant data from pre-stack seismic data and AVO inversion results, process and normalize them, and then input them into the trained CNN model to output the prediction results of low-velocity mudstone.

9. A computer device, characterized in that: It includes a memory, a processor, and a computer program stored on the memory and running on the processor. When the processor executes the program, it implements the low-velocity mudstone intelligent prediction method described in any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that: A computer program is stored thereon. When the program is executed by the processor, it implements the low-velocity mudstone intelligent prediction method described in any one of claims 1 to 7.

Images