A pre-stack intelligent enhanced detection method for multi-scale fractures and small faults in shale gas

Through multi-scale analysis and artificial intelligence intelligent enhanced detection, the problem that traditional seismic exploration technology is difficult to detect small-scale cracks and faults in shale gas reservoirs is solved, and efficient and accurate exploration results are achieved, reducing costs and improving exploration accuracy and reliability.

CN117991370BActive Publication Date: 2024-11-29成都星辉科瑞科技有限责任公司
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202410264887.9
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2024-03-08
Publication Date
2024-11-29
Estimated Expiration
2044-03-08

AI Technical Summary

Technical Problem

Traditional seismic exploration technology is difficult to accurately detect small-scale cracks and faults in shale gas reservoirs, which are affected by resolution limitations, noise interference and complex geological structures, resulting in low exploration efficiency and accuracy.

Method used

Using multi-scale analysis, prestack data processing and artificial intelligence intelligent enhanced detection methods, data is collected through seismic exploration technology, denoising, silence correction and superposition processing is performed, and feature information is extracted using wavelet transform or Fourier transform, combined with deep learning models Carry out intelligent identification and classification to achieve accurate detection of small-scale cracks and faults.

Benefits of technology

Improve the efficiency and accuracy of shale gas reservoir exploration, can quickly and accurately detect small-scale cracks and faults, reduce exploration costs and data processing difficulties, provide more intuitive geological structure information, suitable for various types of shale Gas reservoir.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN117991370B_ABST
    Figure CN117991370B_ABST
Patent Text Reader

Abstract

The present invention belongs to the field of survey technology. The present invention discloses a pre-stack intelligent enhanced detection method for multi-scale fractures and small faults in shale gas, comprising the following steps: data acquisition: using seismic exploration technology to collect data on shale gas reservoirs and obtain seismic data; data processing: pre-stack processing of the collected seismic data, including denoising, static correction and stacking, to improve the quality and resolution of the data; multi-scale analysis: analyzing the processed seismic data to extract characteristic information at different scales, including characteristic information of small-scale fractures and faults; intelligent enhanced detection: using artificial intelligence technology to intelligently identify and enhance the multi-scale analysis results, accurately detect small-scale fractures and faults, and obtain detection results; result output: visually outputting the intelligent enhanced detection results. The present invention realizes rapid and accurate detection of small-scale fractures and faults in shale gas reservoirs.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention belongs to the field of surveying technology, and in particular relates to a pre-stack intelligent enhanced detection method for multi-scale fractures and small faults in shale gas. Background Art

[0002] In traditional geophysical exploration, the exploration of shale gas reservoirs mainly relies on seismic exploration technology. Seismic exploration uses artificially excited seismic waves to propagate in the strata. When encountering the interface between different media, reflection or refraction will occur. These signals are received by seismic detectors, and after processing and analysis, the underground geological structure and lithology distribution can be inferred. However, shale gas reservoirs have the characteristics of strong heterogeneity and complex fracture development, which brings great challenges to exploration. Traditional seismic exploration methods often have difficulty in accurately detecting small-scale fractures and faults, resulting in low exploration efficiency and accuracy.

[0003] At present, there are many problems in the exploration of shale gas reservoirs: Resolution limitation: Traditional seismic exploration technology is limited by resolution and it is difficult to accurately detect small-scale cracks and faults. Due to the propagation characteristics of seismic waves, it is difficult to obtain accurate reflection signals for scales smaller than the seismic wavelength. Noise interference: Seismic exploration is easily affected by environmental noise and human interference. These noises may mask the real geological signals and lead to misjudgment or omission. Complex geological structure: There are a large number of complex geological structures such as cracks, faults and holes in shale gas reservoirs. The morphology and distribution of these structures have an important impact on the production and development of shale gas. However, traditional seismic exploration technology is difficult to accurately identify these complex geological structures. Data processing difficulty: The amount of data obtained from seismic exploration is huge and difficult to process. How to effectively extract useful information, reduce the difficulty of data processing, and improve exploration efficiency is an urgent problem to be solved. Summary of the invention

[0004] In order to solve the above problems, the present invention proposes a shale gas multi-scale fracture and small fault pre-stack high-resolution intelligent enhanced detection technology, which uses artificial intelligence and geophysical exploration technology to achieve rapid and accurate detection of small-scale fractures and faults in shale gas reservoirs. Through multi-scale analysis, pre-stack data processing and intelligent enhanced detection, the exploration efficiency and accuracy of shale gas reservoirs are improved.

[0005] To achieve the above object, the technical solution adopted by the present invention is: a shale gas multi-scale fracture and small fault pre-stack intelligent enhanced detection method, comprising the steps of:

[0006] S10, data acquisition: using seismic exploration technology to collect data on shale gas reservoirs and obtain seismic data;

[0007] S20, data processing: pre-stack processing of the acquired seismic data, including denoising, static correction and stacking, to improve the quality and resolution of the data;

[0008] S30, multi-scale analysis: Analyze the processed seismic data to extract characteristic information at different scales, including characteristic information of small-scale fractures and faults;

[0009] S40, intelligent enhanced detection: using artificial intelligence technology to intelligently identify and enhance multi-scale analysis results, accurately detect small-scale cracks and faults, and obtain detection results;

[0010] S50, result output: visually output the intelligent enhanced detection result.

[0011] Furthermore, when data is collected, an observation system is arranged on the surface of the earth using an array of seismic detectors through seismic exploration technology to stimulate seismic waves and receive seismic data;

[0012] Seismometers are buried underground to capture information about seismic waves traveling through the earth's strata.

[0013] Furthermore, when data processing is performed, it includes preprocessing, static correction, denoising and stacking;

[0014] Preprocessing includes: data format conversion and data sorting;

[0015] Static correction includes: eliminating signal distortion caused by surface topography and surface geological factors;

[0016] Denoising includes: eliminating background noise and interfering signals;

[0017] Superposition includes: combining data from multiple seismic channels to improve signal strength and signal-to-noise ratio.

[0018] Furthermore, when performing multi-scale analysis: the processed seismic data is subjected to multi-scale analysis to extract characteristic information at different scales;

[0019] Wavelet transform or Fourier transform method is used for multi-scale decomposition to decompose seismic data into components of different frequencies and spatial scales to highlight the characteristic information of different scales.

[0020] Furthermore, a wavelet transform method is used to perform multi-scale decomposition to decompose the seismic data into components of different frequencies and different spatial scales, including the steps of:

[0021] Determine the wavelet basis function: select the appropriate wavelet basis function;

[0022] Scale discretization: discretize the scale of wavelet transform;

[0023] Signal decomposition: Input the original seismic data into the wavelet transform algorithm, and perform multi-scale decomposition on the signal according to the determined scale and wavelet basis function; at each scale, the signal is decomposed into two parts: approximate and detailed. The approximate part represents the smooth part of the signal, and the detailed part represents the sudden change part of the signal.

[0024] Feature extraction: extract features from approximate and detailed parts at each scale;

[0025] Feature fusion: Fuse features of different scales to obtain more comprehensive information;

[0026] Result output: The final feature vector is used as the representation of seismic data for subsequent classification and recognition tasks.

[0027] Furthermore, a Fourier transform method is used to perform multi-scale decomposition to decompose the seismic data into components of different frequencies and different spatial scales, including the steps of:

[0028] Determine the window function: select an appropriate window function;

[0029] Signal segmentation: Divide the original seismic data into several segments of equal or approximately equal length; each segment is called a frame, which is used for Fourier transform;

[0030] Fourier transform: Perform a Fourier transform on each frame to convert it from the time domain to the frequency domain; the Fourier transform represents the signal as a superposition of different frequency components, thereby extracting the characteristics of different frequencies;

[0031] Spectrum analysis: Analyze the spectrum of each frame and extract the amplitude and phase information of different frequency components;

[0032] Feature extraction: extract features from the spectrum of each frame;

[0033] Feature fusion: fuse the features of different frames to obtain more comprehensive information;

[0034] Result output: The final feature vector is used as the representation of seismic data for subsequent classification and recognition tasks.

[0035] Furthermore, when performing intelligent enhanced detection: artificial intelligence technology is used to intelligently identify and enhance the results of multi-scale analysis;

[0036] A deep learning method is used to build a model to classify and identify features after multi-scale analysis to detect small-scale cracks and faults;

[0037] Among them, the existing training data set is used for model training to improve the detection accuracy.

[0038] Furthermore, deep learning is used to classify and identify the features after multi-scale analysis to detect small-scale cracks and faults, including the steps of:

[0039] Data collection: Collect a large amount of multi-scale fracture and fault image data with annotated information for training and validating deep learning models;

[0040] Multi-scale feature extraction: Use image processing technology to perform multi-scale analysis on the collected image data;.

[0041] Deep learning model construction: Build deep learning models;

[0042] Model training: Use the extracted multi-scale features and corresponding labels to train the deep learning model;

[0043] Model evaluation and optimization: Evaluate the performance of the model on the validation set, and optimize and adjust the model based on the evaluation results.

[0044] Furthermore, the results of intelligent enhanced detection are output in a visualized manner, and the visualization uses various graphics and image technologies to display the distribution and morphological information of the detected cracks and faults.

[0045] The beneficial effects of adopting this technical solution are:

[0046] The technology of the present invention utilizes artificial intelligence and geophysical exploration technology to achieve rapid and accurate detection of small-scale fractures and faults in shale gas reservoirs. Through methods such as multi-scale analysis, pre-stack data processing and intelligent enhanced detection, the exploration efficiency and accuracy of shale gas reservoirs are improved. The present invention can achieve rapid and accurate detection of small-scale fractures and faults in shale gas reservoirs, improving the exploration efficiency and accuracy. This technology combines geophysical exploration technology and artificial intelligence technology, and through multi-scale analysis and intelligent enhanced detection, it can better reveal the geological structure characteristics of the underground and provide more accurate information for shale gas exploration and development. At the same time, the technical solution of the present invention has high practicality and operability, and can be widely used in actual production.

[0047] The present invention can improve detection accuracy: through multi-scale analysis and intelligent enhanced detection, small-scale cracks and faults in shale gas reservoirs can be detected more accurately, thereby improving the accuracy and reliability of exploration.

[0048] The present invention can reduce exploration costs: the technology can quickly and accurately identify small-scale fractures and faults in shale gas reservoirs, reducing the large amount of tedious geological surveys and drilling verification work required in traditional exploration methods, thereby reducing exploration costs.

[0049] The present invention can improve exploration efficiency: the technology combines geophysical exploration technology and artificial intelligence technology to realize automated and intelligent data processing and analysis, greatly improving exploration efficiency.

[0050] The present invention can output visually: the result output of this technology adopts visualization technology, which can intuitively display the distribution, morphology and other information of the detected cracks and faults, providing a more intuitive basis for subsequent exploration and development.

[0051] The invention has wide applicability: the technology is applicable to various types of shale gas reservoirs and has wide applicability. At the same time, the technology can also be promoted and applied to other types of oil and gas exploration and development fields.

[0052] In summary, the pre-stack high-resolution intelligent enhanced detection method for multi-scale fractures and small faults in shale gas has the advantages of high efficiency, accuracy, and low cost. It provides important technical support for shale gas exploration and development and helps promote the sustainable development of the oil and gas industry. BRIEF DESCRIPTION OF THE DRAWINGS

[0053] Figure 1 It is a schematic diagram of a process flow of a shale gas multi-scale fracture and small fault pre-stack intelligent enhanced detection method of the present invention;

[0054] Figure 2 Schematic diagram of a process of performing multi-scale decomposition using a wavelet transform method in an embodiment of the present invention;

[0055] Figure 3 Schematic diagram of the process of multi-scale decomposition using the Fourier transform method in an embodiment of the present invention. DETAILED DESCRIPTION

[0056] In order to make the purpose, technical solutions and advantages of the present invention more clear, the present invention is further described below with reference to the accompanying drawings.

[0057] In this embodiment, see Figure 1 As shown, the present invention proposes a shale gas multi-scale fracture and small fault pre-stack intelligent enhanced detection method, comprising the steps of:

[0058] S10, data acquisition: using seismic exploration technology to collect data on shale gas reservoirs and obtain seismic data;

[0059] S20, data processing: pre-stack processing of the acquired seismic data, including denoising, static correction and stacking, to improve the quality and resolution of the data;

[0060] S30, multi-scale analysis: Analyze the processed seismic data to extract characteristic information at different scales, including characteristic information of small-scale fractures and faults;

[0061] S40, intelligent enhanced detection: using artificial intelligence technology to intelligently identify and enhance multi-scale analysis results, accurately detect small-scale cracks and faults, and obtain detection results;

[0062] S50, result output: visually output the intelligent enhanced detection result.

[0063] As an optimization solution of the above embodiment, when data is collected, an observation system is arranged on the surface of the earth using a seismic exploration technology and a seismic detector array to stimulate seismic waves and receive seismic data;

[0064] Seismometers are buried underground to capture information about seismic waves traveling through the earth's strata.

[0065] When performing data processing, it includes preprocessing, static correction, denoising and stacking;

[0066] Preprocessing includes: data format conversion and data sorting;

[0067] Static correction includes: eliminating signal distortion caused by surface topography and surface geological factors;

[0068] Denoising includes: eliminating background noise and interfering signals;

[0069] Superposition includes: combining data from multiple seismic channels to improve signal strength and signal-to-noise ratio.

[0070] As an optimization solution of the above embodiment, when performing multi-scale analysis: performing multi-scale analysis on the processed seismic data to extract characteristic information at different scales;

[0071] Wavelet transform or Fourier transform method is used for multi-scale decomposition to decompose seismic data into components of different frequencies and spatial scales to highlight the characteristic information of different scales.

[0072] Solution 1, such as Figure 2 As shown, the wavelet transform method is used for multi-scale decomposition to decompose the seismic data into components of different frequencies and different spatial scales, including the following steps:

[0073] Determine the wavelet basis function: Select a suitable wavelet basis function, such as Haar wavelet, Daubechies wavelet, etc. These wavelet basis functions have different shapes and characteristics and can be selected according to actual needs.

[0074] Scale discretization: discretize the scale of wavelet transform; usually the scale parameter a is binary discretized, and the b parameter can take any value. In this way, the wavelet transform can be applied to different scales, thereby extracting features of different scales.

[0075] Signal decomposition: Input the original seismic data into the wavelet transform algorithm, and perform multi-scale decomposition on the signal according to the determined scale and wavelet basis function; at each scale, the signal is decomposed into two parts: approximate and detailed. The approximate part represents the smooth part of the signal, and the detailed part represents the sudden change part of the signal.

[0076] Feature extraction: Extract features from the approximate and detailed parts at each scale; these features can include information such as the size, shape, and direction of the scale coefficient. Analyzing the performance and change patterns of these features at different scales can reveal the inherent structure and properties of seismic data.

[0077] Feature fusion: Features of different scales are combined to obtain more comprehensive information. Feature fusion can be performed by simple superposition or weighted averaging.

[0078] Result output: The final feature vector is used as the representation of seismic data for subsequent classification and recognition tasks.

[0079] Solution 2, such as Figure 3 As shown, the Fourier transform method is used for multi-scale decomposition to decompose seismic data into components of different frequencies and different spatial scales, including the following steps:

[0080] Determine the window function: Select a suitable window function, such as Hanning window, Hamming window, etc. The function of the window function is to limit the spectrum of the signal so that the signal has different lengths and shapes in the time domain.

[0081] Signal segmentation: Divide the original seismic data into several segments of equal or approximately equal length; each segment is called a frame, which is used for Fourier transform;

[0082] Fourier transform: Perform a Fourier transform on each frame to convert it from the time domain to the frequency domain; the Fourier transform represents the signal as a superposition of different frequency components, thereby extracting the characteristics of different frequencies;

[0083] Spectral analysis: Analyze the spectrum of each frame and extract the amplitude and phase information of different frequency components; this information can be used for subsequent feature extraction and classification and recognition tasks.

[0084] Feature extraction: Extract features from the spectrum of each frame; these features can include center frequency, bandwidth, energy, etc. Analyzing the performance and variation of these features at different scales can reveal the inherent structure and properties of seismic data.

[0085] Feature fusion: Fuse the features of different frames to obtain more comprehensive information; feature fusion can be performed using simple averaging, weighted averaging and other methods.

[0086] Result output: The final feature vector is used as the representation of seismic data for subsequent classification and recognition tasks.

[0087] As an optimization solution of the above embodiment, when performing intelligent enhanced detection: using artificial intelligence technology to intelligently identify and enhance the results of multi-scale analysis;

[0088] A deep learning method is used to build a model to classify and identify features after multi-scale analysis to detect small-scale cracks and faults;

[0089] Among them, the existing training data set is used for model training to improve the detection accuracy.

[0090] Specifically, deep learning is used to classify and identify the features after multi-scale analysis to detect small-scale cracks and faults, including the following steps:

[0091] Data collection: Collect a large amount of multi-scale crack and fault image data with annotated information for training and validating deep learning models; these data should include crack and fault images at different scales, angles, and lighting conditions.

[0092] Multi-scale feature extraction: Use image processing technology to perform multi-scale analysis on the collected image data; use pyramid, wavelet transform and other methods to extract features at different scales; these features should include edges, textures, shapes, etc.

[0093] Deep learning model construction: Build a deep learning model; you can use convolutional neural network CNN, recurrent neural network RNN ​​or generative adversarial network GAN, etc. The specific model structure can be adjusted according to actual needs. For example, you can use pre-trained models such as residual network ResNet and U-Net for fine-tuning.

[0094] Model training: Use the extracted multi-scale features and corresponding labels to train the deep learning model; supervised learning methods can be used to optimize the model's parameters so that the model can better learn and identify the characteristics of cracks and faults.

[0095] Model evaluation and optimization: Evaluate the performance of the model on the validation set, using indicators such as accuracy, recall, and F1 value. Based on the evaluation results, optimize and adjust the model to improve the classification and recognition capabilities of the model.

[0096] As an optimization solution for the above embodiment, the results of intelligent enhanced detection are output visually. The visualization uses various graphics and image technologies, such as contour maps, color profiles, 3D imaging, etc., to display the distribution and morphological information of the detected cracks and faults.

[0097] The above shows and describes the basic principles and main features of the present invention and the advantages of the present invention. It should be understood by those skilled in the art that the present invention is not limited to the above embodiments. The above embodiments and descriptions are only for explaining the principles of the present invention. Without departing from the spirit and scope of the present invention, the present invention may have various changes and improvements, which fall within the scope of the present invention to be protected. The scope of protection of the present invention is defined by the attached claims and their equivalents.

Claims

1. A pre-stack intelligent enhanced detection method for multi-scale fractures and small faults in shale gas, characterized in that: Includes steps: S10, data acquisition: using seismic exploration technology to collect data from shale gas reservoirs and obtain seismic data; when collecting data, using seismic exploration technology, using a seismic detector array, arranging an observation system on the surface, exciting seismic waves and receiving seismic data; seismic detectors are buried underground to capture the propagation information of seismic waves in the strata; S20, data processing: pre-stack processing of the acquired seismic data, including denoising, static correction and stacking, to improve the quality and resolution of the data; When performing data processing, it includes preprocessing, static correction, denoising and stacking; Preprocessing includes: data format conversion and data sorting; Static correction includes: eliminating signal distortion caused by surface topography and surface geological factors; Denoising includes: eliminating background noise and interfering signals; Superposition includes: combining data from multiple seismic channels to increase signal strength and signal-to-noise ratio; S30, multi-scale analysis: Analyze the processed seismic data to extract characteristic information at different scales, including characteristic information of small-scale fractures and faults; When conducting multi-scale analysis: conduct multi-scale analysis on the processed seismic data to extract characteristic information at different scales; Use wavelet transform or Fourier transform method to perform multi-scale decomposition, decomposing seismic data into components of different frequencies and spatial scales to highlight characteristic information of different scales; The wavelet transform method is used to perform multi-scale decomposition to decompose seismic data into components of different frequencies and spatial scales, including the following steps: Determine the wavelet basis function: select the appropriate wavelet basis function; Scale discretization: discretize the scale of wavelet transform; Signal decomposition: Input the original seismic data into the wavelet transform algorithm, and perform multi-scale decomposition on the signal according to the determined scale and wavelet basis function; at each scale, the signal is decomposed into two parts: approximate and detailed. The approximate part represents the smooth part of the signal, and the detailed part represents the sudden change part of the signal. Feature extraction: extract features from approximate and detailed parts at each scale; Feature fusion: Fusion of features at different scales; Result output: The final feature vector is used as the representation of seismic data for subsequent classification and recognition tasks; Or use the Fourier transform method to perform multi-scale decomposition to decompose the seismic data into components of different frequencies and different spatial scales, including the steps of: Determine the window function: select an appropriate window function; Signal segmentation: Divide the original seismic data into several segments of equal or approximately equal length; each segment is called a frame, which is used for Fourier transform; Fourier transform: Perform a Fourier transform on each frame to convert it from the time domain to the frequency domain; the Fourier transform represents the signal as a superposition of different frequency components, thereby extracting the characteristics of different frequencies; Spectrum analysis: Analyze the spectrum of each frame and extract the amplitude and phase information of different frequency components; Feature extraction: extract features from the spectrum of each frame; Feature fusion: fuse the features of different frames; Result output: The final feature vector is used as the representation of seismic data for subsequent classification and recognition tasks; S40, intelligent enhanced detection: using artificial intelligence technology to intelligently identify and enhance multi-scale analysis results, accurately detect small-scale cracks and faults, and obtain detection results; S50, result output: visually output the intelligent enhanced detection result.

2. The method for pre-stack intelligent enhanced detection of multi-scale fractures and small faults in shale gas according to claim 1, characterized in that: When performing intelligent enhanced detection: use artificial intelligence technology to intelligently identify and enhance the results of multi-scale analysis; A deep learning method is used to build a model to classify and identify features after multi-scale analysis to detect small-scale cracks and faults; Among them, the existing training data set is used for model training to improve the detection accuracy.

3. A shale gas multi-scale fracture and small fault pre-stack intelligent enhanced detection method according to claim 2, characterized in that: Deep learning is used to classify and identify the features after multi-scale analysis to detect small-scale cracks and faults, including the following steps: Data collection: Collect a large amount of multi-scale fracture and fault image data with annotated information for training and validating deep learning models; Multi-scale feature extraction: Use image processing technology to perform multi-scale analysis on the collected image data; Deep learning model construction: Build deep learning models; Model training: Use the extracted multi-scale features and corresponding labels to train the deep learning model; Model evaluation and optimization: Evaluate the performance of the model on the validation set, and optimize and adjust the model based on the evaluation results.

4. The method for pre-stack intelligent enhanced detection of multi-scale fractures and small faults in shale gas according to claim 1, characterized in that: The results of intelligent enhanced detection are output visually, and the visualization uses various graphics and image technologies to display the distribution and morphological information of the detected cracks and faults.