A thin layer weak signal enhancement method based on cyclic spectrum enhancement technology
By using cyclic spectrum enhancement technology and processing seismic data with Gaussian smoothing, even-order derivatives, and Butterworth low-pass filters, the problem of identifying weak signals in thin layers under complex geological conditions was solved, achieving efficient and accurate identification of thin reservoirs.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- EAST CHINA UNIV OF TECH
- Filing Date
- 2026-04-08
- Publication Date
- 2026-06-09
AI Technical Summary
Existing technologies suffer from high computational complexity and sensitivity to noise when processing thin, weak signals, making it difficult to accurately identify thin reservoirs under low signal-to-noise ratio conditions, and their effectiveness is limited, especially under complex geological conditions.
By employing cyclic spectrum enhancement technology, precise enhancement of weak signals in thin layers is achieved through Gaussian smoothing preprocessing, even-order derivative operation, Butterworth low-pass filter denoising, and adaptive cycle termination condition.
It improves the identification accuracy and computational efficiency of weak signals in thin layers, separates weak effective wave signals, and makes the phase axis of thin layer reflected waves more continuous and clear, making it suitable for large-scale seismic data processing.
Smart Images
Figure CN122172287A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of seismic signal processing, and specifically discloses a thin-layer weak signal enhancement method based on cyclic spectrum enhancement technology. Background Technology
[0002] With the continuous growth of global energy demand, the exploration and development of conventional oil and gas resources is becoming increasingly difficult, and thin-layer reservoirs and unconventional reservoirs are gradually becoming important targets for oil and gas exploration. Thin-layer reservoirs (such as shale, tight sandstone, and carbonate fracture-cavity formations) are typically characterized by small thickness, strong heterogeneity, and rapid lateral variation, and their seismic responses often exhibit weak reflections or low signal-to-noise ratio signals. In complex geological areas such as the Tarim Basin, due to the superposition of multiple tectonic phases and the complex reservoir structure, the reflection signals of thin layers are often masked by strong background noise or reflection signals from adjacent layers, thus severely restricting the accuracy of detailed reservoir characterization and oil and gas prediction. Therefore, how to effectively enhance the weak signals of thin layers and improve the resolution of seismic data has become an important research problem in the field of seismic signal processing.
[0003] Currently, researchers have proposed various seismic signal processing methods to improve thin-layer identification capabilities. One type of method is based on time-frequency analysis and spectral decomposition techniques. By performing time-frequency decomposition on seismic signals, the energy distribution characteristics of different frequency components can be extracted, thereby enabling thin-layer identification. For example, methods such as S-transform, generalized S-transform, and Gabor transform can improve the time-frequency resolution of seismic signals to some extent and enhance the reflection characteristics of thin layers. However, these methods typically have high computational complexity, are inefficient when processing large-scale seismic data, and are sensitive to noise, easily leading to misjudgments under low signal-to-noise ratio conditions.
[0004] Another type of method is based on spectral extension and inversion techniques. For example, methods such as spectral inversion, spectral restoration, or compressed sensing can broaden the frequency band of seismic signals, thereby improving the vertical resolution of seismic data and enabling clearer identification of thin reservoirs. Furthermore, deconvolution techniques are also widely used in thin-layer identification, recovering the formation reflection coefficient by removing wavelet influence and improving the resolution of thin-layer interfaces. However, these methods are susceptible to noise and wavelet instability under complex geological conditions, and their processing effectiveness has certain limitations.
[0005] In recent years, with the development of artificial intelligence technology, some studies have attempted to use deep learning methods to process seismic data, thereby improving the ability to identify thin layers by constructing neural network models to automatically extract seismic signal features. Although such methods have achieved certain results in some application scenarios, they usually require a large amount of training data, and the model training process is computationally intensive, thus still having certain limitations in practical seismic data processing.
[0006] Overall, existing thin-layer identification techniques have improved the resolution of seismic data to some extent, but some shortcomings remain. For example, some methods have high computational complexity, making it difficult to meet the needs of large-scale seismic data processing; some methods are sensitive to noise, and their identification accuracy is easily reduced under low signal-to-noise ratio conditions; in addition, existing research focuses more on thin-layer thickness or layer identification, while research on the enhancement and identification of extremely weak thin-layer signals generated by waveform interference effects is still relatively insufficient.
[0007] Studies have shown that increasing the dominant frequency of seismic signals is one of the important ways to enhance the identification capability of weak signals in thin layers. Therefore, it is necessary to study a processing method that can effectively enhance weak signals in thin layers while taking into account computational efficiency, from the perspective of signal processing, in order to improve the identification accuracy of thin reservoirs under complex geological conditions. Summary of the Invention
[0008] The purpose of this invention is to enable high-precision identification of weak signals in thin layers, overcome the resolution limitations of traditional methods, improve the sensitivity and accuracy of weak signal detection, and separate weak effective wave signals, thereby making the phase axis of the thin layer reflected wave more continuous and clear. Therefore, a thin layer weak signal enhancement method based on cyclic spectrum enhancement technology is proposed.
[0009] To achieve the above objectives, the technical solution adopted by the present invention is as follows:
[0010] A thin-layer weak signal enhancement method based on cyclic spectrum enhancement technology includes the following steps:
[0011] S1. The original seismic data is preprocessed with Gaussian smoothing along the time axis. The weighted average characteristics of the Gaussian function are used to suppress random noise while accurately preserving the main features of the signal, resulting in a smooth seismic trace.
[0012] S2. Calculate the even-order derivative of the smoothed seismic trace to generate derivative seismic traces. Utilize the nonlinear characteristics of the derivative operation to enhance the high-frequency detail information of the weak signal in the thin layer and highlight the reflection differences at the interlayer interface.
[0013] S3. To address the issue of high-order derivative operations easily introducing additional high-frequency noise, a Butterworth low-pass filter is used for directional noise reduction, achieving accurate separation of effective signals and noise;
[0014] S4. By dynamically selecting the optimal derivative order through adaptive loop termination conditions, even-order derivative seismic traces are cyclically superimposed onto the original seismic traces to avoid over-enhancement or under-enhancement, ultimately completing the identification and enhancement of thin-layer weak signals.
[0015] Furthermore, the Gaussian smoothing is a linear filtering method based on the Gaussian function, and its mathematical expression is:
[0016]
[0017] Where x is the offset from the center point; σ is the standard deviation, which controls the width of the Gaussian function. The larger σ is, the stronger the smoothing effect; G(x) represents the weight at position x.
[0018] Furthermore, for a continuous signal s(t), its nth derivative is defined as follows:
[0019]
[0020] Its first derivative The second derivative represents the rate of change of a signal and is often used to detect edges or abrupt changes in a signal; The rate of change of a signal, i.e., the curvature of the signal, is suitable for detecting inflection points or local extrema of a signal.
[0021] Furthermore, the formula for calculating the dominant frequency of the nth-order derivative seismic trace is as follows:
[0022]
[0023] Furthermore, the adaptive loop termination condition is set as follows:
[0024]
[0025] In the formula, ε is the termination parameter, which is generally between 0.2 and 0.3.
[0026] Furthermore, the differential threshold determination formula for the adaptive loop superposition termination condition is as follows:
[0027]
[0028] in, Let l represent the signal-to-noise ratio of the seismic trace with respect to the first derivative. When the above conditions are met, l is taken as the termination order of the derivative.
[0029] Furthermore, the transfer function of the Butterworth low-pass filter is:
[0030]
[0031] Furthermore, the algorithmic expression for the cyclic spectrum enhancement technique is as follows:
[0032]
[0033] Nor indicates signal normalization. Normalizing before and after processing can avoid incomplete results due to differences in data magnitude. This indicates Gaussian smoothing. They respectively represent the acquisition of the signal First derivative; This represents the difference threshold for calculating the derivative seismic trace. The order after the loop termination condition is determined needs to be negative and cyclically superimposed onto the seismic trace; Let be the order of the seismic trace, which is the largest integer multiple of 4 under this condition. , Find and The subsequent loop is superimposed onto the seismic trace. This indicates that Butterworth low-pass filtering is applied to high-frequency noise. This is the processed, normalized seismic trace.
[0034] The beneficial effects of the present invention are as follows: 1. Compared with the weak signal enhancement method based on empirical mode decomposition, the present invention has higher separation accuracy and can better handle thin-layer weak signal data with high mixing degree, and will not have the problems of reduced strong signal amplitude and unsatisfactory enhancement effect in weak signal region.
[0035] 2. Compared with the weak signal extraction method based on S-transform spectral decomposition technology, the present invention not only improves the identification accuracy of weak signals in thin layers and accelerates the suppression of noise, but also reduces the requirements for computer memory, separates weak effective wave signals, and makes the phase axis of the thin layer reflected wave more continuous and clear.
[0036] 3. Compared with adaptive signal processing methods such as CEEMD, this invention uses a Butterworth low-pass filter for directional noise reduction, which can remove high-frequency interference while retaining the amplitude of strong signals, enhance weak signals caused by interference, and make the layer information of thin interlayers clearer.
[0037] 4. This invention dynamically selects the optimal derivative order through adaptive loop termination conditions, avoiding over-enhancement or under-enhancement. While ensuring solution accuracy, it also enables faster convergence speed to meet the requirements of massive seismic data processing. Attached Figure Description
[0038] To more clearly illustrate the technical solutions in the embodiments of this application, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.
[0039] Figure 1 This is a flowchart of the cyclic spectrum enhancement technology of the present invention;
[0040] Figure 2 These are the unit impulse function and amplitude spectra of each derivative in the embodiments of the present invention;
[0041] Figure 3This is a comparison diagram of the original trace and the seismic traces of each order derivative in an embodiment of the present invention;
[0042] Figure 4 This is a comparison diagram of the waveforms processed by the present invention under different noise reduction methods;
[0043] Figure 5 This is a comparison of waveforms before and after thin-layer weak signal processing under different denoising methods according to the present invention;
[0044] Figure 6 This is a comparison of the amplitude spectra after two different thicknesses of thin interlayered processing according to the present invention;
[0045] Figure 7 This is a comparison chart of earthquake records processed by different denoising methods of the present invention;
[0046] Figure 8 This is a comparison diagram of thick numerical seismic traces and derivative seismic traces of the present invention;
[0047] Figure 9 This is a comparison diagram of thin-layer numerical seismic traces and derivative seismic traces of the present invention;
[0048] Figure 10 This is a comparison diagram of the recognition effect of thin interlayers of different thicknesses before and after the present invention. Detailed Implementation
[0049] The technical solution of the present invention will now be clearly and completely described with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.
[0050] Reference Figure 1 A thin-layer weak signal enhancement method based on cyclic spectrum enhancement technology is proposed. The main steps of implementing this method include Gaussian smoothing preprocessing, even-order derivative operation, Butterworth low-pass filter to remove high-frequency interference, adaptive loop termination condition determination, and cyclic spectrum enhancement algorithm implementation.
[0051] The specific steps are as follows:
[0052] Step 1: Gaussian Smoothing Preprocessing. The raw seismic data is preprocessed using the Gaussian smoothing operator. A weighted average of the signal is applied using a Gaussian function weight distribution to achieve a smoothing effect, accurately preserving the main characteristics of the signal while suppressing random noise. The mathematical expression for the Gaussian function is:
[0053]
[0054] Where x is the offset from the center point; σ is the standard deviation, which controls the width of the Gaussian function. By adjusting the standard deviation σ, the degree of smoothing can be flexibly controlled. The larger σ is, the stronger the smoothing effect; G(x) represents the weight at position x.
[0055] Step 2: Even-order derivative calculation. Calculate the even-order derivative of the smoothed seismic signal to generate the derivative seismic trace. For a continuous signal s(t), its nth-order derivative is defined as:
[0056]
[0057] First derivative The second derivative represents the rate of change of a signal, i.e., the slope of the signal, and is often used to detect edges or abrupt changes in a signal; The rate of change of a signal, i.e., the curvature of the signal, is suitable for detecting inflection points or local extrema of a signal.
[0058] From the perspective of derivative characteristics, odd-order derivatives introduce a 90° phase shift, making the derived seismic trace asymmetrical with the original seismic trace, which is not suitable for thin-layer weak signal enhancement. Even-order derivatives, on the other hand, have no phase shift, and their waveforms remain symmetrical with the original seismic traces. Only the polarity changes regularly with the order: the polarities of second-order and sixth-order derivative seismic traces are opposite, and one of them needs to be negatively corrected before superposition; the polarities of fourth-order and eighth-order derivative seismic traces are consistent, and they can be directly superimposed.
[0059] Step 3: Adaptive loop termination condition determination. Calculate the dominant frequency of each derivative seismic trace:
[0060]
[0061] Set the adaptive loop termination condition as follows:
[0062]
[0063] In the formula, ε is the termination parameter, which is generally between 0.2 and 0.3.
[0064] Simultaneously, the signal-to-noise ratio of seismic traces of each order derivative is calculated. Set the differential threshold judgment conditions:
[0065]
[0066] in, Let l represent the signal-to-noise ratio of the seismic trace with respect to the first derivative. When the above conditions are met, l is taken as the termination order of the derivative.
[0067] Step 4: Butterworth low-pass filter for high-frequency interference removal. To address the issue of high-order derivative operations introducing additional high-frequency noise, a Butterworth low-pass filter is used for targeted noise reduction. The transfer function of the Butterworth low-pass filter is:
[0068]
[0069] In the formula, H(s) is the transfer function of the filter. is the cutoff frequency, and n is the filter order.
[0070] Step 5: Implementation of the Cyclic Spectrum Enhancement Algorithm. Integrating the above steps forms a complete cyclic spectrum enhancement algorithm:
[0071]
[0072] Nor indicates signal normalization. Normalizing before and after processing can avoid incomplete results due to differences in data magnitude. This indicates Gaussian smoothing. They respectively represent the acquisition of the signal First derivative; This represents the difference threshold for calculating the derivative seismic trace. The order after the loop termination condition is determined needs to be negative and cyclically superimposed onto the seismic trace; Let be the order of the seismic trace, which is the largest integer multiple of 4 under this condition. , Find and The subsequent loop is superimposed onto the seismic trace. This indicates that Butterworth low-pass filtering is applied to high-frequency noise. This is the processed, normalized seismic trace.
[0073] Reference Figures 2 to 10 Numerical simulations and tests using actual seismic data both demonstrate that the proposed method can effectively match and enhance the stratigraphic level of the original thin-layer weak signals, thus significantly improving the identification capability of thin-layer weak signals. Model test results show that the method does not significantly enhance strong signals in 40m thick layers, but it can significantly enhance weak signals in 10m (λ / 4) thin layers, especially showing good compensation effects for extremely weak signals in 5m (λ / 8) layers, and it performs excellently in maintaining data fidelity and stratigraphic consistency.
[0074] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit them. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some or all of the technical features therein. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention.
Claims
1. A thin-layer weak signal enhancement method based on cyclic spectrum enhancement technology, characterized in that, Includes the following steps: S1. The original seismic data is preprocessed with Gaussian smoothing along the time axis. The weighted average characteristics of the Gaussian function are used to suppress random noise while accurately preserving the main features of the signal, resulting in a smooth seismic trace. S2. Calculate the even-order derivative of the smoothed seismic trace to generate derivative seismic traces. Utilize the nonlinear characteristics of the derivative operation to enhance the high-frequency detail information of the weak signal in the thin layer and highlight the reflection differences at the interlayer interface. S3. To address the issue of high-order derivative operations easily introducing additional high-frequency noise, a Butterworth low-pass filter is used for directional noise reduction, achieving accurate separation of effective signals and noise; S4. By dynamically selecting the optimal derivative order through adaptive loop termination conditions, even-order derivative seismic traces are cyclically superimposed onto the original seismic traces to avoid over-enhancement or under-enhancement, ultimately completing the identification and enhancement of thin-layer weak signals.
2. The thin-layer weak signal enhancement method based on cyclic spectrum enhancement technology according to claim 1, characterized in that, Gaussian smoothing preprocessing: The raw seismic data is preprocessed using a Gaussian smoothing operator. A weighted average of the signal is applied using a Gaussian function weight distribution to achieve a smoothing effect, accurately preserving the main characteristics of the signal while suppressing random noise. The mathematical expression for the Gaussian function is: Where x is the offset from the center point; σ is the standard deviation, which controls the width of the Gaussian function. By adjusting the standard deviation σ, the degree of smoothing can be flexibly controlled. The larger σ is, the stronger the smoothing effect; G(x) represents the weight at position x.
3. The thin-layer weak signal enhancement method based on cyclic spectrum enhancement technology according to claim 1, characterized in that, The rules for even-order derivative operations are as follows: For a continuous signal s(t), its nth derivative is defined as: First derivative The second derivative represents the rate of change of a signal, i.e., the slope of the signal, and is often used to detect edges or abrupt changes in a signal; The rate of change of a signal, i.e., the curvature of the signal, is suitable for detecting inflection points or local extrema of a signal. From the perspective of derivative characteristics, odd-order derivatives introduce a 90° phase shift, making the derived seismic trace asymmetrical with the original seismic trace, which is not suitable for thin-layer weak signal enhancement. Even-order derivatives, on the other hand, have no phase shift, and their waveforms remain symmetrical with the original seismic traces. Only the polarity changes regularly with the order: the polarities of second-order and sixth-order derivative seismic traces are opposite, and one of them needs to be negatively corrected before superposition; the polarities of fourth-order and eighth-order derivative seismic traces are consistent, and they can be directly superimposed.
4. The thin-layer weak signal enhancement method based on cyclic spectrum enhancement technology according to claim 1, characterized in that, Adaptive loop termination condition determination, calculation of dominant frequencies of seismic traces for each order derivative: Set the adaptive loop termination condition as follows: In the formula, ε is the termination parameter, which generally takes a value between 0.2 and 0.3; Simultaneously, the signal-to-noise ratio of seismic traces of each order derivative is calculated. Set the differential threshold judgment conditions: in, Let l represent the signal-to-noise ratio of the seismic trace with respect to the first derivative. When the above conditions are met, l is taken as the termination order of the derivative.
5. The thin-layer weak signal enhancement method based on cyclic spectrum enhancement technology according to claim 1, characterized in that, The Butterworth low-pass filter is incorporated into the cyclic spectrum enhancement algorithm, and its expression is as follows: The transfer function of the Butterworth low-pass filter is: In the formula, H(s) is the transfer function of the filter. is the cutoff frequency, and n is the filter order.
6. The thin-layer weak signal enhancement method based on cyclic spectrum enhancement technology according to claim 1, characterized in that, The complete algorithmic expression for cyclic spectral enhancement is: Nor indicates signal normalization. Normalizing before and after processing can avoid incomplete results due to differences in data magnitude. This indicates Gaussian smoothing. They respectively represent the acquisition of the signal First derivative; This represents the difference threshold for calculating the derivative seismic trace. The order after the loop termination condition is determined needs to be negative and cyclically superimposed onto the seismic trace; Let be the order of the seismic trace, which is the largest integer multiple of 4 under this condition. , Find and The subsequent loop is superimposed onto the seismic trace; This indicates that Butterworth low-pass filtering is applied to high-frequency noise. This is the processed, normalized seismic trace.