A method and apparatus for identifying faults based on differential cepstrum

Abstract

The present application belongs to the technical field of oil and gas exploration seismic data interpretation. The present application discloses a method and device for identifying faults based on differential cepstrum. The method performs differential cepstrum processing on seismic data of a target layer section, extracts first-order differential cepstrum coefficients of the differential cepstrum, sets a processing factor to calculate eigenvalues of a covariance matrix of the first-order coefficient data body of the differential cepstrum, and generates a high-resolution fault information data body. The differential cepstrum operation adopted by the method strengthens the weak information singularity feature, can identify small faults, and is beneficial to improving the accuracy and precision of fault identification of seismic data. The device for identifying faults based on differential cepstrum provided by the present application comprises a Fourier transform spectrum generator, a logarithmic spectrum generator, a Fourier inverse transform processor, a differential cepstrum first-order coefficient data body generator, and a fault data generator. The device realizes the method for identifying faults based on differential cepstrum.

Description

Technical Field

[0001] This invention relates to the field of seismic data interpretation technology for oil and gas exploration, specifically to a method and apparatus for identifying faults based on differential cepstrum. Background Technology

[0002] Fault identification using seismic data is crucial for defining hydrocarbon reservoir boundaries, understanding hydrocarbon migration processes, analyzing reservoir formation and distribution, and is key to predicting fracture-vuggy reservoirs. Currently, fault identification methods primarily utilize coherence attributes, curvature attributes, frequency division processing, ant tracking techniques, and machine learning. However, coherence attributes, curvature attributes, and frequency division processing are less effective at identifying small faults and strike-slip faults. Furthermore, coherence and curvature attributes are sensitive to noise, resulting in low accuracy. Frequency division processing mainly targets dominant frequency bands, easily missing fault information in some bands. Ant tracking techniques rely heavily on existing geological databases; incomplete databases yield poor results. Machine learning for fault identification often requires extensive labeled data for training, typically using a training set combining model data with a small portion of real-world data, lacking adaptability to complex real-world regions. Additionally, machine learning models often suffer from poor generalization.

[0003] The purpose of this invention is to overcome the shortcomings of traditional seismic data fault identification methods. Based on the differential cepstral algorithm, a new seismic data fault identification method and device are provided. This method is beneficial to improving the accuracy of fault identification and clearly depicting the internal details of the fault zone. Summary of the Invention

[0004] A method for identifying tomography based on differential cepstrum includes the following steps:

[0005] (1) Perform differential cepstral processing on the seismic data of the target section one channel at a time to obtain the differential cepstral data volume of the seismic data of the target section;

[0006] (2) Extract the first-order differential cepstral coefficients of the differential cepstral data for each trace of the seismic data of the target section, and generate the first-order coefficient data of the differential cepstral.

[0007] (3) For the extracted first-order coefficient data volume of differential cepstrum, set the processing factor, calculate the eigenvalues ​​of the covariance matrix, and generate the fault information data volume.

[0008] The core issue of this invention is to transform seismic data into the differential cepstral domain through differential cepstral operations, extract the eigenvalues ​​of the covariance matrix of the first-order differential cepstral coefficient data volume, and generate high-resolution seismic data fault information.

[0009] The specific implementation principle of this invention is as follows:

[0010] 1. Perform differential cepstral processing on each channel of the seismic data of the target segment to obtain the differential cepstral data volume of the seismic data of the target segment.

[0011] 1.1 Zero-padding, framing, and windowing are performed on each trace of the seismic data for the target layer;

[0012] For each seismic data of the target segment The following zero-padding method is used:

[0013] (1)

[0014] in, Represents the zero matrix. It is the length of the window function, and it is an even number;

[0015] right Perform frame division, with each frame having a length of [length missing]. The frame shift is 1, and a Hamming window is added to each frame of the seismic signal;

[0016] 1.2 Perform differential cepstral calculation on each windowed seismic signal frame to obtain the differential cepstral data volume of the seismic data of the target section;

[0017] To better suit the characteristics of seismic data, the formula for calculating differential cepstral spectrum is defined as follows:

[0018] (2)

[0019] In the formula, This is a windowed seismic signal frame. and These represent the forward and inverse Fourier transforms, respectively. For logarithmic operators, For frequency. For a windowed seismic signal frame. The spectrum is calculated using Fourier transform. The absolute value of the spectrum is taken, and then 1 is added. The logarithmic spectrum is calculated using logarithmic operators. The differential value of the logarithmic spectrum is then calculated, and the cepstrum is obtained using inverse Fourier transform. The differential cepstrum is calculated frame by frame to obtain the differential cepstrum data volume of the seismic data of the target section.

[0020] The "1" operation here is to avoid the situation where taking the logarithm of zero is meaningless when calculating the logarithm spectrum. At the same time, this operation does not change the shape distribution of the seismic data results.

[0021] 2. For the differential cepstral data volume of the seismic data of the target layer, extract the first-order differential cepstral coefficients of the differential cepstral one channel at a time to generate the first-order coefficient data volume of the differential cepstral.

[0022] For the seismic data of the target section, extract the differential cepstrum for each frame on a trace-by-trace basis. The first point in the differential cepstral spectrum of each frame in each seismic trace constitutes the first-order differential cepstral coefficients of that seismic trace. The first-order differential cepstral coefficient data volume of the target segment seismic data is generated from the first-order differential cepstral coefficients of all seismic traces, denoted as . .

[0023] 3. For the extracted first-order coefficient data volume of differential cepstrum, set the processing factor, calculate the eigenvalues ​​of the covariance matrix, and generate the fault information data volume.

[0024] The extracted first-order coefficient data volume of the differential cepstral Set processing factor For the adjacent first and the Group First-order coefficient data volume of differential cepstral and The frames are divided into segments, with a frame length of [missing information]. Frame shift is The adjacent first and the Group Each frame of the first-order coefficient data volume of the differential cepstral is denoted as follows: and ;

[0025] Calculate the covariance matrix of each frame in the first-order coefficient data volume of two adjacent sets of differential cepstrum. The following formula is used for calculation:

[0026]

[0027] (3)

[0028] Calculate the covariance matrix The characteristic values ​​are used to generate fault information data volumes for the target seismic segment.

[0029] The present invention provides a method for identifying tomography based on differential cepstrum, which has the following characteristics, mainly manifested as follows:

[0030] (1) This technology uses differential cepstral operation to transform seismic data into differential cepstral domain processing to improve the accuracy and precision of fault identification of seismic data.

[0031] (2) The differential cepstral operation used in this technology strengthens the weak information singularity feature by using a logarithmic operator in the middle of the algorithm, thus enabling the identification of small faults.

[0032] (3) This algorithm can detect multiple amplitude singular values ​​at different scales at the same time, and is suitable for simultaneous detection of multiple faults at different scales in multiple seismic data segments.

[0033] (4) This algorithm runs fast and is suitable for processing large batches of seismic signals.

[0034] This invention also provides a device for identifying tomography based on differential cepstrum, comprising a Fourier transform spectrum generator, a logarithmic spectrum generator, an inverse Fourier transform processor, a differential cepstrum first-order coefficient data volume generator, and a tomography data generator. The Fourier transform spectrum generator, logarithmic spectrum generator, and inverse Fourier transform processor implement differential cepstrum operations. The logarithmic spectrum generator performs the absolute value of the input spectrum, adds "1", and then calculates the logarithm. The differential cepstrum first-order coefficient data volume generator performs a channel-by-channel, frame-by-frame operation to calculate the first-order differential cepstrum coefficients from the differential cepstrum data volume generated by the inverse Fourier transform processor, generating the differential cepstrum first-order coefficient data volume. The tomography data generator performs a channel-by-channel, frame-by-frame calculation to obtain the eigenvalues ​​of the covariance matrix of the differential cepstrum first-order coefficient data volume, generating the tomography information data volume. This device implements the aforementioned method for identifying tomography based on differential cepstrum. Attached Figure Description

[0035] Figure 1 Here is a flowchart of a method for identifying tomography based on differential cepstral spectra.

[0036] Figure 2 Differential cepstrum for amplitude singularity detection map

[0037] Figure 3 This is a profile of post-stack migration seismic data for a gas field.

[0038] Figure 4 This is a cross-sectional view of fault identification after conventional coherence processing.

[0039] Figure 5 This is a fault identification profile after processing using this technique.

[0040] Figure 6 This invention relates to a device for identifying tomography based on differential cepstrum. Detailed Implementation

[0041] (1) Figure 1 This is a flowchart of a method for identifying faults based on differential cepstrum.

[0042] (2) Figure 2Figure 1 shows the detection of amplitude singularities using differential cepstrum. Figure 2 shows a simulated signal with a single amplitude singularity, and Figure 3 shows the result of differential cepstrum calculation. It can be seen from the figure that a maximum value exists at the amplitude discontinuity. Figure 4 shows a simulated signal with multiple amplitude singularities, and Figure 5 shows the result of differential cepstrum calculation. It can be seen from the figure that a maximum value exists at all amplitude discontinuities. This indicates that differential cepstrum can detect multiple amplitude singularities simultaneously.

[0043] (3) Figure 3 This is a back-stack migrated seismic data profile of a gas field. The data sampling frequency is 2ms.

[0044] (4) Figure 4 This is a fault identification profile after conventional coherence processing. As can be seen from the image, information on some large faults is clearly displayed.

[0045] (5) This is a fault identification profile after processing with this technology. As can be seen from the figure, compared with the fault identification profile after conventional coherent processing, the large fault information identified by this technology is more obvious and continuous, and more faults are detected at the same time. Some small fault information is also clearly detected.

[0046] (6) Figure 6 This invention discloses a device for identifying tomography based on differential cepstrum. It includes a Fourier transform spectrum generator, a logarithmic spectrum generator, an inverse Fourier transform processor, a generator for generating first-order coefficient data of the differential cepstrum, and a tomographic data generator. The Fourier transform spectrum generator, logarithmic spectrum generator, and inverse Fourier transform processor perform differential cepstrum operations. Both the generator and processor include data processing and data storage components, which communicate with each other via an internal bus. The data processing components of the generator and processor can be programmable logic devices such as FPGAs and microprocessors such as MCUs. The data storage components of the generator and processor can be various types of memory that store information using magnetic energy, electrical energy, optical methods, or other methods, such as hard disks, USB flash drives, RAM, ROM, and quantum memories. The data processing components can call logic instructions in the data storage components to implement a method for identifying tomography based on differential cepstrum.

Claims

1. A method for identifying tomography based on differential cepstrum, characterized in that... The following steps are adopted: (1) Perform differential cepstral processing on the seismic data of the target section one channel at a time to obtain the differential cepstral data volume of the seismic data of the target section; The seismic data of the target segment is zero-padding, framed, and windowed channel by channel. Differential cepstrum calculation is performed on each windowed seismic signal frame to obtain the differential cepstrum data volume of the seismic data of the target segment. The differential cepstrum calculation formula is defined as follows: (1) In the formula, This is a windowed seismic signal frame. and These represent the forward and inverse Fourier transforms, respectively. For logarithmic operators, For frequency; for a windowed seismic signal frame The spectrum is calculated using Fourier transform. The absolute value of the spectrum is taken, and then a "1" is added. The logarithmic spectrum is calculated using logarithmic operators. The differential value of the logarithmic spectrum is then calculated, and the cepstrum is obtained using inverse Fourier transform. The differential cepstrum is calculated frame by frame to obtain the differential cepstrum data volume of the seismic data of the target section. (2) Extract the first-order differential cepstral coefficients of the differential cepstral data for each trace of the seismic data of the target section, and generate the first-order coefficient data of the differential cepstral. (3) For the extracted first-order coefficient data volume of the differential cepstrum, set the processing factor, calculate the eigenvalues ​​of the covariance matrix, and generate the fault information data volume; wherein, for the extracted first-order coefficient data volume of the differential cepstrum Set processing factor For the adjacent first and the Group First-order coefficient data volume of differential cepstral and The frames are divided into segments, with a frame length of [missing information]. Frame shift is The adjacent first and the Group Each frame of the first-order coefficient data volume of the differential cepstral is denoted as follows: and Calculate the covariance matrix of each frame in the first-order coefficient data volume of two adjacent sets of differential cepstrum. The following formula is used for calculation: (2) Calculate the covariance matrix The characteristic values ​​are used to generate fault information data volumes for the target seismic segment.

2. A device for identifying tomography based on differential cepstrum, the device implementing the method for identifying tomography based on differential cepstrum as described in claim 1, characterized in that, The system includes a Fourier transform spectrum generator, a logarithmic spectrum generator, an inverse Fourier transform processor, a differential cepstral first-order coefficient data volume generator, and a tomographic data generator. The Fourier transform spectrum generator, logarithmic spectrum generator, and inverse Fourier transform processor implement differential cepstral operations. The logarithmic spectrum generator performs the absolute value of the input spectrum, adds "1", and then calculates the logarithm. The differential cepstral first-order coefficient data volume generator performs a channel-by-channel, frame-by-frame operation to calculate the first-order differential cepstral coefficients from the differential cepstral data volume generated by the inverse Fourier transform processor, generating the differential cepstral first-order coefficient data volume. The tomographic data generator performs a channel-by-channel, frame-by-frame calculation to obtain the eigenvalues ​​of the covariance matrix of the differential cepstral first-order coefficient data volume, generating the tomographic information data volume.

Images