Logging curve reconstruction method and device based on Mallat algorithm

Abstract

The invention discloses a logging curve reconstruction method and device based on a Mallat algorithm, and relates to the technical field of logging curve reconstruction, and the key points of the technical scheme are that any one of a GR curve and an RT curve is taken as a lithologic sensitive curve, wavelet decomposition is carried out on the lithologic sensitive curve and a DT curve based on the Mallat algorithm, and the lithologic sensitive curve and the DT curve are obtained; and obtaining a lithology sensitive curve detail part and a DT curve trend part, unifying dimensions of the lithology sensitive curve detail part and the DT curve trend part, and adding the lithology sensitive curve detail part and the DT curve trend part to obtain a reconstructed DT curve. According to the method, the selection range of the GR curve, the RT curve and the DT curve is given based on the actual application scene, sandstone and mudstone in the deep area can be effectively distinguished based on the interval transit time curve, the effective reservoir inversion method is enriched, the reliability of the inversion result is improved, and the method is suitable for large-scale popularization and application. The method has a good application prospect in similar ultra-thin layer sand bodies with cretaceous systems in front of mountains of tower cars in Tarim Basin and the like, and is also suitable for sweet spot identification and reservoir space fine depiction of ultra-thin reservoirs in basins in middle and western areas.

Description

Technical Field

[0001] The present invention relates to the technical field of well logging curve reconstruction, and more particularly to a well logging curve reconstruction method and device based on Mallat algorithm. Background Art

[0002] As one of the three porosity logging curves, the acoustic transit time curve plays an important role in calculating reservoir porosity and evaluating reservoir physical properties. Furthermore, the acoustic transit time curve serves as a bridge between well logging and seismic data. Without the acoustic transit time curve, synthetic records cannot be constructed, and thus, effective matching between well logging depth domain data and seismic time domain data is impossible. Therefore, the accuracy of the acoustic transit time curve is crucial for reservoir parameter calculation in well logging evaluation and for log-constrained seismic inversion.

[0003] However, in practice, due to the influence of factors such as wellbore contamination, mud erosion, and large porosity changes, the acoustic time difference curve often cannot truly reflect the underground lithological changes, resulting in poor matching accuracy between synthetic seismic records and near-well seismic profiles, and the seismic inversion results do not match the actual drilling conditions, making reservoir prediction difficult.

[0004] When acoustic transit time curves are severely distorted, pseudo-acoustic curve construction techniques are often used to reconstruct the original acoustic transit time curves from existing stable, high-resolution logging curves that reflect changes in formation lithology. Seismic inversion is then performed on the reconstructed acoustic transit time curves. The reconstructed acoustic transit time curves not only reflect changes in formation velocity and wave impedance, but also reveal subtle differences in formation lithology. This effectively addresses the challenges of difficult lithology identification and low resolution in thin reservoirs, providing a good initial model for seismic inversion, reducing inversion ambiguity, and improving reservoir identification.

[0005] Based on the above reasons, the present invention provides a logging curve reconstruction method and device based on the Mallat algorithm, which reconstructs the original acoustic time difference curve using either the natural gamma ray curve or the true resistivity curve of the formation as a stable high-resolution logging curve reflecting the changes in the lithology of the formation. Based on the actual application scenario, the selection range of the GR curve, the RT curve and the DT curve is given to improve the accuracy and reliability of the acoustic time difference curve, reduce the multi-solution of the inversion, and improve the recognition ability of the reservoir. Summary of the Invention

[0006] To achieve the above object, the present invention provides the following technical solution: a well logging curve reconstruction method based on the Mallat algorithm, comprising the following steps:

[0007] S1. Obtain a time difference acoustic wave curve (DT) and a lithologic sensitivity curve for a well. The DT has a frequency range of 50-100 Hz. The lithologic sensitivity curve is a natural gamma ray curve (GR) with a frequency range of 27-138 Hz or a true formation resistivity curve (RT) with a frequency range of 0-12 Hz.

[0008] S2. Reconstruct the DT curve based on the following formula:

[0009]

[0010] Where Q is the reconstructed DT value, L is the DT value, H is the lithologic sensitivity value with DT dimension, H is obtained by transforming the lithologic sensitivity value through the regression function relationship between DT and lithologic sensitivity value, h is the low-pass filter coefficient, g is the high-pass filter coefficient, m and m+1 are the layers of wavelet transform, n is the position on the m+1 layer, k is the position on the m layer, and Z is an integer set.

[0011] The present invention is further configured as follows: the regression function relationship between DT and lithologic sensitivity value is as follows:

[0012] H=λ×H′+C

[0013] Where H' is the lithologic sensitivity value, H is the lithologic sensitivity value with the DT dimension, λ and C are constants, and λ and C are obtained by statistical regression analysis of the DT curve and the lithologic sensitivity curve.

[0014] The present invention is further configured such that the frequency range of the reconstructed DT curve is 61-100 Hz.

[0015] The present invention is further configured as follows: S1-S2 is performed on the sandstone well section and the mudstone well section to obtain reconstructed DT curves of the sandstone well section and the mudstone well section, and a threshold value A is obtained based on statistical analysis of the reconstructed DT curves of the sandstone well section and the mudstone well section. S1-S2 is performed on the well section to be analyzed to obtain a reconstructed DT curve of the well section to be analyzed. Parts of the well section to be analyzed with a reconstructed DT value greater than A are sandstone, and parts with a reconstructed DT value less than A are mudstone.

[0016] The present invention is further configured to draw a histogram with the abscissa being the reconstructed DT value and the ordinate being the percentage based on the reconstructed DT curves of the sandstone well section and the mudstone well section for statistical analysis to intuitively determine the threshold value A.

[0017] The present invention is further configured such that the threshold value A is 80 μs / ft.

[0018] The present invention is further configured to calculate the reconstructed P-wave impedance based on the reconstructed DT curves of the sandstone well section and the mudstone well section to obtain the reconstructed wave impedance curves of the sandstone well section and the mudstone well section, and obtain a threshold value B based on the statistical analysis of the reconstructed wave impedance curves of the sandstone well section and the mudstone well section. The reconstructed P-wave impedance is calculated based on the reconstructed DT curve of the well section to be analyzed to obtain the reconstructed wave impedance curve of the well section to be analyzed, and the portion of the well section to be analyzed where the reconstructed P-wave impedance is less than B is sandstone, and the portion where the reconstructed P-wave impedance is greater than B is mudstone.

[0019] The present invention is further configured as follows: based on the reconstructed wave impedance curves of the sandstone well section and the mudstone well section, a histogram with the horizontal axis being the reconstructed longitudinal wave impedance and the vertical axis being the percentage is drawn for statistical analysis to intuitively determine the threshold value B.

[0020] The present invention is further configured as follows: the threshold value B is 11860 kg / m 3 *m / s.

[0021] The present invention is further configured as follows: when the rock sensitivity curve is a GR curve with a 27-138 Hz range, the rock sensitivity curves of the sandstone well section and the mudstone well section are obtained, and a threshold value C is obtained based on statistical analysis of the rock sensitivity curves of the sandstone well section and the mudstone well section. The rock sensitivity curve of the well section to be analyzed is obtained, and the portion of the well section to be analyzed with a GR value less than C is sandstone, and the portion of the well section to be analyzed with a GR value greater than C is mudstone.

[0022] The present invention further comprises the following configurations: when the rock sensitivity curve is an RT curve with a range of 0-12 Hz, rock sensitivity curves for sandstone and mudstone sections are obtained, and a threshold value D is obtained based on statistical analysis of the rock sensitivity curves for the sandstone and mudstone sections. The rock sensitivity curve for the section to be analyzed is obtained; portions of the section to be analyzed with an RT value less than D are mudstone, and portions of the section to be analyzed with an RT value greater than D are sandstone.

[0023] The present invention further provides for the generation of a two-dimensional rock physics intersection map, wherein one parameter of the rock physics intersection map is the reconstructed longitudinal wave impedance and the other parameter is the rock sensitivity value. The rock physics intersection map is generated based on the reconstructed wave impedance curve and the rock sensitivity curve of the well section to be analyzed, thereby visually demonstrating the demarcation between sandstone and mudstone.

[0024] The present invention is further configured to: based on the reconstructed DT curve of the well section to be analyzed, perform random simulation guided by the seismic waveform of the well section to be analyzed to complete the phase-controlled reservoir property prediction.

[0025] The present invention is further configured such that: the random simulation is a seismic waveform indication Markov chain Monte Carlo random simulation.

[0026] The present invention is further configured such that: the lithologic sensitivity curve is a GR with a frequency range of 27-138 Hz.

[0027] The present invention also provides a well logging curve reconstruction device based on the Mallat algorithm, comprising a storage medium and a processor, wherein the storage medium stores a computer program. The processor is configured to implement the above-mentioned well logging curve reconstruction method based on the Mallat algorithm when executing the computer program.

[0028] In summary, the present invention has the following advantages over the prior art: the present invention uses either the GR curve or the RT curve as a lithologic sensitivity curve, performs wavelet decomposition on the lithologic sensitivity curve and the DT curve based on the mallat algorithm, obtains the detail portion of the lithologic sensitivity curve and the trend portion of the DT curve, converts the dimension of the detail portion of the lithologic sensitivity curve into the dimension of the DT curve, and adds the detail portion of the lithologic sensitivity curve with the trend portion of the DT curve with unified dimension to obtain a reconstructed DT curve. This method provides selection ranges for the GR curve, the RT curve, and the DT curve based on actual application scenarios, and can effectively distinguish sandstone from mudstone and gypsum mudstone in deep areas based on the acoustic time difference curve, further enriching the effective reservoir inversion method and improving the reliability of the inversion results. It has good application prospects in similar ultra-thin sand bodies such as the Cretaceous in the Takuche piedmont belt of the Tarim Basin. It is also suitable for the identification of ultra-thin reservoir sweet spots and fine characterization of reservoir space in the central and western basins. It further enriches the effective reservoir inversion method and improves the reliability of the inversion results, and has good application effects and promotion prospects. BRIEF DESCRIPTION OF THE DRAWINGS

[0029] Figure 1 Schematic diagram of the process of the embodiment.

[0030] Figure 2 is the reconstructed DT histogram used to determine the threshold value A for a certain well.

[0031] Figure 3 The reconstructed P-wave impedance histogram of a certain well is used to determine the threshold value B. DETAILED DESCRIPTION

[0032] The technical solution of the present invention will be clearly described below in conjunction with the accompanying drawings. Obviously, the described embodiments are not all embodiments of the present invention. All other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the invention.

[0033] like Figure 1As shown, it is a flow chart of a logging curve reconstruction method based on the Mallat algorithm in a preferred embodiment of the present invention. In this method, any one of the natural gamma ray curve GR and the formation true resistivity curve RT is used as a lithologic sensitivity curve, and the lithologic sensitivity curve and the acoustic time difference curve DT are respectively subjected to wavelet decomposition based on the Mallat algorithm to obtain the detail part of the lithologic sensitivity curve and the trend part of the DT curve, and the detail part of the lithologic sensitivity curve is converted into the detail part of the lithologic sensitivity curve with the dimension of the DT curve, and the detail part of the lithologic sensitivity curve with the unified dimension is added to the trend part of the DT curve to obtain the reconstructed DT curve.

[0034] The wavelet decomposition formula of the mallat algorithm is as follows:

[0035]

[0036]

[0037] Among them, E m-1,j is the trend part after decomposition, F m-1,j is the detail part after decomposition, h is the low-pass filter coefficient, g is the high-pass filter coefficient, m and m-1 are the layers of wavelet transform, j is the position on the m-1 layer, k is the position on the m layer, and Z is an integer set.

[0038] In this embodiment, the formula for converting the details of the lithologic sensitivity curve into the DT dimension is as follows:

[0039] H=λ×H′+C

[0040] Where H' is the lithologic sensitivity value, H is the lithologic sensitivity value with the DT dimension, λ and C are constants, and λ and C are obtained by statistical regression analysis of the DT curve and the lithologic sensitivity curve.

[0041] In this embodiment, the calculation formula for reconstructing the DT curve is as follows:

[0042]

[0043] Where Q is the reconstructed DT value, L is the DT value, H is the lithologic sensitivity value with DT dimension, H is obtained by transforming the lithologic sensitivity value through the regression function relationship between DT and lithologic sensitivity value, h is the low-pass filter coefficient, g is the high-pass filter coefficient, m and m+1 are the layers of wavelet transform, n is the position on the m+1 layer, k is the position on the m layer, and Z is an integer set.

[0044] In the above formula, 2∑ k∈Z (h (n-2k) L m，k ) is the trend part of the DT curve, 2∑ k∈Z (g (n-2k)H m，k ) is the detail part of the lithologic sensitivity curve with DT dimension, about g (n-2k) H m，k In this embodiment, the lithologic sensitivity curve is first decomposed by wavelet based on the mallat algorithm to calculate g (n-2k) H' m，k， Then g (n-2k) H' m，k Substitute H' into H = λ × H' + C to calculate g (n-2k) H m，k , to reduce the amount of computation.

[0045] Based on the rock physics analysis results of the reconstructed DT curve, the frequency ranges of the lithologic sensitivity curve and the DT curve were adjusted multiple times. Ultimately, it was determined that the reconstructed DT curve obtained using the above method, using a 50-100Hz DT curve and a 27-138Hz GR curve or a 0-12Hz RT curve, can effectively delineate the sandstone portion of the wellbore section. The 50-100Hz DT curve and the 27-138Hz GR curve are particularly effective. Furthermore, limiting the reconstructed DT curve frequency range to 61-100Hz further improves the delineation of the sandstone portion of the wellbore section.

[0046] The rock physical analysis includes the following means:

[0047] First, perform S1-S2 on the sandstone well section and the mudstone well section to obtain the reconstructed DT curves of the sandstone well section and the mudstone well section, and obtain the threshold value A based on the statistical analysis of the reconstructed DT curves of the sandstone well section and the mudstone well section. Perform S1-S2 on the well section to be analyzed to obtain its reconstructed DT curve, and divide its lithology based on its reconstructed DT curve with the threshold value A as the boundary. The part of the well section to be analyzed with the reconstructed DT value greater than A is sandstone, and the part with the reconstructed DT value less than A is mudstone. In order to determine the threshold value A intuitively and quickly, this embodiment draws a histogram with the horizontal axis as the reconstructed DT value and the vertical axis as the percentage based on the reconstructed DT curves of the sandstone well section and the mudstone well section for statistical analysis. The histogram divides the reconstructed DT curve into multiple intervals, and the mean in each interval is the reconstructed DT value of the interval, and the number of different reconstructed DT values in the reconstructed DT curve is converted into percentage. As Figure 2 As shown, the threshold value A is finally determined to be 80μs / ft.

[0048] On this basis, this embodiment also calculates the reconstructed longitudinal wave impedance based on the reconstructed DT curves of the sandstone well section and the mudstone well section, obtains the reconstructed wave impedance curves of the sandstone well section and the mudstone well section, and obtains the threshold value B based on the statistical analysis of the reconstructed wave impedance curves of the sandstone well section and the mudstone well section. The reconstructed longitudinal wave impedance of the well section to be analyzed is calculated based on the reconstructed DT curve of the well section to obtain the reconstructed wave impedance curve of the well section to be analyzed, and its lithology is divided based on the reconstructed wave impedance curve with the threshold value B as the boundary. The part of the well section to be analyzed with the reconstructed longitudinal wave impedance less than B is sandstone, and the part with the reconstructed longitudinal wave impedance greater than B is mudstone. In order to determine the threshold value B intuitively and quickly, this embodiment draws a histogram with the horizontal axis as the reconstructed longitudinal wave impedance and the vertical axis as the percentage based on the reconstructed wave impedance curves of the sandstone well section and performs statistical analysis. The histogram divides the reconstructed wave impedance curve into multiple intervals, and the mean value in each interval is the reconstructed longitudinal wave impedance of the interval, and the number of different reconstructed longitudinal wave impedances in the reconstructed wave impedance curve is converted into percentage. As Figure 3 As shown, the threshold value B is finally determined to be 11860kg / m 3 *m / s.

[0049] Furthermore, when the rock sensitivity curve is a 27-138 Hz GR curve, the rock sensitivity curves of the sandstone and mudstone sections can be obtained. Based on the statistical analysis of the rock sensitivity curves of the sandstone and mudstone sections, a threshold value C is obtained. The rock sensitivity curve of the section to be analyzed is obtained. Parts of the section to be analyzed with a GR value less than C are sandstone, and parts of the section to be analyzed with a GR value greater than C are mudstone. When the rock sensitivity curve is a 0-12 Hz RT curve, the rock sensitivity curves of the sandstone and mudstone sections can be obtained. Based on the statistical analysis of the rock sensitivity curves of the sandstone and mudstone sections, a threshold value D is obtained. The rock sensitivity curve of the section to be analyzed is obtained. Parts of the section to be analyzed with an RT value less than D are mudstone, and parts of the section to be analyzed with an RT value greater than D are sandstone.

[0050] Furthermore, a rock physics intersection diagram is drawn. The rock physics intersection diagram is a two-dimensional intersection diagram. One parameter of the rock physics intersection diagram is the reconstructed longitudinal wave impedance, and the other parameter is the rock sensitivity value. The rock physics intersection diagram is drawn according to the reconstructed wave impedance curve and the rock sensitivity curve of the well section to be analyzed, which intuitively reflects the division of sandstone and mudstone.

[0051] Based on the reconstructed DT curve of the well section to be analyzed, a random simulation guided by the seismic waveform of the well section to be analyzed is performed to predict the phase-controlled reservoir properties of the well section to be analyzed. The random simulation method is preferably a seismic waveform-indicated Markov chain Monte Carlo random simulation.

[0052] Taking the Bashkichik Formation as an example, the pre-reconstructed acoustic transit time curves effectively distinguish sandstone from mudstone in the top region, but the boundaries between sandstone, mudstone, and gypsum mudstone in the deeper regions are very blurred. The reconstructed DT curves can effectively distinguish sandstone from mudstone and gypsum mudstone. Because the gypsum layer is located in the Paleogene and gypsum is not developed in the Bashkichik Formation, the reconstructed DT curves of the gypsum are more distinct than those of the sandstone. The histogram obtained by adding the reconstructed DT curves to rock physics analysis software shows that the acoustic transit time for sandstone is greater than 80 μs / ft, while that for mudstone and gypsum mudstone is less than 80 μs / ft. The reconstructed wave impedance shows that the sandstone wave impedance is the smallest, while the wave impedances of mudstone and gypsum mudstone are significantly greater than those of sandstone, and the wave impedance of gypsum mudstone is greater than that of mudstone. The reconstructed wave impedance of the sandstone is less than 11,860 kg / m 3 *m / s, mudstone and gypsum mudstone wave impedance>11860kg / m 3 *m / s, with minimal overlap, essentially converging near the half-amplitude point. The reconstructed curves were added to the inversion software for waveform indication simulation inversion, and the inversion results were highly consistent with actual drilling results, demonstrating excellent results.

[0053] This embodiment further provides a well logging curve reconstruction device based on the Mallat algorithm, comprising a storage medium and a processor, wherein the storage medium stores a computer program. The processor is configured to implement the above-mentioned well logging curve reconstruction method based on the Mallat algorithm when executing the computer program.

[0054] In summary, this embodiment uses either the GR curve or the RT curve as a lithologic sensitivity curve, performs wavelet decomposition on each of the lithologic sensitivity curves and the DT curve based on the mallat algorithm, obtains the detail portion of the lithologic sensitivity curve and the trend portion of the DT curve, converts the dimension of the detail portion of the lithologic sensitivity curve into the dimension of the DT curve, and adds the detail portion of the lithologic sensitivity curve with the trend portion of the DT curve to obtain a reconstructed DT curve. This method provides selection ranges for the GR curve, the RT curve, and the DT curve based on actual application scenarios. It can effectively distinguish sandstone from mudstone and gypsum mudstone in deep areas based on the acoustic time difference curve, further enriches effective reservoir inversion methods, and improves the reliability of inversion results. It has good application prospects in similar ultra-thin sand bodies such as the Cretaceous in the Takuche piedmont belt of the Tarim Basin. It is also suitable for identifying ultra-thin reservoir sweet spots and fine characterization of reservoir space in central and western basins. It further enriches effective reservoir inversion methods, improves the reliability of inversion results, and has good application effects and promotion prospects.

[0055] The foregoing description is merely a preferred embodiment of the present invention and is not intended to limit the present invention. Those skilled in the art will readily appreciate that various modifications and variations of the present invention are possible. Any modifications, equivalent substitutions, or improvements made within the spirit and principles of the present invention are intended to be within the scope of protection of the present invention.

Claims

1. A method for reconstructing well logging curves based on the Mallat algorithm, characterized by: The steps include: S1. Obtaining a time difference curve (DT) and a lithologic sensitivity curve of a certain well; the frequency range of the DT is 50-100 Hz; the lithologic sensitivity curve is a natural gamma ray curve (GR) with a frequency range of 27-138 Hz or a true formation resistivity curve (RT) with a frequency range of 0-12 Hz; S2. Reconstruct the DT curve based on the following formula: Where Q is the reconstructed DT value, L is the DT value, H is the lithologic sensitivity value with DT dimension, H is obtained by transforming the lithologic sensitivity value through the regression function relationship between DT and lithologic sensitivity value, h is the low-pass filter coefficient, g is the high-pass filter coefficient, m and m+1 are the layers of wavelet transform, n is the position on the m+1 layer, k is the position on the m layer, and Z is an integer set.

2. The method for reconstructing well logging curves based on the Mallat algorithm according to claim 1, characterized in that: The regression function relationship between DT and lithologic sensitivity value is as follows: H=λ×H′+C Where H' is the lithologic sensitivity value, H is the lithologic sensitivity value with the DT dimension, λ and C are constants, and λ and C are obtained by statistical regression analysis of the DT curve and the lithologic sensitivity curve.

3. The method for reconstructing well logging curves based on the Mallat algorithm according to claim 1, characterized in that: The frequency range of the reconstructed DT curve is 61-100 Hz.

4. The method for reconstructing well logging curves based on the Mallat algorithm according to claim 1, characterized in that: Perform S1-S2 on the sandstone well section and the mudstone well section to obtain the reconstructed DT curves of the sandstone well section and the mudstone well section. The threshold value A is obtained based on the statistical analysis of the reconstructed DT curves of the sandstone well section and the mudstone well section. Perform S1-S2 on the well section to be analyzed to obtain the reconstructed DT curve of the well section to be analyzed. The part of the well section to be analyzed with the reconstructed DT value greater than A is sandstone, and the part with the reconstructed DT value less than A is mudstone.

5. The method for reconstructing well logging curves based on the Mallat algorithm according to claim 4, characterized in that: Based on the reconstructed DT curves of the sandstone and mudstone sections, a histogram with the horizontal axis as the reconstructed DT value and the vertical axis as the percentage was drawn for statistical analysis to intuitively determine the threshold value A.

6. The method for reconstructing well logging curves based on the Mallat algorithm according to claim 4, characterized in that: The threshold value A is 80 μs / ft.

7. The method for reconstructing well logging curves based on the Mallat algorithm according to claim 4, characterized in that: The reconstructed P-wave impedance is calculated based on the reconstructed DT curves of the sandstone and mudstone sections to obtain the reconstructed wave impedance curves of the sandstone and mudstone sections. The threshold value B is obtained based on the statistical analysis of the reconstructed wave impedance curves of the sandstone and mudstone sections. The reconstructed P-wave impedance is calculated based on the reconstructed DT curve of the section to be analyzed to obtain the reconstructed wave impedance curve of the section to be analyzed. The part of the section to be analyzed with the reconstructed P-wave impedance less than B is sandstone, and the part with the reconstructed P-wave impedance greater than B is mudstone.

8. The method for reconstructing well logging curves based on the Mallat algorithm according to claim 7, characterized in that: Based on the reconstructed wave impedance curves of the sandstone section and the mudstone section, a histogram with the horizontal axis as the reconstructed longitudinal wave impedance and the vertical axis as the percentage is drawn for statistical analysis to intuitively determine the threshold value B.

9. The method for reconstructing well logging curves based on the Mallat algorithm according to claim 7, characterized in that: The threshold value B is 11860 kg / m 3 *m / s.

10. The method for reconstructing well logging curves based on the Mallat algorithm according to claim 7, characterized in that: When the rock sensitivity curve is a GR curve with a 27-138 Hz range, obtain the rock sensitivity curves of the sandstone and mudstone sections. Statistical analysis of the rock sensitivity curves of the sandstone and mudstone sections yields a threshold value C. Obtain the rock sensitivity curve of the section to be analyzed. The section with a GR value less than C in the section to be analyzed is sandstone, and the section with a GR value greater than C is mudstone.

11. The method for reconstructing well logging curves based on the Mallat algorithm according to claim 7, characterized in that: When the rock sensitivity curve is an RT curve with a range of 0-12 Hz, obtain the rock sensitivity curves of the sandstone well section and the mudstone well section, and obtain the threshold value D based on the statistical analysis of the rock sensitivity curves of the sandstone well section and the mudstone well section. Obtain the rock sensitivity curve of the well section to be analyzed. The portion of the well section to be analyzed with an RT value less than D is mudstone, and the portion of the well section to be analyzed with an RT value greater than D is sandstone.

12. A method for reconstructing well logging curves based on the Mallat algorithm according to claim 10 or 11, characterized in that: A rock physics intersection diagram is drawn. The rock physics intersection diagram is a two-dimensional intersection diagram. One parameter of the rock physics intersection diagram is the reconstructed longitudinal wave impedance, and the other parameter is the rock sensitivity value. The rock physics intersection diagram is drawn based on the reconstructed wave impedance curve and the rock sensitivity curve of the well section to be analyzed, which intuitively reflects the division of sandstone and mudstone.

13. A method for reconstructing well logging curves based on the Mallat algorithm according to any one of claims 1 to 3, characterized in that: Based on the reconstructed DT curve of the well section to be analyzed, random simulation guided by the seismic waveform of the well section to be analyzed is carried out to complete the phase-controlled reservoir property prediction.

14. The method for reconstructing well logging curves based on the Mallat algorithm according to claim 13, characterized in that: The random simulation is a seismic waveform indication Markov chain Monte Carlo random simulation.

15. The method for reconstructing well logging curves based on the Mallat algorithm according to claim 13, characterized in that: The lithologic sensitivity curve is a GR with a frequency range of 27-138 Hz.

16. A well logging curve reconstruction device based on the Mallat algorithm, characterized by: The method comprises a storage medium and a processor, wherein the storage medium stores a computer program; the processor is configured to implement a method for reconstructing well logging curves based on the Mallat algorithm as described in any one of claims 1 to 15 when executing the computer program.

Images