Thin interbed lithologic interface identification method and system

Abstract

The invention provides a thin interbed lithologic interface identification method and system. The identification method comprises the steps of S1, preferably selecting a sensitive logging curve; s2, performing multi-scale high and low frequency signal decomposition on the thin interbed lithologic interface GR logging curve by using wavelet transform; s3, analyzing the decomposed high-frequency signal and the decomposed low-frequency signal, selecting the decomposed high-frequency signal and the decomposed low-frequency signal which can represent an effective signal, filtering out a curve carrying noise, and reconstructing an REGR curve; and S4, the REGRr curve reconstructed in the step S3 is processed through a deconvolution method, and finally the REGRr curve is obtained. According to the method, the longitudinal logging curve resolution is improved through deconvolution, the thin interbed lithologic layered interface information is highlighted, and the identification accuracy of the thin interbed lithologic layered interface is effectively improved.

Description

Technical Field

[0001] The present invention relates to the technical field of oil and gas exploration, and particularly to a method and system for identifying thin interbedded lithologic interfaces. Background Art

[0002] The lithology type of a reservoir is closely related to physical properties, oil and gas distribution, etc. Identifying the thin interbedded lithologic stratification interface is a basic research work in well logging interpretation. Accurately identifying the thin interbedded lithologic stratification interface is of great significance for later reservoir effectiveness evaluation, accurate parameter calculation, reserve evaluation, etc. However, due to the large differences in lithology of different grain sizes and minerals in the formation and the strong vertical lithologic heterogeneity, the thin interbedded lithology frequently alternates in a single sand body, posing a great challenge to the identification of the thin interbedded lithologic stratification interface.

[0003] Currently, there are many methods for identifying lithologic interfaces. The commonly used method is to identify based on well logging data by means of single signal analysis and experimental tests, etc. However, this method often does not consider the existence of thin interbedded lithology, and thus has a poor identification effect on the lithologic interfaces of thin interbedded alternations. Therefore, how to solve the problem of low accuracy in identifying the thin interbedded lithologic stratification interface is a technical problem that needs to be urgently solved by those skilled in the art.

[0004] For example, the uranium geology journal, volume 37, issue 2, pages 262 - 268, introduces the identification of formation lithologic interfaces based on wavelet analysis of well logging curves, including the following steps: S1: performing continuous wavelet and discrete wavelet transforms on the triple lateral resistivity curve; S2: plotting the wavelet coefficient scale diagram; S3: plotting the high-frequency wavelet coefficient curve diagram; S4: comprehensively using the wavelet coefficient scale diagram and the high-frequency wavelet coefficient curve diagram to identify the formation lithologic interface; this method realizes the identification of the interfaces of large sections of sandstone or thin interbedded sandstone sections in the formation and solves the interface identification of lithology in sandstone-shale alternating strata. However, it still does not involve the identification of the lithology of the thin interbedded sandstone stratification interface, and only has a certain identification effect on the identification of the interfaces of large sections of sandstone or thin interbedded sandstone sections, and has limitations in the later division of sedimentary environment and cycles, as well as reserve evaluation.

[0005] For example, the patent application number is CN202111484350.6, a DC detection method for identifying the lithological interface in front of the tunnel excavation face, in which power supply electrode A, receiving electrode M and receiving electrode N are arranged in sequence in the tunnel along the direction away from the excavation working face, and the power supply electrode A and the receiving electrodes M and N are arranged along the same straight line parallel to the tunnel; when the power supply electrode A is fixed at point A1, point A2 and point A3 respectively, the receiving electrodes M and N repeatedly perform the following actions: the receiving electrodes M and N both start from their starting positions and move alternately along the straight line where they are located in the direction away from the excavation working face. Each time the receiving electrode moves once, the power supply electrode A emits current once, and the receiving electrodes M and N measure the potential difference once. At the same time, the measurement value corresponds to a measuring point. After the receiving electrodes M and N repeatedly move alternately, three groups of measuring points are obtained. The three groups of measuring points correspond to the power supply electrode A at points A1, A2 and A3, respectively. A3 position, where points A1, A2 and A3 are located between the receiving electrode and the excavation working face and are located in the same straight line as the receiving electrodes M and N; the spherical shell apparent resistivity of each measuring point in the three groups of measuring points is calculated respectively: where ΔU is the receiving potential difference between the receiving electrodes M and N, and I is the power supply current of the power supply electrode A; L is the distance between the power supply electrode A and the measuring point, and MN is the distance between the receiving electrodes M and N, and the measuring point is located at the midpoint of the receiving electrodes M and N; with the distance L between the power supply electrode A and the measuring point as the abscissa and the corresponding spherical shell apparent resistivity ρ as the ordinate, three spherical shell apparent resistivity curves corresponding to the three groups of measuring points are drawn; interference suppression and data superposition processing are performed on the three groups of spherical shell resistivity curves to obtain the spherical shell apparent resistivity curve in front of the excavation face; the low-resistance anomaly area is delineated according to the spherical shell apparent resistivity value in the spherical shell apparent resistivity curve in front of the excavation face, and the influence range of the low-resistance anomaly area is the range of the interface between different lithologies. Although this method can identify lithologic interfaces, it does not involve the identification of stratified interfaces of thin interbedded lithologies. Moreover, it is conducted through experimental testing and is not applicable to the identification of lithologic stratified interfaces in strata. Summary of the invention

[0006] In view of the above problems, the present invention is proposed to provide a thin interbedded lithology interface identification method and system that overcomes the above problems or at least partially solves the above problems.

[0007] According to one aspect of the present invention, a method for identifying a thin interbedded lithology interface is provided, the method comprising:

[0008] Step S1: Optimizing sensitive logging curves;

[0009] Step S2: using wavelet transform to perform multi-scale high- and low-frequency signal decomposition on the GR logging curve of the thin interbedded lithological interface;

[0010] Step S3: Analyze the decomposed high- and low-frequency signals, select the high- and low-frequency signals that can represent the effective signals after decomposition, filter out the curves carrying noise, and reconstruct the REGR curve;

[0011] Step S4: Process the REGR curve reconstructed in Step S3 through the deconvolution method to finally obtain the REGRr curve.

[0012] Optionally, the Step S1: Optimize the sensitive logging curve specifically includes:

[0013] The logging curve is a record of the variation of formation physical properties with well depth and includes geological formation information;

[0014] Using the natural gamma ray GR logging is a method of measuring the natural radioactivity of rock formations along the wellbore to study the properties of rock formations.

[0015] Optionally, the radioactivity of the rock formation is related to the radioactivity intensity and content of the minerals contained in the rock;

[0016] Using the variation of the natural gamma ray GR logging value is used to reflect the differences in complex lithologies composed of various minerals and has a relatively obvious discrimination for different types of rocks.

[0017] Optionally, the Step S2: Use wavelet transform to perform multi-scale high- and low-frequency signal decomposition on the GR logging curve of the thin interbed lithology interface specifically includes:

[0018] Step S2: Use the db4 wavelet basis function in the compactly supported orthogonal wavelet - db wavelet in wavelet transform to perform multi-scale high- and low-frequency signal decomposition on the GR logging curve of the thin interbed lithology interface.

[0019] Optionally, the REGRr curve can identify the thin interbed lithology stratification interface.

[0020] Optionally, the Step S2: Use wavelet transform to perform multi-scale high- and low-frequency signal decomposition on the GR logging curve of the thin interbed lithology interface specifically includes:

[0021] Use wavelet transform to perform multi-scale wavelet coefficient decomposition on the logging curve data;

[0022] The curve shape after decomposition is related to the original formation information;

[0023] The non-formation information is in the highest frequency information after decomposition.

[0024] Optionally, the Step S3: Analyze the decomposed high- and low-frequency signals, select the high- and low-frequency signals that can represent the effective signals after decomposition, filter out the curves carrying noise, and reconstruct the REGR curve specifically includes:

[0025] After the signal is decomposed, it is necessary to reconstruct the decomposed signal. Wavelet reconstruction is the inverse process of wavelet decomposition and is a process of reconstructing the signal using the decomposed signal;

[0026] Based on the decomposition of high- and low-frequency signals in wavelet transform, signals of different scales in logging data can be quickly separated, and the reconstruction of signals with different frequencies can filter out wavelet coefficients carrying noise;

[0027] Select the high- and low-frequency signals that can represent the effective signal after decomposition for reconstruction, and finally obtain the signal that can represent the true information.

[0028] Optionally, step S4: processing the REGR curve reconstructed in step S3 by deconvolution method to finally obtain the REGRr curve specifically includes:

[0029] Using db4 wavelet to decompose the high- and low-frequency wavelet coefficients of logging data, filtering out the wavelet coefficients carrying noise, and reconstructing the signal representing the true information of the formation;

[0030] Based on the reconstructed logging curve, perform deconvolution to obtain the REGRr curve.

[0031] Optionally, step S4: processing the REGR curve reconstructed in step S3 by deconvolution method to finally obtain the REGRr curve specifically includes:

[0032] The deconvolution process is to construct a linear equation system with the error value between the expected output and the actual output, and perform convolution processing on the deconvolution factor obtained by solving the equation system and the logging response value to obtain the true formation logging response value. The determination of the instrument response function and the deconvolution factor is the key to the deconvolution technology;

[0033] Taking natural gamma logging as an example, its response function K(z) can be expressed by an exponential decay function:

[0034]

[0035] In the formula, α is a geological characteristic parameter, which depends on the content of mineral radioactive elements and determines the decay rate of the exponential function; z is the coordinate of the logging point in the vertical direction;

[0036] Convolve both sides of the above formula with K-1(z), where K-1(z) is the deconvolution factor of K(z), and finally obtain the true logging response value of the formation through Fourier transform.

[0037] The present invention also provides a thin interbed lithology interface recognition system, applying the above-mentioned thin interbed lithology interface recognition method, and the recognition system specifically includes:

[0038] Logging curve optimization module, used to optimize sensitive logging curves;

[0039] Wavelet transform processing module, which is used to perform multi-scale high- and low-frequency signal decomposition on the GR logging curve of thin interbedded lithologic interfaces by using wavelet transform;

[0040] REGR curve reconstruction module, which is used to analyze the decomposed high- and low-frequency signals, select the high- and low-frequency signals that can represent the effective signals after decomposition, filter out the curves carrying noise, and reconstruct the REGR curve;

[0041] Deconvolution processing module, which is used to process the reconstructed REGR curve by deconvolution method to finally obtain the REGRr curve.

[0042] A method and system for identifying thin interbedded lithologic interfaces provided by the present invention, the identification method includes:

[0043] Step S1: Optimize sensitive logging curves; Step S2: Perform multi-scale high- and low-frequency signal decomposition on the GR logging curve of thin interbedded lithologic interfaces by using wavelet transform; Step S3: Analyze the decomposed high- and low-frequency signals, select the high- and low-frequency signals that can represent the effective signals after decomposition, filter out the curves carrying noise, and reconstruct the REGR curve; Step S4: Process the REGR curve reconstructed in Step S3 by deconvolution method to finally obtain the REGRr curve. The present invention improves the resolution of longitudinal logging curves through deconvolution, highlights the information of thin interbedded lithologic stratification interfaces, and effectively improves the accuracy of identifying thin interbedded lithologic stratification interfaces.

[0044] The above description is only an overview of the technical solution of the present invention. In order to be able to understand the technical means of the present invention more clearly, it can be implemented according to the content of the specification. And in order to make the above and other purposes, features and advantages of the present invention more obvious and understandable, the following specifically gives the specific embodiments of the present invention. Description of the Drawings

[0045] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings required for the description of the embodiments will be briefly introduced below. Obviously, the drawings in the following description are only some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can also be obtained according to these drawings.

[0046] Figure 1 It is a flowchart of a method for identifying thin interbedded lithologic interfaces provided by an embodiment of the present invention;

[0047] Figure 2 It is a schematic diagram of the principle of wavelet transform provided by Embodiment 1 of the present invention.

[0048] Figure 3 It is a schematic diagram of the principle of thin layer logging deconvolution in Embodiment 1 of the present invention.

[0049] Figure 4 This is the application effect diagram of identifying thin interbed lithologic interfaces by combining wavelet high-frequency decomposition and deconvolution in Embodiment 1 of the present invention. Specific implementation manners

[0050] Hereinafter, exemplary embodiments of the present disclosure will be described in more detail with reference to the accompanying drawings. Although the exemplary embodiments of the present disclosure are shown in the drawings, it should be understood that the present disclosure can be implemented in various forms and should not be limited by the embodiments set forth herein. On the contrary, these embodiments are provided so that the present disclosure can be more thoroughly understood and the scope of the present disclosure can be completely conveyed to those skilled in the art.

[0051] The terms "comprising" and "having" and any variations thereof in the description, embodiments and claims of the present invention are intended to cover non-exclusive inclusion. For example, a series of steps or units are included.

[0052] Hereinafter, with reference to the accompanying drawings and embodiments, the technical solutions of the present invention will be described in further detail.

[0053] As Figure 1 shown, a method for identifying thin interbed lithologic interfaces by combining wavelet and deconvolution includes the following steps:

[0054] Step 1: Optimize sensitive logging curves. Logging curves are records of the variation of formation physical properties with well depth and contain a lot of geological formation information. GR (natural gamma) logging is a method of measuring the natural radioactivity of rock formations along the wellbore to study the properties of rock formations. The radioactivity of rock formations is mainly related to the radioactivity intensity and content of minerals contained in the rocks. The change of natural gamma logging values can be used to reflect the differences in complex lithologies composed of various minerals and has a relatively obvious discrimination degree for different types of rocks. Therefore, GR is selected as the sensitive logging curve for identifying thin interbed lithologic interfaces;

[0055] Step 2: From the perspective of signal analysis, considering the characteristics that wavelet transform multiple decomposition and reconstruction are helpful for improving the signal-to-noise ratio, use the db4 wavelet basis function in the compactly supported orthogonal wavelet - db (Daubechies) wavelet with good regularity and time-frequency compact support in wavelet transform to perform high- and low-frequency signal decomposition of the GR logging curve that is sensitive to thin interbed lithologic interfaces at multiple decomposition levels;

[0056] The basic principle of wavelet transform decomposition is as follows: Wavelet transform is a method of converting a signal into different forms of signals for processing. This conversion can observe the signal at different scales, thereby revealing the hidden information in the signal. Usually, the signals collected from the device have a certain amount of noise. The signal contaminated by noise is mainly a combination of a clean signal and noise. Therefore, it is necessary to denoise the signal contaminated by noise to obtain a signal that can represent the true information. According to the characteristic that the wavelet decomposition coefficients of the signal in different frequency bands have different intensity distributions, analyze the wavelet coefficients in each frequency band, so as to distinguish the signal from the noise, remove the wavelet coefficients generated by the noise, and retain the wavelet coefficients generated by the signal. Therefore, using wavelet transform to perform multi-scale wavelet coefficient decomposition on well logging curve data, the decomposed curve morphology has a certain connection with the original formation information, and the non-formation information is mainly in the highest frequency information after decomposition ( Figure 2 ).

[0057] The specific implementation steps of wavelet transform decomposition are as follows:

[0058] In the dbN wavelet, N is the wavelet order, also known as the vanishing moment. Generally, as the order (sequence N) increases, the vanishing moment order becomes larger, the vanishing moment becomes higher, the smoothness is better, the frequency domain localization ability is stronger, and the frequency band division effect is better. However, at the same time, it will weaken the time domain tight support, increase the calculation amount, and deteriorate the real-time performance. Among them, the db4 wavelet has a 4th-order vanishing moment and an effective support length of 7. Theoretically, when the decomposition scale is large enough, the signal can be infinitely decomposed. However, in practice, when the required signal is decomposed from the original signal into high- and low-frequency signals, when the decomposition scale is within a certain range, the optimal result can be obtained. For the GR well logging curve here, when the decomposition scale is 5, the optimal processing of the data can be achieved. Therefore, use the db4 wavelet to perform 5-layer high- and low-frequency decomposition on GR to obtain the high-frequency decomposition curves REGR-D1~REGR-D5 and low-frequency decomposition curves REGR-A1~REGR-A5 of different wavelets. The decomposition process is to use the appcoef, detcoef, and wrcoef functions in MATLAB for processing. The function call methods are as follows:

[0059] CAN = appcoef(C, L, 'db4', N)

[0060] CDN = detcoef(C, L, N)

[0061] AN = wrcoef('a', C, L, 'db4', N)

[0062] DN = wrcoef('d', C, L, 'db4', N)

[0063] Where: C is a vector storing the approximation coefficient (CA) and the detail coefficient (CD), L is the length corresponding to CA and CD, CAN is the N - layer low - frequency coefficient extracted by one - dimensional wavelet transform, CDN is the N - layer high - frequency coefficient extracted by one - dimensional wavelet transform, AN is the N - layer low - frequency signal after decomposition, and DN is the N - layer high - frequency signal after decomposition.

[0064] Step 3: Analyze the decomposed high - and low - frequency signals, select the high - and low - frequency signals that can represent the effective signal after decomposition, filter out the curves carrying noise, and reconstruct the REGR curve that can represent the true formation information.

[0065] The basic principle of wavelet transform decomposition and reconstruction is as follows: After the signal is decomposed, the decomposed signal needs to be reconstructed. Wavelet reconstruction is the inverse process of wavelet decomposition and is a process of reconstructing the signal using the decomposed signal. Based on the decomposition of high - and low - frequency signals in wavelet transform, signals of different scales in well logging data can be quickly separated, and the reconstruction of signals with different frequencies can filter out the wavelet coefficients carrying noise. Select the high - and low - frequency signals that can represent the effective signal after decomposition for reconstruction, and finally obtain the signal that can represent the true information ( Figure 2 ).

[0066] The specific implementation steps of wavelet transform decomposition and reconstruction are as follows:

[0067] As the sequence numbers of the high - and low - frequency decomposition curves increase, on the basis of not losing the changing trend of the original well logging curve, the decomposed high - frequency curve REGR - D3 and the low - frequency decomposition curve REGR - A1 become smoother compared with the other 4 decomposed high - and low - frequency curves, and have a higher correlation with the original well logging curve. Therefore, in order to retain the reflection of the GR curve on the lithology change of the formation, a REGR curve without noise that reflects the original formation lithology information is reconstructed using the high - frequency decomposition curve REGR - D3 and the low - frequency decomposition curve REGR - A1. Through db4 wavelet decomposition and reconstruction, noise can be screened out, and while reducing the noise of the GR curve, the original formation information can be retained. The reconstruction is processed by superimposing the low - frequency and high - frequency reconstruction signals, and the reconstruction process is as follows:

[0068] READ = A1 + D3

[0069] Where: A1 is the low - frequency signal after 1 - layer decomposition and reconstruction, D3 is the high - frequency signal after 3 - layer decomposition and reconstruction, and READ is the final reconstructed curve with noise filtered out based on wavelet transform.

[0070] Step 4: Process the REGR curve reconstructed in Step 3 through a deconvolution method that can improve the vertical resolution of the well logging curve, and finally obtain the REGRr curve that can accurately identify the lithology stratification interface of thin interbeds.

[0071] Stratigraphic deconvolution is a method commonly used in the field of geophysics to improve resolution and is widely applied in seismic and logging. From the perspective of considering signal principles, logging deconvolution believes that logging curves are the superposition of the comprehensive response signals detected by the instrument in the formation. The logging data at a certain depth point is not the response value of the real formation, but the weighted average of the measurement values at that point and the surrounding rock. Ignoring the influence of noise, the logging response formula of stratigraphic deconvolution is as follows:

[0072] S(z) = K(z) * T(z)

[0073] Where: S(z) is the logging response value, K(z) is the instrument response function, and T(z) is the true logging response value of the formation. Its basic principle is that when the formation is a thick layer, the formation signal reflected by the logging instrument is the real formation information. When there are thin layers in the formation, the logging response information is the convolution of the real signal of the formation and the logging instrument response function. The result of the convolution makes the logging curve smoother, resulting in unclear display of the logging response characteristics for the thin layers of the formation and their stratification interfaces ( Figure 3 ). Therefore, in order to improve the vertical logging resolution of thin layers, it is necessary to perform deconvolution on the logging curve data, restore the true logging response signal of the thin layers, and finally obtain the logging curve data that can reflect the real formation information.

[0074] The deconvolution process is to construct a linear equation system with the error value between the expected output and the actual output, and perform convolution processing on the deconvolution factor obtained by solving the equation system and the logging response value to obtain the true logging response value of the formation. The determination of the instrument response function and the deconvolution factor is the key to the deconvolution technology. Taking natural gamma logging as an example, its response function K(z) can be expressed by an exponential decay function:

[0075]

[0076] Where, α is a geological characteristic parameter, which depends on the content of mineral radioactive elements and determines the decay rate of the exponential function; z is the coordinate of the logging point in the vertical direction. Convolve both sides of the above formula with K-1(z) respectively, and K-1(z) is the deconvolution factor of K(z). Finally, the true logging response value of the formation can be obtained through Fourier transform.

[0077] The basic theoretical method of using deconvolution to measure the resolution of logging curves is as follows: In view of the advantages of wavelet transform in high-frequency and low-frequency coefficient decomposition, reconstruction, and deconvolution, in order to enhance the recognition ability of lithological interfaces without losing the real signal of the original formation. First, use the db4 wavelet to decompose the high-frequency and low-frequency wavelet coefficients of the logging data, filter out the wavelet coefficients carrying noise, reconstruct the signal representing the real information of the formation, and then perform deconvolution based on the reconstructed logging curve to improve its vertical resolution.

[0078] The specific implementation steps for improving the resolution of logging curves using deconvolution are as follows: Based on the reconstruction of the REGR curve by wavelet transform, the REGR curve is processed using deconvolution to obtain the REGRr curve. By observing the morphology of the processed curve, a characteristic "amplification" phenomenon appears at the interface of lithological changes, improving the vertical logging resolution of thin beds and highlighting the information of the lithological stratification interface of thin interbeds. The REGRr curve after deconvolution processing has a greater response amplitude and clearer boundaries when the lithology changes. Therefore, from the perspective of signal analysis, integrating the advantages of wavelet transform and deconvolution methods can accurately and quickly identify the lithological stratification interface of thin interbeds, and finally establish a method for identifying the lithological stratification interface of thin interbeds.

[0079] Figure 4 This is the application of the present invention in identifying the lithological interface of thin interbeds in the reservoir section of a single well.

[0080] The method for identifying the lithological stratification interface of thin interbeds of the present invention can finely identify the lithological stratification interface of thin interbeds, providing a new idea for lithological interface identification and laying a foundation for the later division of sedimentary environment and cycles, as well as fluid analysis and reserve evaluation. Through the decomposition and reconstruction methods of wavelet transform, the present invention can achieve noise reduction and retain the original formation information, making the lithological stratification interface information more accurate and obvious. By using deconvolution to improve the resolution of vertical logging curves, the present invention highlights the information of the lithological stratification interface of thin interbeds and effectively improves the accuracy of identifying the lithological stratification interface of thin interbeds. By combining wavelet and deconvolution, the present invention not only weakens the influence of noise in formation logging signals, but also improves the resolution of logging curves to a certain extent.

[0081] Beneficial effects: (1) The method for identifying the lithological stratification interface of thin interbeds of the present invention can finely identify the lithological stratification interface of thin interbeds, providing a new idea for lithological interface identification and laying a foundation for the later division of sedimentary environment and cycles, as well as fluid analysis and reserve evaluation;

[0082] (2) Through the decomposition and reconstruction methods of wavelet transform, the present invention can achieve noise reduction and retain the original formation information, making the lithological stratification interface information more accurate and obvious;

[0083] (3) By using deconvolution to improve the resolution of vertical logging curves, the present invention highlights the information of the lithological stratification interface of thin interbeds and effectively improves the accuracy of identifying the lithological stratification interface of thin interbeds;

[0084] (4) By combining wavelet and deconvolution, the present invention not only weakens the influence of noise in formation logging signals, but also improves the resolution of logging curves to a certain extent.

[0085] In the above specific embodiments, the object, technical solution and beneficial effects of the present invention have been further described in detail. It should be understood that the above are only specific embodiments of the present invention and are not used to limit the protection scope of the present invention. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. A method for identifying thin interbedded lithologic interfaces, characterized in that, the identification method includes: Step S1: Optimize sensitive logging curves; Step S2: Use wavelet transform to perform multi-scale high- and low-frequency signal decomposition on the GR logging curve of the thin interbedded lithologic interface; Step S3: Analyze the decomposed high- and low-frequency signals, select the high- and low-frequency signals that can represent the effective signals after decomposition, filter out the curves carrying noise, and reconstruct the REGR curve; Step S4: Process the REGR curve reconstructed in Step S3 by deconvolution method to finally obtain the REGRr curve.

2. A method for identifying thin interbedded lithologic interfaces according to claim 1, characterized in that, the Step S1: Optimize sensitive logging curves specifically includes: The logging curve is a record of the variation of formation physical properties with well depth and includes geological formation information; Using natural gamma ray GR logging is a method of measuring the natural radioactivity of rock formations along the wellbore to study the properties of rock formations.

3. A method for identifying thin interbedded lithologic interfaces according to claim 2, characterized in that, the radioactivity of the rock formation is related to the radioactivity intensity and content of the minerals contained in the rock; Using the variation of natural gamma ray GR logging values is used to reflect the differences in complex lithologies composed of various minerals, and has a relatively obvious discrimination for different types of rocks.

4. A method for identifying thin interbedded lithologic interfaces according to claim 1, characterized in that, the Step S2: Use wavelet transform to perform multi-scale high- and low-frequency signal decomposition on the GR logging curve of the thin interbedded lithologic interface specifically includes: Step S2: Use the compactly supported orthogonal wavelet - db4 wavelet basis function in wavelet transform to perform multi-scale high- and low-frequency signal decomposition on the GR logging curve of the thin interbedded lithologic interface.

5. A method for identifying thin interbedded lithologic interfaces according to claim 1, characterized in that, the REGRr curve can identify the thin interbedded lithologic stratification interface.

6. A method for identifying thin interbedded lithologic interfaces according to claim 1, characterized in that, the Step S2: Use wavelet transform to perform multi-scale high- and low-frequency signal decomposition on the GR logging curve of the thin interbedded lithologic interface specifically includes: Use wavelet transform to perform multi-scale wavelet coefficient decomposition on the logging curve data; The morphology of the decomposed curve is related to the original formation information; The non-formation information is in the highest frequency information after decomposition.

7. A method for identifying thin interbedded lithologic interfaces according to claim 1, characterized in that, the Step S3: Analyze the decomposed high- and low-frequency signals, select the high- and low-frequency signals that can represent the effective signals after decomposition, filter out the curves carrying noise, and reconstruct the REGR curve specifically includes: After the signal is decomposed, it is necessary to perform reconstruction processing on the decomposed signal, and wavelet reconstruction is the inverse process of wavelet decomposition, which is a process of using the decomposed signal to reconstruct the signal; Based on the high- and low-frequency signal decomposition in wavelet transform, different scale signals in logging data can be quickly separated, and the reconstruction of different frequency signals can filter out the wavelet coefficients carrying noise; Select the high- and low-frequency signals that can represent the effective signals after decomposition for reconstruction to finally obtain a signal that can represent the real information.

8. A method for identifying thin interbedded lithologic interfaces according to claim 1, characterized in that, the step S4: processing the REGR curve reconstructed in step S3 by deconvolution method to finally obtain the REGRr curve specifically includes: Using db4 wavelet to decompose the logging data into high and low frequency wavelet coefficients, filtering out the wavelet coefficients carrying noise, and reconstructing the signal representing the true formation information; Based on the reconstructed logging curve, deconvolution is performed to obtain the REGRr curve.

9. A method for identifying thin interbedded lithologic interfaces according to claim 1, characterized in that, the step S4: processing the REGR curve reconstructed in step S3 by deconvolution method to finally obtain the REGRr curve specifically includes: The deconvolution process is to construct a linear equation system with the error values between the desired output and the actual output, and perform convolution processing on the deconvolution factor obtained by solving the equation system and the logging response value to obtain the true formation logging response value. The determination of the instrument response function and the deconvolution factor is the key to the deconvolution technology; Taking natural gamma logging as an example, its response function K(z) can be expressed by an exponential decay function: In the formula, α is a geological characteristic parameter, which depends on the content of mineral radioactive elements and determines the decay speed of the exponential function; z is the coordinate of the logging point in the vertical direction; Convolve both sides of the above formula with K-1(z), where K-1(z) is the deconvolution factor of K(z), and finally obtain the true logging response value of the formation through Fourier transform.

10. A thin interbedded lithologic interface identification system applying the thin interbedded lithologic interface identification method described in any one of claims 1-9 above, characterized in that, the identification system specifically includes: A logging curve optimization module for optimizing sensitive logging curves; A wavelet transform processing module for performing multi-scale high and low frequency signal decomposition on the GR logging curve of the thin interbedded lithologic interface by wavelet transform; A REGR curve reconstruction module for analyzing the decomposed high and low frequency signals, selecting the high and low frequency signals that can represent the effective signals after decomposition, filtering out the curves carrying noise, and reconstructing the REGR curve; A deconvolution processing module for processing the reconstructed REGR curve by deconvolution method to finally obtain the REGRr curve.

Images