Method and device for predicting mineral composition content of salt-containing stratum, electronic equipment and storage medium

Abstract

The application discloses a salt-containing stratum mineral component content prediction method and device, electronic equipment and a storage medium, and relates to the technical field of geological survey. The method comprises the following steps: acquiring target logging parameter values of at least two target logging parameters of a target salt-containing stratum in a target area; according to each target logging parameter, screening a target mineral component content prediction model from each candidate mineral component content prediction model; and using the target mineral component content prediction model to predict the target mineral component content of the target salt-containing stratum in the target area based on the target logging parameter values. The technical scheme of the embodiment of the application realizes the characterization of the complex nonlinear coupling relationship between multiple parameters in an actual stratum, and improves the prediction accuracy of the mineral component content under complex stratum conditions.

Description

Technical Field

[0001] This invention relates to the field of geological exploration technology, and in particular to a method, apparatus, electronic device and storage medium for predicting the mineral composition content of saline strata. Background Technology

[0002] In the site selection and construction of salt cavern gas storage facilities, accurate prediction of the mineral composition of saline strata is a fundamental prerequisite for tasks such as selecting reservoir sites, designing water-soluble caverns, and evaluating cavern stability. Currently, salinity prediction mainly relies on geophysical logging methods, and two representative approaches have been developed: one is the mineral composition model method, which establishes a mineral volume model and solves the logging response equation to achieve quantitative calculation of salinity; the other is the statistical regression empirical formula method, which relies on core analysis data from specific areas and interprets the relationship between logging parameters and salinity by constructing an empirical relationship.

[0003] Overall, existing methods for predicting mineral content in saline formations suffer from the following problems and shortcomings: The interpretation accuracy of mineral composition models heavily depends on the quality and completeness of well logging data, and the models themselves are significantly affected by regional geological factors, resulting in poor universality. This significantly weakens the applicability and reliability of mineral composition models when faced with incomplete well logging data or complex geological conditions. Statistical regression empirical formulas typically use linear or exponential relationships to establish mapping models between well logging parameters and salinity. While simple in form, they struggle to accurately characterize the complex nonlinear coupling relationships between multiple parameters in actual formations, leading to poor prediction performance under complex formation conditions. Summary of the Invention

[0004] This invention provides a method, apparatus, electronic device, and storage medium for predicting the mineral composition content of saline strata, which enables the characterization of complex nonlinear coupling relationships between multiple parameters in actual strata and improves the prediction accuracy of mineral composition content under complex strata conditions.

[0005] According to one aspect of the present invention, a method for predicting the mineral composition content of saline strata is provided, the method comprising: Obtain target logging parameter values ​​for at least two target logging parameters of the target saline formation in the target area; Based on the target logging parameters, the target mineral composition prediction model is selected from the candidate mineral composition prediction models. Using the target mineral composition content prediction model, the target mineral composition content of the target saline strata in the target area is predicted based on the target logging parameter values.

[0006] According to another aspect of the present invention, a device for predicting the mineral composition of saline strata is provided, the device comprising: The target logging parameter value acquisition module is used to acquire the target logging parameter values ​​of at least two target logging parameters of the target saline formation in the target area. The mineral composition content prediction model screening module is used to screen the target mineral composition content prediction model from among the candidate mineral composition content prediction models based on the target logging parameters. The target mineral composition content prediction module is used to predict the target mineral composition content of the target saline strata in the target area based on the target logging parameter values ​​using the target mineral composition content prediction model.

[0007] According to another aspect of the present invention, an electronic device is provided, the electronic device comprising: At least one processor; and A memory communicatively connected to the at least one processor; wherein, The memory stores a computer program that can be executed by the at least one processor, which enables the at least one processor to perform the method for predicting the mineral composition of saline strata according to any embodiment of the present invention.

[0008] According to another aspect of the present invention, a computer-readable storage medium is provided, the computer-readable storage medium storing computer instructions for causing a processor to execute and implement the method for predicting the mineral composition content of saline strata according to any embodiment of the present invention.

[0009] According to another aspect of the present invention, a computer program product is provided, the computer program product comprising a computer program that, when executed by a processor, implements the method for predicting the mineral composition content of saline strata according to any embodiment of the present invention.

[0010] The technical solution of this invention obtains the target logging parameter values ​​of at least two target logging parameters of the target saline formation in the target area. Based on each target logging parameter, a target mineral composition content prediction model is selected from each candidate mineral composition content prediction model. Using the target mineral composition content prediction model, the target mineral composition content of the target saline formation in the target area is predicted based on each target logging parameter value. This takes into account the differences between the logging parameters that can be obtained in different areas. By using different candidate mineral composition content prediction models, the complex nonlinear coupling relationship between multiple parameters in the actual formation is characterized, thus improving the prediction accuracy of mineral composition content under complex formation conditions.

[0011] It should be understood that the description in this section is not intended to identify key or essential features of the embodiments of the present invention, nor is it intended to limit the scope of the invention. Other features of the invention will become readily apparent from the following description. Attached Figure Description

[0012] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0013] Figure 1 This is a flowchart of a method for predicting the mineral composition of saline strata according to Embodiment 1 of the present invention; Figure 2 This is a flowchart of a method for predicting the mineral composition of saline strata according to Embodiment 2 of the present invention; Figure 3 This is a flowchart of a method for predicting the mineral composition of saline strata according to Embodiment 3 of the present invention; Figure 4 This is a graph showing the prediction results of the method for predicting the content of different mineral components provided in Embodiment 3 of the present invention; Figure 5 This is a schematic diagram of the structure of a saline stratum mineral composition prediction device provided in Embodiment 4 of the present invention; Figure 6 This is a schematic diagram of the structure of an electronic device for implementing the method for predicting the mineral composition content of saline strata according to embodiments of the present invention. Detailed Implementation

[0014] To enable those skilled in the art to better understand the present invention, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of the present invention.

[0015] It should be noted that the terms "first," "second," etc., in the specification, claims, and accompanying drawings of this invention are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of the invention described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover a non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0016] Example 1 Figure 1 This is a flowchart illustrating a method for predicting the mineral content of saline strata according to Embodiment 1 of the present invention. This embodiment of the invention is applicable to situations where the mineral content of complex saline strata needs to be predicted. The method can be executed by a saline strata mineral content prediction device, which can be implemented in hardware and / or software and can be configured in an electronic device that performs the function of predicting the mineral content of saline strata.

[0017] See Figure 1 The methods for predicting the mineral composition of saline strata shown include: S101. Obtain the target logging parameter values ​​of at least two target logging parameters for the target saline formation in the target area.

[0018] The target area is the region for real-time monitoring of the mineral composition of saline formations. The target area is a geographical or subsurface space with clearly defined boundaries, used to define subsequent saline formation identification, well logging curve analysis, and parameter extraction. Optionally, at least one well log may be included within the target area.

[0019] A target saline formation refers to a stratum containing layers of salt minerals such as rock salt, whose mineral composition is monitored in real time. Within the target area, a target saline formation is a stratigraphic unit containing salt minerals (including rock salt, gypsum, and sodium sulfate) identified based on lithological identification, logging data, or well logging response characteristics. This stratigraphic unit has identifiable top and bottom interfaces and serves as the geological object for subsequent well logging curve analysis and parameter extraction. Target saline formations may include salt rocks, mudstones, gypsum, and other evaporites.

[0020] Target logging parameters are parameters used to characterize the formation properties of a target saline formation detected during logging operations in a target area. Target logging parameters are categories of target logging curves or target logging parameter values. For example, target logging parameters may include at least two of the following: natural gamma logging parameters, compensated neutron logging parameters, density logging parameters, and sonic transit time logging parameters.

[0021] The target logging curve refers to a continuous or discrete logging response sequence obtained from the depth range of the target saline formation, reflecting the formation's physical properties. The logging response includes, but is not limited to, at least two of natural gamma ray, sonic transit time, compensated neutron, and density. The target logging curve is used to extract or calculate the target logging parameter values. Optionally, the target logging curve may include the original target logging curve or the processed target logging curve.

[0022] The target logging parameter value is the specific numerical value of the target logging parameter. The target logging parameter value refers to the quantified value obtained by extracting from the target logging curve. Referring to the example above, the target logging parameter value can include at least two of the following: natural gamma logging value, compensated neutron logging value, density logging value, and sonic transit time logging value.

[0023] In an optional example, during or after drilling in the target area, the target raw curves of at least two target logging parameters of the target saline formation in the target area are obtained, and the target logging curves are extracted to obtain the target logging parameter values.

[0024] In an optional example, after obtaining the target raw logging curves for at least two target logging parameters of the target saline formation in the target area, the target raw logging curves can be denoised and their resolution improved to obtain the target processed logging curves. The target logging parameter values ​​can then be extracted from these processed logging curves. The target processed logging curves are the result of denoising and resolving the target raw logging curves.

[0025] Specifically, wavelet transform formulas can be used to perform wavelet transform on the original target logging curve to obtain the target denoised logging curve. The target denoised logging curve is the denoised result of the original target logging curve.

[0026] By performing wavelet transform on the original target logging curve, a denoised target logging curve is obtained. This removes the spike noise of the original target logging curve while preserving its waveform characteristics, thus improving the signal-to-noise ratio. Furthermore, by performing wavelet reconstruction on the denoised target logging curve, a processed target logging curve is obtained, which improves the resolution of the denoised target logging curve. This results in clearer boundaries, more accurate amplitudes, and more precise values ​​for the processed target logging curve, thereby enhancing the prediction effect of mineral composition in saline formations.

[0027] For example, the wavelet transform formula can be expressed using the following formula: ; In the formula, Noise reduction logging curve for the target; The original logging curve for the target; It is the mother wavelet The complex conjugate; This indicates the signal sampling interval of the target's original logging curve; As a scaling variable, it is used to control the target's original logging curve. Scaling (frequency characteristics); This is a translation variable used to control the target's original logging curve. Translation (time domain characteristics).

[0028] Specifically, the sym5 wavelet (5th-order approximate symmetric wavelet) from the Symlet wavelet family can be selected for wavelet decomposition of the target denoising logging curve. The decomposition layer number is set to 3. The target denoising low-frequency signal, the target denoising two-layer decomposition high-frequency signal, and the target denoising three-layer decomposition high-frequency signal are extracted from the target denoising logging curve. This allows for the detection of the development status and mineral composition changes of saline strata in the target area. The target denoising low-frequency signal is the low-frequency signal of the target denoising logging curve. The target denoising two-layer decomposition high-frequency signal is the high-frequency signal of the second-layer decomposition of the target denoising logging curve. The target denoising three-layer decomposition high-frequency signal is the high-frequency signal of the third-layer decomposition of the target denoising logging curve.

[0029] Specifically, wavelet reconstruction formulas can be used to reconstruct the wavelet decomposition results of the target noise reduction logging curve using wavelet reconstruction to obtain the target processed logging curve.

[0030] For example, when the saline strata in the target area are stably developed and the rate of change of mineral composition content is less than or equal to the preset threshold for change of mineral composition content, the target noise-reduced low-frequency signal, the target noise-reduced two-layer decomposition high-frequency signal, and the target noise-reduced three-layer decomposition high-frequency signal are superimposed based on the first wavelet reconstruction formula to obtain the first target processing logging curve.

[0031] Among them, the first target processing logging curve is the target processing logging curve obtained by wavelet reconstruction when the saline strata in the target area are stably developed and the rate of change of mineral composition content is less than or equal to the preset mineral composition content change threshold.

[0032] The first wavelet reconstruction formula can be represented by the following formula: ; In the formula, Process the logging curves for the target; The low-frequency signal of the target noise reduction logging curve; High-frequency signals from two layers of noise-reducing logging curves; The high-frequency signal of the three-layer decomposition of the target noise reduction logging curve.

[0033] For example, when the saline strata in the target area have interlayers and the rate of change of mineral composition content is greater than the preset threshold for change of mineral composition content, the three-layer decomposition weights of the target noise reduction three-layer decomposition high-frequency signal are obtained in advance. Based on the second wavelet reconstruction formula, the three-layer decomposition weights are used to superimpose the target noise reduction low-frequency signal, the target noise reduction two-layer decomposition high-frequency signal, and the target noise reduction three-layer decomposition high-frequency signal to obtain the second target processing logging curve.

[0034] Among them, the second target processing logging curve is the target processing logging curve obtained by wavelet reconstruction when the salt-bearing strata in the target area have well-developed interlayers and the rate of change of mineral composition content is greater than the preset mineral composition content change threshold.

[0035] The second wavelet reconstruction formula can be represented by the following formula: ; In the formula, Process the logging curves for the target; The low-frequency signal of the target noise reduction logging curve; High-frequency signals from two layers of noise-reducing logging curves; Three-layer decomposition of high-frequency signals for noise reduction of logging curves; The weights are determined by a three-level decomposition.

[0036] By extracting the target noise-reduced low-frequency signal, the target noise-reduced two-layer decomposition high-frequency signal, and the target noise-reduced three-layer decomposition high-frequency signal from the target noise-reduced logging curve, the development status and mineral composition content change status of the saline strata in the target area are detected. Based on whether the saline strata in the target area are stably developed and the rate of change of mineral composition content is less than or equal to the preset mineral composition content change threshold, or whether the saline strata in the target area have interlayer development and the rate of change of mineral composition content is greater than the preset mineral composition content change threshold, a first target-processed logging curve or a second target-processed logging curve is generated differently. This takes into account the development status and mineral composition content change status of the saline strata, which can eliminate noise signals, improve the resolution of logging curves at lithological change interfaces, and enhance the resolution of thin-layer salt rocks, thereby improving the flexibility and accuracy of the target-processed logging curve.

[0037] S102. Based on the logging parameters of each target, select the target mineral composition content prediction model from among the candidate mineral composition content prediction models.

[0038] The candidate mineral content prediction model is a pre-trained model for predicting mineral content. Specifically, it represents the data relationship between well logging parameter combinations and mineral content. Different well logging parameter combinations correspond to different candidate mineral content prediction models. Optionally, the mapping relationship between well logging parameter combinations and candidate mineral content prediction models can be predetermined and stored in the device. The input data for the candidate mineral content prediction model is the combination of well logging parameters, and the output is the predicted mineral content. Therefore, based on each target well logging parameter, a target mineral content prediction model can be selected from the candidate models. The target mineral content prediction model is the candidate mineral content prediction model corresponding to the well logging parameter combination for each target well logging parameter.

[0039] Specifically, based on the combination of logging parameters for each target logging parameter, the mapping relationship between the logging parameter combination and the candidate mineral composition content prediction model is queried, and the target mineral composition prediction model is selected from among the candidate mineral composition content prediction models.

[0040] S103. Using a target mineral composition content prediction model, based on the logging parameter values ​​of each target, the target mineral composition content of the target saline strata in the target area is predicted.

[0041] The target mineral composition content refers to the predicted mineral composition content of the target saline strata in the target area. For example, the target mineral composition may include salt, insoluble matter, mirabilite, and gypsum. Correspondingly, the target mineral composition content may include salinity, insoluble matter content, mirabilite content, and gypsum content.

[0042] Specifically, the target logging parameter values ​​can be input into the target mineral composition content prediction model to predict the target mineral composition content of the target saline strata in the target area.

[0043] The technical solution of this invention obtains the target logging parameter values ​​of at least two target logging parameters of the target saline formation in the target area. Based on each target logging parameter, a target mineral composition content prediction model is selected from each candidate mineral composition content prediction model. Using the target mineral composition content prediction model, the target mineral composition content of the target saline formation in the target area is predicted based on each target logging parameter value. This takes into account the differences between the logging parameters that can be obtained in different areas. By using different candidate mineral composition content prediction models, the complex nonlinear coupling relationship between multiple parameters in the actual formation is characterized, thus improving the prediction accuracy of mineral composition content under complex formation conditions.

[0044] Example 2 Figure 2This is a flowchart illustrating a method for predicting the mineral composition content of saline formations according to Embodiment 2 of the present invention. Based on the above embodiments, this embodiment further adds the following steps before "obtaining the target logging parameter values ​​of at least two target logging parameters for the target saline formation in the sample area": ​​"obtaining the original logging curves of at least two sample logging parameters sensitive to mineral composition content in the sample area and the core measured mineral composition content dataset; denoising and improving the resolution of the original logging curves to obtain processed logging curves; sampling the processed logging curves based on the core recording depth of the core measured mineral composition content dataset for a single sample logging parameter to obtain a sample logging parameter dataset; and constructing a probability relationship between mineral composition content and a single sample logging parameter based on the core measured mineral composition content dataset and the sample logging parameter dataset for a single sample logging parameter." The algorithm generates a mineral composition content prediction sub-model corresponding to a single sample logging parameter based on the probability density function between mineral composition content and individual sample logging parameters, as well as the sample logging parameter dataset. It then obtains the logging parameter weights for each sample logging parameter and uses these weights to fuse the corresponding mineral composition content prediction sub-models, resulting in a candidate mineral composition content prediction model. By reducing noise, improving resolution, and adjusting the probability density function, a highly robust multi-logging parameter weighted fusion candidate mineral composition content prediction model is constructed. This model effectively overcomes the influence of logging data quality fluctuations under complex geological conditions with frequent changes in lithology and mineral composition content, achieving probabilistically optimal mineral composition content prediction and improving the prediction accuracy. It should be noted that parts not detailed in this embodiment can be found in other embodiments.

[0045] See Figure 2 The methods for predicting the mineral composition of saline strata shown include: S201. Obtain the original logging curves of at least two sample logging parameters that are sensitive to mineral composition content in the sample area, as well as the core measured mineral composition content dataset.

[0046] The sample region serves as a reference area when constructing the candidate mineral composition content prediction model. Optionally, the sample region may include at least one well log.

[0047] Sample logging parameters are parameters used to characterize the formation properties of a sample saline formation detected during logging operations in a sample area. The sample saline formation refers to a formation containing layers of salt minerals such as rock salt, for which mineral composition analysis is performed. Sample saline formations may include salt rocks, mudstones, gypsum, and other evaporites. For example, sample logging parameters may include at least two of the following: natural gamma logging parameters, compensated neutron logging parameters, density logging parameters, and sonic transit time logging parameters.

[0048] The original logging curves of the sample are used to characterize the trend of the sample logging parameter values ​​as the logging depth changes.

[0049] The core measured mineral content dataset is a discrete dataset of core measured mineral content corresponding to different core recording depths. Specifically, it represents the mineral content of a core sample at a given core recording depth detected during drilling in a saline formation within the sample area.

[0050] The sample logging parameter curves and core measured mineral composition datasets can be pre-determined and stored in the logging data of the sample area.

[0051] Specifically, the original logging curves of at least two sample logging parameters that are sensitive to mineral composition content are obtained from the logging data of the sample area, along with the core measured mineral composition content dataset.

[0052] Optionally, the integrity of the logging data for the sample area is checked. When the logging data for the sample area is complete, the original logging curves for the sample are obtained from the logging data of the sample area, including natural gamma-ray logging curves, compensated neutron logging curves, density logging curves, and sonic transit-time logging curves, which are sensitive to mineral composition content. When the logging data for the sample area is incomplete, at least two of the following are obtained from the logging data of the sample area: natural gamma-ray logging curves, compensated neutron logging curves, density logging curves, and sonic transit-time logging curves, which are sensitive to mineral composition content, to obtain the original logging curves for the sample. Therefore, based on the differences in logging data from different sample areas, multiple logging parameters can be dynamically adjusted to generate different candidate mineral composition content prediction models, ensuring accurate prediction of mineral composition content under different regions, different stratigraphic levels, and different logging data source conditions.

[0053] In an optional embodiment of the present invention, after obtaining the original logging curves of at least two sample logging parameters that are sensitive to the mineral composition content in the sample area and the core measured mineral composition content dataset, the method further includes: detecting whether the original logging curve of the sample belongs to the wellbore enlargement section; and removing the original logging curve of the sample if it belongs to the wellbore enlargement section.

[0054] In the enlarged section of the wellbore, logging instruments may move away from the wellbore, leading to distortion of the detected logging parameters. Therefore, it is necessary to discard the original logging curves of samples from the enlarged section to ensure the prediction accuracy of the candidate mineral composition content prediction model.

[0055] Specifically, it can detect which type of original logging curve a sample belongs to: natural gamma logging curve, compensated neutron logging curve, density logging curve, or sonic transit time logging curve.

[0056] Specifically, for the natural gamma ray logging curve, does the statistical noise of the natural gamma ray logging curve double? If so, then the natural gamma ray logging curve belongs to the wellbore enlargement zone; if not, then the natural gamma ray logging curve does not belong to the wellbore enlargement zone.

[0057] Specifically, for compensated neutron logging curves, the variation degree of the shale segment in the compensated neutron logging curve is checked to see if it exceeds a preset shale segment variation threshold. If so, the compensated neutron logging curve is determined to belong to the wellbore enlargement zone; if not, it is determined not to belong to the wellbore enlargement zone. The preset shale segment variation threshold is a pre-set upper limit for the shale segment variation degree of a normal compensated neutron logging curve. This preset shale segment variation threshold is used to measure whether the compensated neutron logging curve belongs to the wellbore enlargement zone. For example, the preset shale segment variation threshold can be -2%. Specifically, for density logging curves, the cycle jump rate of the density logging curve is checked to see if it exceeds a preset cycle jump rate threshold. If it does, the density logging curve is determined to belong to the wellbore enlargement field; if not, it is determined not to belong to the wellbore enlargement field. The preset cycle jump rate threshold is a pre-defined upper limit for the cycle jump rate of a normal density logging curve. This threshold is used to determine whether the density logging curve belongs to the wellbore enlargement section. For example, the preset cycle jump rate threshold is 15%.

[0058] Specifically, for the sonic transit time logging curve, the change in the shale segment of the curve is checked to see if it exceeds a preset threshold. If so, the sonic transit time logging curve is determined to belong to the wellbore enlargement field; otherwise, it is determined not to belong to the wellbore enlargement field. The preset threshold for shale segment change is a pre-set upper limit for the shale segment change in a normal sonic transit time logging curve. This threshold is used to determine whether the sonic transit time logging curve belongs to the wellbore enlargement zone. For example, the preset threshold for shale segment change can be 2.2.

[0059] Specifically, when the original logging curve of the sample is detected to belong to the wellbore enlargement section, the original logging curve of the sample is removed.

[0060] This approach eliminates the original well logging curves of a sample when they are detected to belong to a wellbore enlargement zone, thus ensuring the accuracy of the original well logging curves and consequently guaranteeing the prediction accuracy of the candidate mineral composition content prediction model.

[0061] S202. Noise reduction and resolution enhancement are performed on the original logging curves of the sample to obtain the processed logging curves of the sample.

[0062] The sample-processed logging curves are the results of noise reduction and resolution improvement of the original logging curves.

[0063] Specifically, moving average filtering, median filtering, least squares smoothing filtering, or wavelet transform can be used to denoise the original well logging curves of the sample. Deconvolution, wavelet reconstruction, spectral bluening, or machine learning can be used to improve the resolution of the denoised original well logging curves of the sample, resulting in processed well logging curves.

[0064] In an optional embodiment of the present invention, the original sample logging curve is denoised and its resolution is improved to obtain a processed sample logging curve, including: performing wavelet transform on the original sample logging curve to obtain a denoised sample logging curve; and performing wavelet reconstruction on the denoised sample logging curve to obtain the processed sample logging curve.

[0065] The sample denoised logging curve is the result of denoising the original logging curve of the sample.

[0066] Specifically, wavelet transform formulas can be used to perform wavelet transform on the original logging curves of the sample to obtain the denoised logging curves of the sample.

[0067] For example, the wavelet transform formula can be represented by the following formula: ; In the formula, For the sample noise reduction logging curve; The original well logging curve for the sample; It is the mother wavelet The complex conjugate; This indicates the signal sampling interval of the original well logging curve of the sample; As a scaling variable, it is used to control the original logging curves of the sample. Scaling (frequency characteristics); This is a translation variable used to control the original logging curve of the sample. Translation (time domain characteristics).

[0068] Specifically, the sym5 wavelet (5th-order approximate symmetric wavelet) from the Symlet wavelet family can be selected for wavelet decomposition of the sample denoising logging curves, with a decomposition level of 3. Wavelet reconstruction formulas can then be used to reconstruct the wavelet decomposition results of the sample denoising logging curves, yielding the processed logging curves.

[0069] For example, the wavelet reconstruction formula can be represented by the following formula: ; In the formula, Process the well logging curves for the sample; The low-frequency signal of the sample noise-reduced logging curve; The high-frequency signal of the second-layer decomposition of the sample noise-reduced logging curve; The high-frequency signal of the three-layer decomposition of the sample noise-reduced logging curve.

[0070] This scheme obtains denoised logging curves by performing wavelet transform on the original sample logging curves. This removes spike noise from the original sample logging curves while preserving their waveform characteristics, thus improving the signal-to-noise ratio. Furthermore, wavelet reconstruction of the denoised logging curves yields processed logging curves, which improves their resolution, resulting in clearer boundaries, more accurate amplitudes, and more precise values. This enhances the prediction of mineral composition in saline formations.

[0071] In an optional embodiment of the present invention, wavelet reconstruction is performed on the sample denoised logging curve to obtain the sample processed logging curve, including: extracting the sample denoised low-frequency signal, the sample denoised two-layer decomposition high-frequency signal, and the sample denoised three-layer decomposition high-frequency signal from the sample denoised logging curve; detecting the development status and mineral composition content change status of the saline strata in the sample area; when the saline strata in the sample area are stably developed and the rate of change of mineral composition content is less than or equal to a preset mineral composition content change threshold, superimposing the sample denoised low-frequency signal, the sample denoised two-layer decomposition high-frequency signal, and the sample denoised three-layer decomposition high-frequency signal to obtain the first sample processed logging curve; when the saline strata in the sample area have interlayer development and the rate of change of mineral composition content is greater than a preset mineral composition content change threshold, obtaining the three-layer decomposition weight of the sample denoised three-layer decomposition high-frequency signal, and using the three-layer decomposition weight, superimposing the sample denoised low-frequency signal, the sample denoised two-layer decomposition high-frequency signal, and the sample denoised three-layer decomposition high-frequency signal to obtain the second sample processed logging curve.

[0072] The sample denoising low-frequency signal is the low-frequency signal of the sample denoising logging curve. The sample denoising two-layer decomposition high-frequency signal is the high-frequency signal of the sample denoising logging curve after two-layer decomposition. The sample denoising three-layer decomposition high-frequency signal is the high-frequency signal of the sample denoising logging curve after three-layer decomposition.

[0073] The developmental state is used to characterize the lithology of saline strata. Stable development of saline strata in the sample area can be understood as the strata not being interrupted by tectonic movements or dissolved, and having homogeneous lithology. Development of interlayers in the saline strata of the sample area can be understood as the presence of numerous thin or medium layers of non-salt materials (such as mudstone, shale, gypsum, and carbonates) within the predominantly salt-rock strata. These non-salt layers act as interlayers separating the upper and lower salt layers, breaking the originally continuous salt layers into multiple salt-non-salt interbedded segments.

[0074] The mineral composition content variation status is used to characterize the degree of change in the mineral composition content of salt rocks. A preset mineral composition content variation threshold is used to measure whether the change in the mineral composition content of salt rocks is significant. The preset mineral composition content variation threshold is a pre-defined lower limit value for the mineral composition content when it is considered significant. A mineral composition content variation rate less than or equal to the preset mineral composition content variation threshold can be interpreted as a significant change in the mineral composition content of the salt rocks. A mineral composition content variation rate greater than the preset mineral composition content variation threshold can be interpreted as a minor change in the mineral composition content of the salt rocks. The three-layer decomposition weight is used to enhance the resolution of thin-layer salt rocks. The three-layer decomposition weight is greater than 1.

[0075] The first sample-processed logging curve is the wavelet-reconstructed logging curve obtained when the saline strata in the sample area are stably developed and the rate of change of mineral composition is less than or equal to a preset mineral composition change threshold. The second sample-processed logging curve is the wavelet-reconstructed logging curve obtained when the saline strata in the sample area have interlayered formations and the rate of change of mineral composition is greater than a preset mineral composition change threshold.

[0076] Specifically, the sym5 wavelet from the Symlet wavelet family can be selected for wavelet decomposition of the sample denoised logging curves. The decomposition layer number is set to 3. This extracts the sample denoised low-frequency signal, the sample denoised high-frequency signal from the second-layer decomposition, and the sample denoised high-frequency signal from the third-layer decomposition. This allows for the detection of the development status and mineral composition changes of saline strata in the sample area.

[0077] Specifically, when the saline strata in the sample area are stably developed and the rate of change of mineral composition content is less than or equal to the preset threshold for change of mineral composition content, the sample denoised low-frequency signal, the sample denoised two-layer decomposition high-frequency signal, and the sample denoised three-layer decomposition high-frequency signal are superimposed based on the first wavelet reconstruction formula to obtain the first sample processed logging curve.

[0078] For example, the first wavelet reconstruction formula can be represented by the following formula: ; In the formula, Process the well logging curves for the sample; The low-frequency signal of the sample noise-reduced logging curve; The high-frequency signal of the second-layer decomposition of the sample noise-reduced logging curve; The high-frequency signal of the three-layer decomposition of the sample noise-reduced logging curve.

[0079] Specifically, when the saline strata in the sample area have interlayers and the rate of change of mineral content is greater than the preset threshold for change of mineral content, the three-layer decomposition weights of the sample noise reduction three-layer decomposition high-frequency signal are obtained in advance. Based on the second wavelet reconstruction formula, the three-layer decomposition weights are used to superimpose the sample noise reduction low-frequency signal, the sample noise reduction two-layer decomposition high-frequency signal, and the sample noise reduction three-layer decomposition high-frequency signal to obtain the second sample processing logging curve.

[0080] For example, the second wavelet reconstruction formula can be represented by the following formula: ; In the formula, Process the well logging curves for the sample; The low-frequency signal of the sample noise-reduced logging curve; The high-frequency signal of the second-layer decomposition of the sample noise-reduced logging curve; The high-frequency signal of the three-layer decomposition of the sample noise-reduced logging curve; The weights are determined by a three-level decomposition.

[0081] This scheme extracts low-frequency signals, high-frequency signals from two-layer decomposition, and high-frequency signals from three-layer decomposition of sample denoised logging curves to detect the development status and mineral content variation status of saline strata in the sample area. Based on whether the saline strata in the sample area are stably developed and the rate of change in mineral content is less than or equal to a preset threshold, or whether the saline strata in the sample area have interlayer development and the rate of change in mineral content is greater than a preset threshold, a first-sample-processed logging curve or a second-sample-processed logging curve is generated differentially. This approach considers both the development status and mineral content variation status of the saline strata, eliminating noise signals, improving the resolution of logging curves at lithological change interfaces, and enhancing the resolution of thin-layered salt rocks, thereby improving the flexibility and accuracy of the sample-processed logging curves.

[0082] S203. For a single sample logging parameter, the sample processing logging curve is sampled based on the core recording depth of the core measured mineral composition content dataset to obtain the sample logging parameter dataset.

[0083] The sample logging parameter dataset is a discrete dataset of sample logging parameter values ​​corresponding to different core recording depths.

[0084] Specifically, for a single sample logging parameter, the sample processing logging curve is sampled according to the core recording depth of the core measured mineral composition content dataset, and the sample logging parameter values ​​corresponding to each core recording depth are extracted to obtain the sample logging parameter dataset.

[0085] In an optional embodiment of the present invention, before sampling the sample-processed logging curve based on the core recording depth of the core measured mineral content dataset for a single sample logging parameter, the method further includes: adjusting the core recording depth of the core measured mineral content to obtain an adjusted core recording depth; detecting the degree of matching between the sample logging parameter value with the core measured mineral content at the same logging recording depth as the adjusted core recording depth; and correcting the core recording depth of the core measured mineral content dataset based on the degree of matching.

[0086] Core recording depth refers to the depth at which the measured mineral composition of the core is recorded during the drilling process. Core recording depth is the original depth position of that core segment within the wellbore. Logging recording depth is the well depth detected by the logging instrument during the logging process. Logging core recording depth refers to the depth position corresponding to the logging instrument's response signal recorded during logging operations. Because there is a difference between the core recording depth corresponding to the drill bit during drilling and the logging recording depth detected by the logging instrument during the logging process, it is necessary to correct the core recording depth corresponding to the drill bit during drilling to ensure a unified depth benchmark between core measurement data and logging data. Adjusting the core recording depth is the result of adjusting the core recording depth for the measured mineral composition of the core. There is a correspondence between the measured mineral composition of the core at the same depth and the sample logging parameter values. Specifically, high mineral composition content corresponds to low natural gamma logging values, low compensated neutron logging values, low density logging values, and high sonic transit time logging values.

[0087] Specifically, the core recording depth for the measured mineral composition content is adjusted according to a preset depth adjustment step size to obtain the adjusted core recording depth. The degree of matching between the sample logging parameter values ​​corresponding to the adjusted core recording depth and the measured mineral composition content is detected. Based on the adjusted core recording depth corresponding to the maximum matching degree, the corresponding core recording depth for the measured mineral composition content is updated. The preset depth adjustment step size characterizes the degree of a single adjustment to the pre-defined core recording depth for the measured mineral composition content. For example, the preset depth adjustment step size can be 1 meter.

[0088] This scheme adjusts the core recording depth for measured mineral composition content in core samples to obtain an adjusted core recording depth. It then tests the matching degree between the sample logging parameter values ​​at the same adjusted core recording depth and the measured mineral composition content in the core samples. Based on the matching degree, the core recording depth of the measured mineral composition content dataset is corrected. This approach considers the differences between the core recording depth corresponding to the drill bit during drilling and the logging depth detected by the instrument during logging. By correcting the core recording depth corresponding to the drill bit during drilling, a unified depth benchmark is ensured between the core measurement data and the logging data, thus improving the accuracy of mineral composition content detection in saline formations.

[0089] S204. For a single sample logging parameter, construct a probability density function between the mineral composition content and the single sample logging parameter based on the core measured mineral composition content dataset and the sample logging parameter dataset.

[0090] The probability density function is the probability density function of the mineral composition content corresponding to the logging parameters of a single sample.

[0091] Specifically, for a single sample logging parameter, the mean calculation formula is used to calculate the average mineral content based on the core measured mineral content dataset.

[0092] For example, the following formula can be used to represent the dataset of measured mineral composition from core samples: ; In the formula, is the dataset of measured mineral composition from rock cores; n is the total amount of data in the dataset of measured mineral composition from rock cores.

[0093] For example, the following formula can be used to represent the sample well logging parameter dataset: ; In the formula, is the sample logging parameter dataset for the j-th sample logging parameter; n is the total amount of data in the sample logging parameter dataset for the j-th sample logging parameter.

[0094] For example, the average mineral content can be calculated using the following formula: ; In the formula, This represents the average content of mineral components. denoted as the measured mineral composition content of the i-th core in the measured mineral composition content dataset; n represents the total amount of data in the measured mineral composition dataset.

[0095] Specifically, the mean calculation formula is used to calculate the average value of the logging parameters based on the sample logging parameter dataset.

[0096] For example, the following formula can be used to calculate the average value of well logging parameters: ; In the formula, The average logging parameters of the j-th sample are denoted as . is the value of the i-th sample logging parameter in the sample logging parameter dataset of the j-th sample logging parameter; n is the total amount of data in the sample logging parameter dataset of the j-th sample logging parameter.

[0097] Specifically, the standard deviation calculation formula is used to calculate the standard deviation of mineral composition content based on the measured mineral composition content dataset and the average value of mineral composition content in the core sample.

[0098] For example, the standard deviation of mineral component content can be calculated using the following formula: ; In the formula, This represents the standard deviation of mineral content. This represents the average content of mineral components. denoted as the measured mineral composition content of the i-th core in the measured mineral composition content dataset; n represents the total amount of data in the measured mineral composition dataset.

[0099] Specifically, the standard deviation calculation formula is used to calculate the standard deviation of the logging parameters based on the sample logging parameter dataset and the average value of the logging parameters.

[0100] For example, the standard deviation of well logging parameters can be calculated using the following formula: ; In the formula, Let be the standard deviation of the logging parameters for the j-th sample. The average logging parameters of the j-th sample are denoted as . Let be the value of the i-th sample logging parameter in the sample logging parameter dataset for the j-th sample logging parameter; n is the total amount of data in the sample logging parameter dataset.

[0101] Specifically, the correlation coefficient calculation formula is used to calculate the correlation coefficient between mineral composition content and well logging parameters based on the core measured mineral composition content dataset, sample well logging parameter dataset, average mineral composition content, average well logging parameter, sum of squares of mineral composition content deviation, and sum of squares of well logging parameter deviation.

[0102] For example, the correlation coefficient between mineral composition content and logging parameters can be calculated using the following formula: ; In the formula, is the correlation coefficient between the mineral composition content and the logging parameters of the j-th sample; This represents the measured mineral composition content of the i-th core in the dataset of measured mineral composition content. This represents the average content of mineral components. Let i be the value of the i-th sample logging parameter in the sample logging parameter dataset for the j-th sample logging parameter. The average logging parameters of the j-th sample are denoted as . This represents the sum of squares of the deviations in mineral component content; is the sum of squared deviations of the logging parameters for the j-th sample logging parameter; n is the total amount of data in the core measured mineral composition dataset or the sample logging parameter dataset for the j-th sample logging parameter.

[0103] Specifically, for the logging parameters of a single sample, the sum of squares of deviations is calculated using the formula for calculating the sum of squares of deviations. This is based on the measured mineral content of each core and the average value of the mineral content in the core measured mineral content dataset.

[0104] For example, the following formula for calculating the sum of squares of deviations can be used to calculate the sum of squares of deviations in mineral composition content: ; In the formula, This represents the sum of squares of the deviations in mineral component content; This represents the measured mineral composition content of the i-th core in the dataset of measured mineral composition content. denoted as the average mineral composition content; n represents the total amount of data in the core sample mineral composition dataset.

[0105] Specifically, for a single sample logging parameter, the sum of squared deviations formula is used to calculate the sum of squared deviations of the logging parameters based on the values ​​of each sample logging parameter and the average content of logging parameters in the sample logging parameter dataset.

[0106] For example, the following formula for calculating the sum of squares of deviations can be used to calculate the sum of squares of well logging parameter deviations: ; In the formula, This represents the sum of squares of the well logging parameter deviations; This represents the i-th sample logging parameter value in the sample logging parameter dataset. is the average value of the logging parameters; n is the total amount of data in the sample logging parameter dataset for the j-th sample logging parameter.

[0107] Specifically, for a single sample logging parameter, the probability density function formula between mineral composition content and a single logging parameter is adopted. Based on the average value of mineral composition content, the average value of logging parameters, the standard deviation of mineral composition content, the standard deviation of logging parameters, and the correlation coefficient between mineral composition content and logging parameters, the probability density function between mineral composition content and a single logging parameter is constructed.

[0108] For example, the probability density function formula for the relationship between mineral composition content and a single logging parameter can be used: ; In the formula, Let be the probability density function between the mineral composition content and the logging parameters of the j-th sample. This represents the standard deviation of mineral content. Let be the standard deviation of the logging parameters for the j-th sample. is the correlation coefficient between the mineral composition content and the logging parameters of the j-th sample; This refers to the content of mineral components; This represents the average content of mineral components. The average logging parameters of the j-th sample are denoted as . Let be the logging parameter value of the j-th sample.

[0109] S205. Based on the probability density function between mineral composition content and single sample logging parameters, and the sample logging parameter dataset, generate a mineral composition content prediction sub-model corresponding to a single sample logging parameter.

[0110] The mineral composition content prediction sub-model corresponding to the sample logging parameters is used to predict the mineral composition content based on the logging parameter values ​​corresponding to the sample logging parameters.

[0111] Specifically, based on the probability density function between mineral composition content and logging parameters, and the sample logging parameter dataset, the mineral composition content corresponding to the sample logging parameters under the highest probability condition is calculated, thus obtaining a mineral composition content prediction sub-model for a single sample logging parameter.

[0112] For example, the following formula can be used to calculate the mineral composition content corresponding to the sample logging parameters under the highest probability condition, thus obtaining the mineral composition content prediction sub-model for a single sample logging parameter: ; In the formula, The mineral composition content is predicted based on the value of the i-th logging parameter corresponding to the j-th sample logging parameter. Let be the probability density function relating the mineral composition content to the i-th logging parameter value of the j-th sample.

[0113] S206. Obtain the logging parameter weights of each sample logging parameter, and use the logging parameter weights to fuse the mineral composition content prediction sub-models corresponding to each sample logging parameter to obtain the candidate mineral composition content prediction model.

[0114] The logging parameter weights are the weights of the sample logging parameters in the candidate mineral composition content prediction model. Optionally, the logging parameter weights can be preset and adjusted by technicians based on experience.

[0115] Specifically, predetermined logging parameter weights can be obtained. These weights can then be used to weight and fuse the mineral composition content prediction sub-models corresponding to the logging parameters of each sample, resulting in a candidate mineral composition content prediction model.

[0116] For example, the following weighted fusion formula can be used to weight and fuse the mineral composition content prediction sub-models corresponding to the well logging parameters of each sample to obtain the candidate mineral composition content prediction model: ; In the formula, The final output of the candidate mineral composition content prediction model is the mineral composition content. Let be the logging parameter weights for the j-th sample; The mineral composition content is predicted based on the logging parameter values ​​of the j-th sample. Let be the logging parameter weights for the m-th sample. The mineral composition content is predicted based on the logging parameter value of the m-th sample; where m is the total number of sample logging parameters corresponding to the input data of the candidate mineral composition content prediction model.

[0117] In an optional embodiment of the present invention, obtaining the logging parameter weights of each sample logging parameter includes: for a single sample logging parameter, calculating the mineral composition content accuracy coefficient corresponding to the sample logging parameter based on the mineral composition content prediction sub-model and the core measured mineral composition content dataset; for a single sample logging parameter, normalizing the mineral composition content accuracy coefficient corresponding to the sample logging parameter to obtain the normalized mineral composition content accuracy coefficient corresponding to the sample logging parameter; constructing the normalized mineral composition content accuracy coefficient matrix corresponding to each sample logging parameter based on the normalized mineral composition content accuracy coefficient matrix corresponding to each sample logging parameter; for a single sample logging parameter, calculating the mineral composition content entropy value corresponding to the sample logging parameter based on the normalized mineral composition content accuracy coefficient matrix corresponding to the sample logging parameter; and calculating the logging parameter weights corresponding to each sample logging parameter based on the mineral composition content entropy values ​​corresponding to each sample logging parameter.

[0118] The mineral composition content accuracy coefficient is the reciprocal of the error in the mineral composition content calculated by the candidate mineral composition content prediction sub-model corresponding to the sample logging parameters. The mineral composition content accuracy coefficient characterizes the closeness between the predicted and measured values ​​of mineral composition content. Here, the measured mineral composition content from the core is the actual measured value. It can be understood that the closer the predicted value of the mineral composition content is to the measured value, the larger the value of the mineral composition content accuracy coefficient. The normalized mineral composition content accuracy coefficient is the result of normalizing the mineral composition content accuracy coefficient. The normalized mineral composition content accuracy coefficient matrix is ​​the summary result of the normalized mineral composition content accuracy coefficients corresponding to each sample logging parameter. The mineral composition content entropy value is used to measure the uniformity (or dispersion) of the normalized mineral composition content accuracy coefficients corresponding to the sample logging parameters among all normalized mineral composition content accuracy coefficients. It can be understood that the mineral composition content entropy value is used to measure the stability of the inversion accuracy of the sample logging parameters. Correspondingly, the higher the mineral composition content entropy value, the more stable the inversion accuracy of the corresponding sample logging parameters.

[0119] Specifically, for a single sample logging parameter, the accuracy coefficient calculation formula is used to calculate the mineral composition content accuracy coefficient corresponding to the sample logging parameter based on the mineral composition content prediction sub-model of the sample logging parameter and the core measured mineral composition content dataset.

[0120] For example, the following formula can be used to represent the formula for calculating the precision coefficient: ; In the formula, The accuracy coefficient of mineral composition content corresponding to the i-th logging parameter value of the j-th sample logging parameter; The mineral composition content is predicted based on the value of the i-th logging parameter corresponding to the j-th sample logging parameter. Let represent the measured mineral composition content of the i-th core in the dataset of measured mineral composition content.

[0121] Specifically, for a single sample logging parameter, a normalization formula is used to normalize the accuracy coefficient of the mineral composition content corresponding to the sample logging parameter, thus obtaining the normalized accuracy coefficient of the mineral composition content corresponding to the sample logging parameter.

[0122] For example, the normalization formula can be represented by the following formula: ; In the formula, The normalized mineral composition content accuracy coefficient is the value of the ith logging parameter corresponding to the ith logging parameter of the j-th sample. The accuracy coefficient of mineral composition content corresponding to the i-th logging parameter value of the j-th sample logging parameter; This represents the minimum value of the accuracy coefficient for the mineral composition content corresponding to the logging parameters of the j-th sample. This represents the minimum value of the accuracy coefficient for the mineral composition content corresponding to the logging parameters of the j-th sample.

[0123] Specifically, the accuracy coefficient matrix formula is used to construct the normalized mineral composition content accuracy coefficient matrix corresponding to the sample logging parameters.

[0124] For example, the precision coefficient matrix formula can be represented by the following formula: ; In the formula, is the normalized mineral composition content accuracy coefficient corresponding to the i-th logging parameter value of the j-th sample logging parameter; m is the total number of sample logging parameter types corresponding to the input data of the candidate mineral composition content prediction model.

[0125] Specifically, for a single sample logging parameter, the entropy value calculation formula is used to calculate the entropy value of the mineral composition content corresponding to the sample logging parameter based on the normalized mineral composition content accuracy coefficient matrix.

[0126] For example, the following formula can be used to represent the formula for calculating the accuracy coefficient of mineral composition content: ; In the formula, Let be the entropy value of mineral composition content corresponding to the logging parameters of the j-th sample; is the normalized mineral composition content accuracy coefficient corresponding to the i-th logging parameter value of the j-th sample logging parameter; n is the total amount of data in the sample logging parameter dataset of the j-th sample logging parameter.

[0127] Specifically, the weighting formula is used to calculate the weight of each well logging parameter based on the entropy value of the mineral composition content corresponding to each sample well logging parameter.

[0128] For example, the following formula can be used to represent the weight calculation formula: ; In the formula, Let be the logging parameter weights corresponding to the j-th type of sample logging parameters; denoted as the entropy value of mineral composition content corresponding to the j-th sample logging parameter; m represents the total number of sample logging parameter types.

[0129] This scheme achieves adaptive optimization of logging parameter weights by quantifying the actual contribution (accuracy) and stability (entropy) of each logging parameter under specific geological conditions, thereby significantly improving the accuracy and noise resistance of mineral composition content inversion.

[0130] S207. Obtain the target logging parameter values ​​of at least two target logging parameters for the target saline formation in the target area.

[0131] S208. Based on the logging parameters of each target, select the target mineral composition content prediction model from among the candidate mineral composition content prediction models.

[0132] S209. Using a target mineral composition content prediction model, based on the logging parameter values ​​of each target, the target mineral composition content of the target saline strata in the target area is predicted.

[0133] The technical solution of this invention constructs a highly robust candidate mineral composition content prediction model by weighted fusion of multiple logging parameters through noise reduction, resolution enhancement, and probability density function. This model can effectively overcome the influence of logging data quality fluctuations under complex geological conditions where lithology and mineral composition content change frequently, achieving the optimal mineral composition content prediction in a probabilistic sense and improving the prediction accuracy of mineral composition content.

[0134] Example 3 Figure 3 This invention provides a method for predicting the salinity of saline formations, as described in Embodiment 3. Based on the above embodiments, Figure 3 This is a preferred embodiment of a method for predicting the mineral composition of saline strata provided by the present invention. See also Figure 3 The methods for predicting the salinity of saline formations shown include: S301. Selection of well logging data.

[0135] To improve the accuracy of salinity prediction calculations, priority is given to using original well logs that are sensitive to salinity response, and all original well logs are taken from sections without significant wellbore enlargement. The combination of original well logs is selected based on the following principles: 1) When logging data is complete, use a combination of natural gamma logging curves, compensated neutron logging curves, density logging curves and sonic transit time logging curves that are sensitive to salinity.

[0136] 2) When logging data is insufficient, at least the above two logging curves should be included to ensure the accuracy of salinity calculation.

[0137] S302, Well logging curve preprocessing.

[0138] Specifically, noise reduction and resolution improvement are performed on well logging data.

[0139] For example, wavelet transform expressions are used to perform wavelet transform on the original well logging curves of the sample.

[0140] For example, the wavelet transform expression can be represented by the following formula: ; In the formula, For the sample noise reduction logging curve; The original well logging curve for the sample; It is the mother wavelet The complex conjugate; This indicates the signal sampling interval of the original well logging curve of the sample; As a scaling variable, it is used to control the original logging curves of the sample. Scaling (frequency characteristics); This is a translation variable used to control the original logging curve of the sample. Translation (time domain characteristics).

[0141] Specifically, the sym5 wavelet in the Symlet wavelet family can be selected for wavelet decomposition of the sample denoising logging curve. The number of decomposition layers is selected as 3, and the sample denoising low-frequency signal, sample denoising two-layer decomposition high-frequency signal, and sample denoising three-layer decomposition high-frequency signal can be extracted from the sample denoising logging curve.

[0142] Specifically, wavelet reconstruction formulas can be used to reconstruct the wavelet decomposition results of the sample denoised logging curves to obtain the sample processed logging curves.

[0143] Specifically, under the condition that the salt rock is stably developed and the salinity does not change significantly, wavelet reconstruction is performed using the first wavelet reconstruction formula: ; In the formula, Process the well logging curves for the sample; The low-frequency signal of the sample noise-reduced logging curve; The high-frequency signal of the second-layer decomposition of the sample noise-reduced logging curve; The high-frequency signal of the three-layer decomposition of the sample noise-reduced logging curve.

[0144] This can eliminate noise signals and improve the resolution of logging curves at lithological change interfaces.

[0145] Specifically, when salt-bearing strata have well-developed interlayers or when the salinity of salt rocks varies significantly, wavelet reconstruction is performed using the second wavelet reconstruction formula: ; In the formula, Process the well logging curves for the sample; The low-frequency signal of the sample noise-reduced logging curve; The high-frequency signal of the second-layer decomposition of the sample noise-reduced logging curve; The high-frequency signal of the three-layer decomposition of the sample noise-reduced logging curve; The weights are decomposed into three levels. .

[0146] This can enhance the resolution of thin-layered salt rocks.

[0147] S303, Core repositioning.

[0148] Specifically, the core recording depth for the measured salt content of the core is adjusted to obtain the adjusted core recording depth. The degree of match between the sample logging parameter values ​​corresponding to the adjusted core recording depth and the measured salt content of the core is then checked. Based on the adjusted core recording depth corresponding to the maximum match, the core recording depth for the corresponding measured salt content is updated.

[0149] By repositioning the core samples, the core recording depth of the samples to be tested is corrected to the logging recording depth to ensure that the core measurement data and the logging data have a unified depth benchmark. Specifically, high salinity corresponds to low natural gamma logging values, low compensated neutron logging values, low density logging values, and high sonic transit time logging values.

[0150] S304, Dataset Construction and Parameter Calculation.

[0151] Specifically, for a single sample logging parameter, the sample processing logging curve is sampled according to the core recording depth of the measured salt content dataset, and the sample logging parameter values ​​corresponding to each core recording depth are extracted to obtain the sample logging parameter dataset.

[0152] For example, the following formula can be used to represent the core measured salinity dataset and the well logging parameter dataset for each sample: ; In the formula, This represents a dataset of measured salt content from rock cores, where... This represents the measured salt content of the i-th core. , , , These represent the natural gamma-ray logging dataset, compensated neutron logging dataset, density logging dataset, and sonic transit time logging data matched with core data, respectively. , , , These represent the i-th natural gamma logging value, compensated neutron logging value, density logging value, and sonic transit time logging value, respectively. This indicates the quantity of salinity, natural gamma logging values, compensated neutron logging values, density logging values, or sonic transit time logging values.

[0153] Specifically, the following formula can be used to calculate the average well logging parameters for each sample in the dataset: ; In the formula, , , , , These represent the average values ​​of salinity, natural gamma logging, compensated neutron logging, density logging, and sonic transit time logging, respectively. , , , , These represent the i-th salinity, natural gamma logging value, compensated neutron logging value, density logging value, and sonic transit time logging value, respectively. This indicates the quantity of salinity, natural gamma logging values, compensated neutron logging values, density logging values, or sonic transit time logging values.

[0154] Specifically, the following formula can be used to calculate the sum of squared deviations of each parameter in the dataset: ; In the formula, , , , , These represent the sum of squared deviations of salinity, natural gamma logging value, compensated neutron logging value, density logging value, and sonic transit time logging value, respectively. , , , , These represent the i-th salinity, natural gamma logging value, compensated neutron logging value, density logging value, and sonic transit time logging value, respectively. , , , , These represent the average values ​​of salinity, natural gamma logging, compensated neutron logging, density logging, and sonic transit time logging, respectively. This indicates the quantity of salinity, natural gamma logging values, compensated neutron logging values, density logging values, or sonic transit time logging values.

[0155] Specifically, the standard deviation of each parameter in the dataset can be calculated using the following formula: ; In the formula, , , , , These represent the standard deviations of salinity, natural gamma ray logging value, compensated neutron logging value, density logging value, and sonic transit time logging value, respectively. , , , , These represent the i-th salinity, natural gamma logging value, compensated neutron logging value, density logging value, and sonic transit time logging value, respectively. , , , , These represent the average values ​​of salinity, natural gamma logging, compensated neutron logging, density logging, and sonic transit time logging, respectively. This indicates the quantity of salinity, natural gamma logging values, compensated neutron logging values, density logging values, or sonic transit time logging values.

[0156] Specifically, the correlation coefficient between the salinity of the dataset and each logging parameter can be calculated using the following formula: ; In the formula, , , , These represent the correlation coefficients between salinity and natural gamma logging values, salinity and compensated neutron logging values, salinity and density logging values, and salinity and sonic transit time logging values, respectively. , , , , These represent the i-th salinity, natural gamma logging value, compensated neutron logging value, density logging value, and sonic transit time logging value, respectively. , , , , These represent the average values ​​of salinity, natural gamma logging, compensated neutron logging, density logging, and sonic transit time logging, respectively. , , , , These represent the sum of squared deviations of salinity, natural gamma logging value, compensated neutron logging value, density logging value, and sonic transit time logging value, respectively. This indicates the quantity of salinity, natural gamma logging values, compensated neutron logging values, density logging values, or sonic transit time logging values.

[0157] Specifically, the following formula can be used to calculate the bivariate normal probability density function between salinity and various logging parameters: ; In the formula, , , , These represent the binary normal probability density functions relating salinity to natural gamma logging parameters, salinity to compensated neutron logging parameters, salinity to density logging parameters, and salinity to sonic transit time logging parameters, respectively. , , , , These represent the standard deviations of salinity, natural gamma ray logging value, compensated neutron logging value, density logging value, and sonic transit time logging value, respectively. , , , These represent the correlation coefficients between salinity and natural gamma logging values, salinity and compensated neutron logging values, salinity and density logging values, and salinity and sonic transit time logging values, respectively. , , , , These represent the salinity, natural gamma logging value, compensated neutron logging value, density logging value, and sonic transit time logging value, respectively. , , , , These represent the average values ​​of salinity, natural gamma logging, compensated neutron logging, density logging, and sonic transit time logging, respectively. This indicates the quantity of salinity, natural gamma logging values, compensated neutron logging values, density logging values, or sonic transit time logging values.

[0158] S305, a single logging parameter salt content prediction sub-model.

[0159] Specifically, the bivariate normal probability density function between salinity and natural gamma logging parameters is obtained. The bivariate normal probability density function between salinity and compensated neutron logging parameters The bivariate normal probability density function between salinity and density logging parameters The bivariate normal probability density function between salinity and sonic transit time logging parameters Based on this, given specific natural gamma logging values Or specific compensated neutron logging values or specific density logging values or specific acoustic time-of-flight logging values The salinity under the highest probability condition can be calculated separately, thereby obtaining the salinity prediction sub-model (i.e., candidate salinity prediction sub-model) corresponding to the logging parameters of each sample.

[0160] For example, the following formula can be used to represent the salinity prediction sub-model corresponding to each logging parameter: ; In the formula, , , , These represent the salinity calculated based on natural gamma logging, compensated neutron logging, density logging, and sonic transit time logging, respectively. , , , These represent the binary normal probability density functions relating salinity to natural gamma logging parameters, salinity to compensated neutron logging parameters, salinity to density logging parameters, and salinity to sonic transit time logging parameters, respectively.

[0161] S306, Construction of a weighted fusion model for predicting salinity using multiple logging parameters.

[0162] Specifically, the reciprocal of the salinity error calculated based on the salinity prediction sub-model corresponding to a single logging parameter is defined as the accuracy coefficient.

[0163] Specifically, the following formula can be used to calculate the salinity accuracy coefficient corresponding to the logging parameters: ; in, , , , These represent the salt content accuracy coefficients when calculating salt content based on natural gamma logging values, compensated neutron logging values, density logging values, and sonic transit time logging values, respectively. In other words, they represent the salt content accuracy coefficients corresponding to the natural gamma logging parameters, compensated neutron logging parameters, density logging parameters, and sonic transit time logging parameters. , , , These represent the salinity calculated based on natural gamma logging, compensated neutron logging, density logging, and sonic transit time logging, respectively. This represents the measured value of the salt content.

[0164] Specifically, the following formula can be used to normalize the accuracy coefficient of salinity calculated from a single logging parameter: ; in, , , , These represent the normalized values ​​of the salt content accuracy coefficients when calculating salt content based on natural gamma logging, compensated neutron logging, density logging, and sonic transit time logging, respectively. , , , These represent the salt content accuracy coefficients when calculating salt content based on natural gamma logging values, compensated neutron logging values, density logging values, and sonic transit time logging values, respectively. In other words, they represent the salt content accuracy coefficients corresponding to the natural gamma logging parameters, compensated neutron logging parameters, density logging parameters, and sonic transit time logging parameters. , , , These represent the maximum values ​​of the salinity accuracy coefficients when calculating salinity based on natural gamma logging values, compensated neutron logging values, density logging values, and sonic transit time logging values, respectively. , , , These represent the minimum values ​​of the salinity accuracy coefficients when calculating salinity based on natural gamma logging, compensated neutron logging, density logging, and sonic transit time logging, respectively.

[0165] Specifically, the following formula can be used to construct a normalized salinity accuracy coefficient matrix based on the normalized salinity accuracy coefficients calculated from various logging parameters: ; In the formula, , , , These represent the normalized values ​​of the salinity accuracy coefficients when calculating salinity based on natural gamma logging, compensated neutron logging, density logging, and sonic transit time logging, respectively.

[0166] Specifically, the logging parameter weights of a single logging parameter can be obtained through the entropy weight method.

[0167] For example, the following formula can be used to calculate the salinity entropy value corresponding to each sample logging parameter: ; In the formula, , , , These represent the entropy values ​​of salinity obtained based on natural gamma logging, salinity obtained based on compensated neutron logging, salinity obtained based on density logging, and salinity obtained based on sonic transit time logging, respectively. , , , These represent the normalized values ​​of the salinity accuracy coefficients when calculating salinity based on natural gamma logging, compensated neutron logging, density logging, and sonic transit time logging, respectively.

[0168] For example, the following formula can then be used to calculate the logging parameter weights corresponding to each sample logging parameter: ; In the formula, , , , These represent the weights of the salinity obtained based on natural gamma logging, the salinity obtained based on compensated neutron logging, the salinity obtained based on density logging, and the salinity obtained based on sonic transit time logging, respectively. , , , These represent the entropy values ​​of salinity obtained based on natural gamma logging, salinity obtained based on compensated neutron logging, salinity obtained based on density logging, and salinity obtained based on sonic transit time logging, respectively.

[0169] Specifically, the following formula can be used to construct a multi-logging parameter weighted fusion salinity calculation model (i.e., a candidate salinity prediction model) based on the logging parameter weights corresponding to each sample logging parameter: ; in, The salinity is calculated based on a weighted fusion algorithm of multiple logging parameters. , , , These represent the weights of the salinity obtained based on natural gamma logging, the salinity obtained based on compensated neutron logging, the salinity obtained based on density logging, and the salinity obtained based on sonic transit time logging, respectively. , , , These represent the salinity calculated based on natural gamma logging, compensated neutron logging, density logging, and sonic transit time logging, respectively.

[0170] When complete well logging data is available, i.e., natural gamma ray logging, compensated neutron logging, density logging, and sonic transit time logging are selected for salinity calculation, the above formula is used for weight calculation and construction of a weighted fusion salinity calculation model for multiple logging parameters. When well logging data is insufficient, i.e., the selected well logging curves are a subset of the above four types of well logging curves, the set of logging parameters corresponding to the subset is used to generate corresponding well logging parameter weights, and the weighted fusion salinity calculation model for multiple logging parameters is adjusted accordingly to adapt to different data completeness conditions.

[0171] like Figure 4 As shown, , , , These represent the natural gamma logging curve, compensated neutron logging curve, density logging curve, and sonic transit time logging curve after wavelet transform processing, respectively. This indicates the measured salt content of the core after it has been returned to its original position. , , , These represent the salinity predicted by statistical regression empirical formulas based on natural gamma logging values, compensated neutron logging values, density logging values, and sonic transit time logging values, respectively. This represents the salinity predicted by a multi-logging parameter weighted fusion salinity calculation model constructed using natural gamma logging values ​​and sonic transit time logging values ​​as input data. This represents the salinity predicted by a multi-logging parameter weighted fusion salinity calculation model constructed using natural gamma logging values, compensated neutron logging values, density logging values, and sonic transit time logging values ​​as input data.

[0172] like Figure 4 As shown in the figure, natural gamma ray logging values, compensated neutron logging values, density logging values, and sonic transit time logging values ​​were used as input values, and salinity was predicted using a statistical regression empirical formula method. The prediction results are shown in the figure. , , , The errors were 11.03%, 13.65%, 15.11%, and 12.66%, respectively.

[0173] like Figure 4 As shown in the figure, a weighted fusion model for calculating salinity using natural gamma logging values ​​and sonic transit time logging values ​​processed by wavelet transform as input data is constructed. The calculation results are shown in the figure. The error was 7.24%, and the prediction effect was significantly better than that of the statistical regression empirical formula method.

[0174] like Figure 4 As shown in the figure, a multi-logging parameter weighted fusion salinity calculation model is constructed using natural gamma logging values, compensated neutron logging values, density logging values, and sonic transit time logging values ​​processed by wavelet transform as input data. The calculation results are shown in the figure. The error was 6.69%, which is a certain improvement compared to the prediction effect of the multi-logging parameter weighted fusion salinity calculation model constructed by using natural gamma logging values ​​and sonic transit time logging values ​​after wavelet transform processing as input data.

[0175] Compared with existing methods, this invention integrates wavelet transform, binary normal distribution probability density function, and entropy weighting method to construct a highly robust multi-logging parameter weighted fusion salinity prediction model. This model effectively overcomes the influence of fluctuating logging data quality under complex geological conditions with frequent changes in lithology and salinity, achieving probabilistically optimal salinity prediction. Furthermore, this invention can dynamically adjust the multi-logging parameter weighted fusion salinity calculation model based on differences in logging data from different study areas, ensuring accurate salinity prediction under different regional, stratigraphic, and logging data source conditions. Finally, this invention is not only applicable to the prediction of salinity in saline formations but also has strong adaptability to the content of other mineral components (such as insoluble matter, glaucophane, or gypsum), and its application breadth is significantly superior to existing technologies.

[0176] Example 4 Figure 5 This is a schematic diagram of a device for predicting the mineral content of saline strata according to Embodiment 4 of the present invention. This embodiment of the invention is applicable to situations where the mineral content of complex saline strata needs to be predicted. The device can execute a method for predicting the mineral content of saline strata. The device can be implemented in hardware and / or software, and can be configured in an electronic device that performs the function of predicting the mineral content of saline strata.

[0177] See Figure 5The saline formation mineral content prediction device shown includes: a target logging parameter value acquisition module 501, a mineral content prediction model screening module 502, and a target mineral content prediction module 503. Specifically, the target logging parameter value acquisition module 501 is used to acquire target logging parameter values ​​for at least two target logging parameters of the target saline formation in the target area; the mineral content prediction model screening module 502 is used to screen target mineral content prediction models from among candidate mineral content prediction models based on each target logging parameter; and the target mineral content prediction module 503 is used to predict the target mineral content of the target saline formation in the target area based on each target logging parameter value using the target mineral content prediction model.

[0178] The technical solution of this invention obtains the target logging parameter values ​​of at least two target logging parameters of the target saline formation in the target area. Based on each target logging parameter, a target mineral composition content prediction model is selected from each candidate mineral composition content prediction model. Using the target mineral composition content prediction model, the target mineral composition content of the target saline formation in the target area is predicted based on each target logging parameter value. This takes into account the differences between the logging parameters that can be obtained in different areas. By using different candidate mineral composition content prediction models, the complex nonlinear coupling relationship between multiple parameters in the actual formation is characterized, thus improving the prediction accuracy of mineral composition content under complex formation conditions.

[0179] In an optional embodiment of the present invention, the apparatus includes: a sample raw logging curve acquisition module, used to acquire sample raw logging curves and a core measured mineral composition content dataset of at least two sample logging parameters sensitive to mineral composition content in a sample area; a sample raw logging curve processing module, used to reduce noise and improve resolution of the sample raw logging curves to obtain sample processed logging curves; a sample logging parameter dataset generation module, used to sample the sample processed logging curves for a single sample logging parameter based on the core recording depth of the core measured mineral composition content dataset to obtain a sample logging parameter dataset; and a probability density function construction module, used to construct a probability density function for a single sample logging parameter. Based on the core measured mineral composition content dataset and the sample logging parameter dataset, this well logging parameter constructs a probability density function between mineral composition content and individual sample logging parameters. A mineral composition content prediction sub-model generation module generates a mineral composition content prediction sub-model corresponding to each sample logging parameter based on the probability density function between mineral composition content and individual sample logging parameters, as well as the sample logging parameter dataset. A candidate mineral composition content prediction model generation module obtains the logging parameter weights for each sample logging parameter and uses these weights to fuse the mineral composition content prediction sub-models corresponding to each sample logging parameter, resulting in a candidate mineral composition content prediction model.

[0180] In an optional embodiment of the present invention, the device further includes: a wellbore enlargement zone detection module, used to detect whether the original well log curve of the sample belongs to the wellbore enlargement zone after acquiring the original well log curve of at least two sample logging parameters that are sensitive to the mineral composition content in the sample area and the core measured mineral composition content dataset; and a sample original well log curve rejection module, used to reject the original well log curve of the sample when it belongs to the wellbore enlargement zone.

[0181] In an optional embodiment of the present invention, the sample original logging curve processing module includes: a sample original logging curve denoising unit, used to perform wavelet transform on the sample original logging curve to obtain a sample denoised logging curve; and a sample denoised logging curve resolution enhancement unit, used to perform wavelet reconstruction on the sample denoised logging curve to obtain a sample processed logging curve.

[0182] In an optional embodiment of the present invention, the sample noise reduction logging curve resolution enhancement unit includes: a sample noise reduction logging curve decomposition subunit, used to extract sample noise reduction low-frequency signals, sample noise reduction second-layer decomposition high-frequency signals, and sample noise reduction third-layer decomposition high-frequency signals from the sample noise reduction logging curve; a saline formation state detection subunit, used to detect the development state and mineral composition content change state of the saline formation in the sample area; and a first sample processing logging curve generation subunit, used to process the sample noise reduction logging curve when the saline formation in the sample area is stably developed and the mineral composition content change rate is less than or equal to a preset mineral composition content change threshold. The low-frequency signal, the high-frequency signal from the sample denoising two-layer decomposition, and the high-frequency signal from the sample denoising three-layer decomposition are superimposed to obtain the first sample-processed logging curve. The second sample-processed logging curve generation sub-unit is used to obtain the three-layer decomposition weight of the sample denoising three-layer decomposition high-frequency signal when the saline strata interlayers are developed in the sample area and the rate of change of mineral composition content is greater than the preset mineral composition content change threshold. The three-layer decomposition weight is then used to superimpose the sample denoising low-frequency signal, the sample denoising two-layer decomposition high-frequency signal, and the sample denoising three-layer decomposition high-frequency signal to obtain the second sample-processed logging curve. The three-layer decomposition weight is greater than 1.

[0183] In an optional embodiment of the present invention, the device further includes: a core recording depth adjustment module, used to adjust the core recording depth of the measured mineral composition content in the core to obtain an adjusted core recording depth; a matching degree detection module, used to detect the matching degree between the sample logging parameter values ​​of the same logging depth as the adjusted core recording depth and the measured mineral composition content in the core; and a core repositioning module, used to correct the core recording depth of the measured mineral composition content dataset in the core according to the matching degree.

[0184] In an optional embodiment of the present invention, the candidate mineral composition content prediction model generation module includes: a mineral composition content accuracy coefficient calculation unit, used to calculate the mineral composition content accuracy coefficient corresponding to a single sample logging parameter based on the mineral composition content prediction sub-model of the sample logging parameter and the core measured mineral composition content dataset; a mineral composition content accuracy coefficient normalization unit, used to normalize the mineral composition content accuracy coefficient corresponding to a single sample logging parameter to obtain the normalized mineral composition content accuracy coefficient corresponding to the sample logging parameter; a mineral composition content accuracy coefficient matrix generation unit, used to construct the normalized mineral composition content accuracy coefficient matrix corresponding to each sample logging parameter based on the normalized mineral composition content accuracy coefficient corresponding to each sample logging parameter; a mineral composition content entropy value calculation unit, used to calculate the mineral composition content entropy value corresponding to a single sample logging parameter based on the normalized mineral composition content accuracy coefficient matrix corresponding to the sample logging parameter; and a logging parameter weight calculation unit, used to calculate the logging parameter weight corresponding to each sample logging parameter based on the mineral composition content entropy value corresponding to each sample logging parameter.

[0185] The saline strata mineral composition prediction device provided in this embodiment of the invention can execute the saline strata mineral composition prediction method provided in any embodiment of the invention, and has the corresponding functional modules and beneficial effects of the method.

[0186] In the technical solutions of this invention, the acquisition, storage, and application of the target logging parameter values ​​of at least two target logging parameters of the target saline strata in the target area, the original logging curves of at least two sample logging parameters that are sensitive to the mineral composition content in the sample area, the core measured mineral composition content dataset, and the logging parameter weights of the sample logging parameters, etc., all comply with the provisions of relevant laws and regulations and do not violate public order and good morals.

[0187] Example 4 Figure 6 A schematic diagram of an electronic device 600 that can be used to implement embodiments of the present invention is shown. The electronic device is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device can also represent various forms of mobile devices, such as personal digital processors, cellular phones, smartphones, wearable devices (e.g., helmets, glasses, watches, etc.), and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely illustrative and are not intended to limit the implementation of the invention described and / or claimed herein.

[0188] like Figure 6As shown, the electronic device 600 includes at least one processor 601 and a memory, such as a read-only memory (ROM) 602 or a random access memory (RAM) 603, communicatively connected to the at least one processor 601. The memory stores computer programs executable by the at least one processor. The processor 601 can perform various appropriate actions and processes based on the computer program stored in the ROM 602 or loaded into the RAM 603 from storage unit 608. The RAM 603 may also store various programs and data required for the operation of the electronic device 600. The processor 601, ROM 602, and RAM 603 are interconnected via a bus 604. An input / output (I / O) interface 605 is also connected to the bus 604.

[0189] Multiple components in electronic device 600 are connected to I / O interface 605, including: input unit 606, such as keyboard, mouse, etc.; output unit 607, such as various types of displays, speakers, etc.; storage unit 608, such as disk, optical disk, etc.; and communication unit 609, such as network card, modem, wireless transceiver, etc. Communication unit 609 allows electronic device 600 to exchange information / data with other devices through computer networks such as the Internet and / or various telecommunications networks.

[0190] Processor 601 can be a variety of general-purpose and / or special-purpose processing components with processing and computing capabilities. Some examples of processor 601 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various special-purpose artificial intelligence (AI) computing chips, various processors running machine learning model algorithms, a digital signal processor (DSP), and any suitable processor, controller, microcontroller, etc. Processor 601 performs the various methods and processes described above, such as methods for predicting the mineral composition of saline strata.

[0191] In some embodiments, the method for predicting the mineral content of saline formations can be implemented as a computer program tangibly contained in a computer-readable storage medium, such as storage unit 608. In some embodiments, part or all of the computer program can be loaded and / or installed on electronic device 600 via ROM 602 and / or communication unit 609. When the computer program is loaded into RAM 603 and executed by processor 601, one or more steps of the method for predicting the mineral content of saline formations described above can be performed. Alternatively, in other embodiments, processor 601 can be configured to perform the method for predicting the mineral content of saline formations by any other suitable means (e.g., by means of firmware).

[0192] Various implementations of the systems and techniques described above herein can be implemented in digital electronic circuit systems, integrated circuit systems, field-programmable gate arrays (FPGAs), application-specific integrated circuits (ASICs), application-specific standard products (ASSPs), systems-on-a-chip (SoCs), complex programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various implementations may include: implementations in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which may be a dedicated or general-purpose programmable processor, capable of receiving data and instructions from a storage system, at least one input device, and at least one output device, and transmitting data and instructions to the storage system, the at least one input device, and the at least one output device.

[0193] Computer programs used to implement the methods of the present invention may be written in any combination of one or more programming languages. These computer programs may be provided to a processor of a general-purpose computer, a special-purpose computer, or other programmable data processing device, such that when executed by the processor, the computer programs cause the functions / operations specified in the flowcharts and / or block diagrams to be performed. The computer programs may be executed entirely on a machine, partially on a machine, or as a standalone software package, partially on a machine and partially on a remote machine, or entirely on a remote machine or server.

[0194] In the context of this invention, a computer-readable storage medium can be a tangible medium that may contain or store a computer program for use by or in conjunction with an instruction execution system, apparatus, or device. A computer-readable storage medium may include, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, apparatus, or devices, or any suitable combination thereof. Alternatively, a computer-readable storage medium may be a machine-readable signal medium. More specific examples of machine-readable storage media include electrical connections based on one or more wires, portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fibers, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination thereof.

[0195] To provide interaction with a user, the systems and techniques described herein can be implemented on an electronic device having: a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to the user; and a keyboard and pointing device (e.g., a mouse or trackball) through which the user provides input to the electronic device. Other types of devices can also be used to provide interaction with the user; for example, feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form (including sound input, voice input, or tactile input).

[0196] The systems and technologies described herein can be implemented in computing systems that include backend components (e.g., as data servers), or computing systems that include middleware components (e.g., application servers), or computing systems that include frontend components (e.g., user computers with graphical user interfaces or web browsers through which users can interact with implementations of the systems and technologies described herein), or any combination of such backend, middleware, or frontend components. The components of the system can be interconnected via digital data communication of any form or medium (e.g., communication networks). Examples of communication networks include local area networks (LANs), wide area networks (WANs), blockchain networks, and the Internet.

[0197] A computing system can include clients and servers. Clients and servers are generally geographically separated and typically interact via communication networks. The client-server relationship is created by computer programs running on the respective computers and having a client-server relationship with each other. The server can be a cloud server, also known as a cloud computing server or cloud host, which is a hosting product within the cloud computing service system. It addresses the shortcomings of traditional physical hosts and VPS (Virtual Private Server) services, such as high management difficulty and weak business scalability.

[0198] It should be understood that the various forms of processes shown above can be used to reorder, add, or delete steps. For example, the steps described in this invention can be executed in parallel, sequentially, or in different orders, as long as the desired result of the technical solution of this invention can be achieved, and this is not limited herein.

[0199] The specific embodiments described above do not constitute a limitation on the scope of protection of this invention. Those skilled in the art should understand that various modifications, combinations, sub-combinations, and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of this invention should be included within the scope of protection of this invention.

Claims

1. A method for predicting the mineral composition content of saline strata, characterized in that, The method includes: Obtain target logging parameter values ​​for at least two target logging parameters of the target saline formation in the target area; Based on the target logging parameters, the target mineral composition prediction model is selected from the candidate mineral composition prediction models. Using the target mineral composition content prediction model, the target mineral composition content of the target saline strata in the target area is predicted based on the target logging parameter values.

2. The method according to claim 1, characterized in that, Before obtaining the target logging parameter values ​​for at least two target logging parameters in the sample area, the method further includes: Obtain the original logging curves of at least two sample logging parameters that are sensitive to mineral composition content in the sample area, as well as the core measured mineral composition content dataset; The original logging curves of the sample are denoised and the resolution is improved to obtain the processed logging curves of the sample. For a single sample logging parameter, the sample processing logging curve is sampled based on the core recording depth of the core measured mineral composition content dataset to obtain the sample logging parameter dataset; For a single sample logging parameter, a probability density function is constructed between the mineral composition content and the single sample logging parameter, based on the core measured mineral composition content dataset and the sample logging parameter dataset. Based on the probability density function between the mineral composition content and a single sample logging parameter, and the sample logging parameter dataset, a mineral composition content prediction sub-model corresponding to a single sample logging parameter is generated. Obtain the logging parameter weights of each sample logging parameter, and use the logging parameter weights to fuse the mineral composition content prediction sub-models corresponding to each sample logging parameter to obtain the candidate mineral composition content prediction model.

3. The method according to claim 2, characterized in that, After obtaining the original well log curves of at least two sample logging parameters that are sensitive to mineral composition content in the sample area and the core measured mineral composition content dataset, the method further includes: The test is conducted to determine whether the original logging curve of the sample belongs to the wellbore enlargement section. If the original logging curve of the sample belongs to the wellbore enlargement section, the original logging curve of the sample will be discarded.

4. The method according to claim 2, characterized in that, The process of denoising and improving the resolution of the original logging curves of the sample to obtain the processed logging curves includes: Wavelet transform is performed on the original logging curves of the sample to obtain the denoised logging curves of the sample; Wavelet reconstruction is performed on the sample denoised logging curves to obtain the sample processed logging curves.

5. The method according to claim 4, characterized in that, The step of performing wavelet reconstruction on the sample denoised logging curves to obtain the sample processed logging curves includes: Extract the sample denoising low-frequency signal, the sample denoising two-layer decomposition high-frequency signal, and the sample denoising three-layer decomposition high-frequency signal from the sample denoising logging curve; The development status and mineral composition variation status of the saline strata in the sample area were detected. When the saline strata in the sample area are stably developed and the rate of change of mineral composition content is less than or equal to the preset threshold for change of mineral composition content, the sample noise-reduced low-frequency signal, the sample noise-reduced second-layer decomposition high-frequency signal, and the sample noise-reduced third-layer decomposition high-frequency signal are superimposed to obtain the first sample processing logging curve. When the saline strata in the sample area have interlayers and the rate of change of mineral composition content is greater than the preset threshold for change of mineral composition content, the three-layer decomposition weight of the sample denoised three-layer decomposition high-frequency signal is obtained, and the sample denoised low-frequency signal, the sample denoised two-layer decomposition high-frequency signal, and the sample denoised three-layer decomposition high-frequency signal are superimposed using the three-layer decomposition weight to obtain the second sample processed logging curve; wherein, the three-layer decomposition weight is greater than 1.

6. The method according to claim 2, characterized in that, Before sampling the sample-processed logging curve based on the core recording depth of the core measured mineral composition dataset for a single sample logging parameter, the method further includes: The core recording depth is adjusted to reflect the measured mineral composition of the core, resulting in the adjusted core recording depth. The degree of matching between the sample logging parameter values ​​at the same logging depth as the adjusted core recording depth and the measured mineral composition content of the core is detected. Based on the degree of matching, the core recording depth of the core measured mineral composition content dataset is corrected.

7. The method according to claim 2, characterized in that, The step of obtaining the logging parameter weights for each of the sample logging parameters includes: For a single sample logging parameter, the mineral composition content prediction sub-model of the sample logging parameter and the core measured mineral composition content dataset are used to calculate the mineral composition content accuracy coefficient corresponding to the sample logging parameter; For a single sample logging parameter, the accuracy coefficient of mineral composition content corresponding to the sample logging parameter is normalized to obtain the normalized accuracy coefficient of mineral composition content corresponding to the sample logging parameter. Based on the normalized mineral composition content accuracy coefficients corresponding to the logging parameters of each sample, a matrix of normalized mineral composition content accuracy coefficients corresponding to the logging parameters of the sample is constructed. For a single sample logging parameter, the entropy value of the mineral composition content corresponding to the sample logging parameter is calculated based on the normalized mineral composition content accuracy coefficient matrix corresponding to the sample logging parameter. Based on the entropy value of mineral composition content corresponding to each sample logging parameter, the logging parameter weights corresponding to each sample logging parameter are calculated respectively.

8. A device for predicting the mineral composition of saline strata, characterized in that, The device includes: The target logging parameter value acquisition module is used to acquire the target logging parameter values ​​of at least two target logging parameters of the target saline formation in the target area. The mineral composition content prediction model screening module is used to screen the target mineral composition content prediction model from among the candidate mineral composition content prediction models based on the target logging parameters. The target mineral composition content prediction module is used to predict the target mineral composition content of the target saline strata in the target area based on the target logging parameter values ​​using the target mineral composition content prediction model.

9. An electronic device, characterized in that, The electronic device includes: At least one processor; and A memory communicatively connected to the at least one processor; wherein, The memory stores a computer program that can be executed by the at least one processor, the computer program being executed by the at least one processor to enable the at least one processor to perform the method for predicting the mineral composition content of saline strata according to any one of claims 1-7.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer instructions that cause a processor to execute the method for predicting the mineral composition content of saline strata according to any one of claims 1-7.

Images