Multi-well consistency correction method, system and equipment based on Warisstein distance

Abstract

The invention discloses a multi-well consistency correction method, system and equipment based on Warisstein distance, and relates to the field of oil and gas field exploration and development, and the method comprises the steps: constructing an equal-width histogram of a standard well based on a logging curve of the standard well; according to the equal-width histogram of the standard well and the logging curves of the multiple to-be-corrected wells, on the basis of an empirical distribution function method, with the minimum Warisstein distance between each to-be-corrected well and the standard well as the target, an annealing algorithm is adopted to conduct iterative optimization on parameters of the consistency mapping function, and optimal parameters corresponding to each to-be-corrected well are obtained; inputting a consistency mapping function to obtain an optimal matching correction model corresponding to each well to be corrected; and inputting the logging curve of each to-be-corrected well into the corresponding optimal matching correction model to obtain a corrected logging curve of each to-be-corrected well, and completing multi-well consistency correction. According to the method, the accuracy of the corrected logging curve can be improved, so that a solid data foundation is laid for subsequent seismic inversion and reservoir prediction.

Description

Technical Field

[0001] This application relates to the field of oil and gas field exploration and development technology, and in particular to a multi-well consistency correction method, system and equipment based on Wasserstein distance. Background Technology

[0002] In oil and gas exploration and development, physical property logging is an indispensable key technology for quantitatively evaluating the lithology, physical properties, and fluid properties of subsurface formations. However, in actual production, multi-well logging data commonly suffers from systematic baseline inconsistencies. This inconsistency mainly stems from differences in logging instrument models, calibration standards, wellbore environments, and drilling fluid properties used in different well runs. These factors lead to significant background shifts or amplitude scaling differences in curve values ​​measured in different wells, even for the same geological formation. This "multi-well inconsistency" severely restricts the application of logging data in cross-well correlation, seismic attribute calibration, and quantitative interpretation models, reducing the accuracy and reliability of reservoir prediction results.

[0003] In some cases, the commonly used multi-well consistency correction methods in the industry, such as the mean-variance method and histogram matching method based on statistical characteristics, have some limitations, such as: (1) Insufficient accuracy in matching distribution patterns: Most methods focus on aligning the statistical moments (such as mean and variance) of the corrected data with the standard well, but do not adequately consider the matching of the overall probability distribution pattern of the data. For example, simple histogram translation or scaling is difficult to finely adjust complex patterns such as non-Gaussian distribution and multimodal distribution, resulting in an essential difference in statistical distribution structure between the corrected curve and the standard curve, which cannot meet the stringent requirements of high-precision reservoir characterization for data distribution consistency. (2) The correction parameters (such as translation amount and scaling factor) of most technologies rely on manual experience or are determined by simple linear regression, lacking an objective and adaptive optimization framework. This mode, which is mainly based on "manual intervention", is not only inefficient in correction, but also difficult to guarantee the optimality and repeatability of the results due to its strong subjectivity. Overall, the existing technologies have not yet been able to effectively solve the problem of high-precision and adaptive matching from the perspective of data distribution patterns. Summary of the Invention

[0004] The purpose of this application is to provide a multi-well consistency correction method, system, and device based on Wasserstein distance, which can improve the accuracy of the corrected logging curves, thereby laying a solid data foundation for subsequent seismic inversion and reservoir prediction.

[0005] To achieve the above objectives, this application provides the following solution.

[0006] Firstly, this application provides a multi-well consistency correction method based on Wassstein distance, including: Obtain logging curves from standard wells and logging curves from multiple wells to be calibrated; Based on the logging curves of standard wells, construct an equal-width histogram of the standard wells; Based on the equal width histogram of the standard well and the logging curves of multiple wells to be calibrated, and using the empirical distribution function method, with the goal of minimizing the Wasserstein distance between each well to be calibrated and the standard well, the parameters of the consistency mapping function are iteratively optimized using the annealing algorithm to obtain the optimal parameters for each well to be calibrated. The optimal parameters corresponding to each well to be corrected are input into the consistency mapping function to obtain the optimal matching correction model corresponding to each well to be corrected. The logging curves of each well to be calibrated are input into the corresponding optimal matching calibration model to obtain the calibrated logging curves of each well, thus completing the multi-well consistency calibration.

[0007] Secondly, this application provides a multi-well consistency correction system based on Wassstein distance, comprising the following modules: The acquisition module is used to acquire the logging curves of a standard well and the logging curves of multiple wells to be calibrated; The module is used to construct an isowidth histogram of a standard well based on its logging curves. The optimization module is used to iteratively optimize the parameters of the consistency mapping function based on the equal width histogram of the standard well and the logging curves of multiple wells to be calibrated, using the empirical distribution function method and with the goal of minimizing the Wasserstein distance between each well to be calibrated and the standard well, and to obtain the optimal parameters for each well to be calibrated. The calibration model generation module is used to input the optimal parameters corresponding to each well to be calibrated into the consistency mapping function to obtain the optimal matching calibration model corresponding to each well to be calibrated. The calibration module is used to input the logging curves of each well to be calibrated into the corresponding optimal matching calibration model to obtain the calibrated logging curves of each well and complete the consistency calibration of multiple wells.

[0008] Thirdly, this application provides a computer device, including: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the steps of the above-described multi-well consistency correction method based on Wasserstein distance.

[0009] According to the specific embodiments provided in this application, this application has the following technical effects: In this application, when correcting the logging curves of multiple wells to be corrected based on the logging curves of a standard well, the process involves several steps. First, an equal-width histogram of the standard well is constructed using its logging curves. Then, the Wasserstein distance is introduced into the annealing algorithm, and the parameters of the consistency mapping are iteratively optimized with the goal of minimizing the Wasserstein distance calculated using the empirical distribution function method. This yields the optimal parameters for each well to be corrected. Finally, an optimal matching correction model is constructed using these optimal parameters, and the logging curves of each well are corrected using this optimal matching correction model. In this process, the Wasserstein distance comprehensively considers the distribution characteristics between the standard well and the wells to be corrected, avoiding insufficient matching caused by relying solely on statistics. Simultaneously, the annealing algorithm dynamically adjusts the parameters of the consistency mapping with the goal of minimizing the Wasserstein distance, achieving adaptive parameter adjustment, reducing manual intervention, and thus improving the accuracy of the corrected logging curves. Attached Figure Description

[0010] To more clearly illustrate the technical solutions in the embodiments of this application or related technologies, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0011] Figure 1 A flowchart illustrating a multi-well consistency correction method based on Wasserstein distance provided in this application.

[0012] Figure 2 for Figure 1 A detailed flowchart of step S3.

[0013] Figure 3 This is a schematic diagram of the uncorrected equal-width histogram of the well to be corrected and the equal-width histogram of the standard well provided in this application.

[0014] Figure 4 A schematic diagram showing the Wasserstein distance of the well to be calibrated before calibration and the Wasserstein distance of the standard well provided in this application.

[0015] Figure 5 This is a schematic diagram of the corrected equal-width histogram of the well to be corrected and the equal-width histogram of the standard well provided in this application.

[0016] Figure 6 A schematic diagram of the corrected Wasserstan distance of the well to be corrected and the Wasserstan distance of the standard well provided for this application. Figure 7 Comparison chart of density curve consistency correction for two example wells provided in this application embodiment. Figure 8This is a schematic diagram of the functional modules of a multi-well consistency correction system based on Wassstein distance, provided for another embodiment of this application.

[0017] Figure 9 This is a schematic diagram of the structure of a computer device provided in an embodiment of this application.

[0018] Attached reference numerals: Acquisition module-1, Construction module-2, Optimization module-3, Calibration model generation module-4, Calibration module-5. Detailed Implementation

[0019] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0020] To make the above-mentioned objectives, features and advantages of this application more apparent and understandable, the application will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0021] In one exemplary embodiment, such as Figure 1 As shown, a multi-well consistency correction method based on Wassstein distance is provided. This method is executed by a computer device, specifically by a terminal or server alone, or by both a terminal and a server. In this embodiment, the method is described using a server as an example, and includes the following steps S1 to S5.

[0022] Step S1: Obtain the logging curves of the standard well and the logging curves of multiple wells to be calibrated.

[0023] As one feasible approach, step S1 specifically includes steps S11 to S13: Step S11: Select any one of the multiple target wells as the standard well and the remaining wells as the wells to be calibrated.

[0024] Step S12: Obtain the initial logging curves of multiple target wells.

[0025] Step S13: Clean the initial logging data of multiple target wells to obtain the logging curves of the standard well and the logging curves of multiple wells to be calibrated.

[0026] Specifically, the logging curves of standard wells in the target work area are selected as reference benchmarks, and the wells in the target work area other than the standard wells are selected as wells to be calibrated; logging data of all wells in the target work area are obtained, and the logging data of all wells are subjected to quality control to remove outliers and invalid data points.

[0027] Step S2: Based on the logging curves of the standard well, construct the equal width histogram of the standard well.

[0028] Step S3: Based on the equal width histogram of the standard well and the logging curves of multiple wells to be calibrated, and using the empirical distribution function method, with the goal of minimizing the Wasserstein distance between each well to be calibrated and the standard well, the parameters of the consistency mapping function are iteratively optimized using the annealing algorithm to obtain the optimal parameters for each well to be calibrated.

[0029] Specifically, the annealing algorithm automatically searches the parameter space to find the optimal parameters by sequentially performing initialization, generating new parameters, evaluating the standardization effect, accepting / rejecting new parameters according to the Metropolis criterion, and cooling. After the annealing algorithm completes, the global output represents the found optimal parameters. The annealing algorithm is used to optimize and adjust the optimal parameters to minimize the Wasserstein distance. The annealing algorithm process is as follows: (1) Starting from the initial parameters, perform intelligent search within a reasonable parameter range and set the initial parameters.

[0030] (2) High temperature stage: Explore different parameter combinations on a large scale and quickly locate the high-quality parameter area.

[0031] (3) Cooling stage: Gradually narrow the search range and finely adjust the parameter values.

[0032] (4) Termination stage: After reaching the distribution matching threshold according to the set termination conditions, the optimal parameters are obtained.

[0033] As an feasible approach, such as Figure 2 As shown, for any well to be calibrated, the following steps are used to obtain the optimal parameters corresponding to each well. Step S3 specifically includes steps S31 to S37: Step S31: Determine the parameters of the consistency mapping function at the current iteration number to obtain the current parameters; if the current iteration number is the initial iteration number, determine the initial parameters of the consistency mapping function based on the equal width histogram of the standard well.

[0034] As an feasible approach, if the current iteration number is the initial iteration number, the initial parameters of the consistency mapping function are determined based on the equal-width histogram of the standard well. Specifically, this includes: determining the standard numerical range based on the equal-width histogram of the standard well; the standard numerical range is the numerical range of the logging curve of the standard well; using the standard numerical range as a constraint on the numerical range to be corrected, the initial parameters of the consistency mapping function are determined; the numerical range to be corrected is the numerical range of the logging curve of the well after the transformation.

[0035] Specifically, after constructing the equal-width histogram of the standard well, the numerical range of the logging curve of the standard well can be determined. For example, porosity (sonic, compensated neutron, and density) logging curves, spontaneous potential logging curves, and natural gamma logging curves.

[0036] Define the consistency mapping function as follows: ,in, The initial calibration logging curve for a standard well; The initial slope parameter controls data scaling. The main compensations are for differences in instrument scale and the influence of skeleton density; The logging curves of the well to be calibrated; The initial intercept parameter controls the data translation. The main correction is for baseline shifts caused by factors such as clay content and pore fluid. The result of normalizing the logging curves; constraints and Initial values, such that the numerical range of the initial calibration logging curve is within... .

[0037] Step S32: Input the current parameters into the consistency mapping function to obtain the current matching correction model.

[0038] Step S33: Input the logging curve of the well to be calibrated into the current matching calibration model to obtain the current calibration logging curve of the well to be calibrated.

[0039] Step S34: Construct the current equal width histogram of the well to be calibrated based on the current calibration logging curve.

[0040] Step S35: The empirical distribution function method is used to calculate the isowidth histogram of the standard well and the current isowidth histogram of the well to be corrected, so as to obtain the Wasserstein distance between the well to be corrected and the standard well at the current iteration number.

[0041] Specifically, the Empirical Distribution Function (EDF) method is used to calculate the isowidth histograms of the standard well and the current isowidth histogram of the well to be calibrated, obtaining the Cumulative Distribution Function (CDF) of the standard well and the current Cumulative Distribution Function of the well to be calibrated. Then, based on the Cumulative Distribution Function (CDF) of the standard well and the current Cumulative Distribution Function of the well to be calibrated, the Wasserstein distance between the well to be calibrated and the standard well is calculated at the current iteration number.

[0042] The Wasserstein distance is used as a measure of the distribution difference in well logging data, measuring the overall distribution difference between the current cumulative distribution function of the well to be corrected and the cumulative distribution function of the standard well. The Wasserstein distance is calculated using the same formula.

[0043] As an feasible approach, the Wasserstein distance can be calculated using the following formula: ; in, The Wasserstein distance between the well to be calibrated and the standard well; Let be the cumulative probability distribution function of the well to be corrected; Let be the cumulative probability distribution function of a standard well; The corrected logging curves for the well to be corrected; The corrected logging curves for a standard well; Let be the integral variable, representing the value of the logging curve (e.g., the density range from 1.5 to 3.0). In the one-dimensional case, since the cumulative distribution function of the well to be calibrated and the cumulative distribution function of the standard well are in curve form, the Wasserstein distance is the area between the two cumulative distribution function curves.

[0044] Step S36: Obtain the first parameter based on the Wasserstein distance between the well to be calibrated and the standard well in the current iteration and the Wasserstein distance between the well to be calibrated and the standard well in the previous iteration. The first parameter is either the current parameter or the parameter in the previous iteration.

[0045] As an implementable approach, step S36 specifically includes: determining whether the Wasserstein distance between the well to be calibrated and the standard well in the current iteration number is less than the Wasserstein distance between the well to be calibrated and the standard well in the previous iteration number, and obtaining a first judgment result; if the first judgment result is yes, then the current parameter is determined as the first parameter; if the first judgment result is no, then the current parameter is accepted according to the Metropolis acceptance criterion, and the first parameter is obtained.

[0046] As an implementable approach, the first parameter is obtained by deciding whether to accept the current parameter according to the Metropolis acceptance criterion. Specifically, this includes: deciding whether to accept the current parameter according to the Metropolis acceptance criterion to obtain a second judgment result; if the second judgment result is yes, then the current parameter is determined as the first parameter; if the second judgment result is no, then the parameter from the previous iteration is determined as the first parameter.

[0047] Step S37: Determine whether the Wasserstein distance between the well to be calibrated and the standard well corresponding to the first parameter reaches the distribution matching threshold; if yes, determine the first parameter as the optimal parameter; if no, execute the update process; the update process includes: determining the current annealing temperature based on the annealing temperature of the previous iteration and the current iteration; obtaining the updated parameter based on the current annealing temperature and the first parameter, and using the updated parameter as the parameter of the consistency mapping function in the next iteration, updating the iteration number, and returning to determine the parameter of the consistency mapping function in the current iteration, thus obtaining the current parameter.

[0048] Specifically, set a distribution matching threshold. The value is 0.04, which achieves a matching effect when the Wasserstein distance is <0.04, thus meeting the calibration quality acceptance standard.

[0049] Step S4: Input the optimal parameters corresponding to each well to be calibrated into the consistency mapping function to obtain the optimal matching calibration model corresponding to each well to be calibrated.

[0050] As an implementable approach, the expression for the optimal matching correction model is: .

[0051] in, The corrected logging curves for the well to be corrected; This is the slope parameter; For normalization; The logging curves of the well to be calibrated; This is the intercept parameter.

[0052] Step S5: Input the logging curves of each well to be corrected into the corresponding optimal matching correction model to obtain the corrected logging curves of each well to be corrected, thus completing the multi-well consistency correction.

[0053] Specifically, the optimal parameters are used to obtain the optimal matching correction model, and then the logging curves of the well to be corrected are used to complete the curve distribution probability morphology standardization process and achieve multi-well consistency correction.

[0054] Taking a certain oilfield as an example, the multi-well consistency correction method based on Wasserstan distance provided in this application is used to perform multi-well consistency correction on multiple target wells in the block. The specific process is as follows.

[0055] (1) Distribution analysis of physical property logging data.

[0056] Using density as the physical property logging data, the density logging curves of standard wells exhibit a normal distribution with a mean of 0.15 and a standard deviation of 0.25. Logging curves of multiple wells to be calibrated: skewed distribution, mainly concentrated in the range [-0.6, 0.2]. (2) Set the key parameters involved in the annealing algorithm.

[0057] Annealing temperature is used to control the probability of accepting inferior solutions.

[0058] (Cooling rate) controls the rate at which the temperature decreases.

[0059] (Distribution matching threshold), optimize termination conditions.

[0060] For example, set the following parameters: Initial parameters: =0.76, =-0.24; initial temperature =1.0; Cooling rate =0.95; Maximum number of iterations = N, set according to machine performance, initially N=200; =0.03.

[0061] Under the control of initial parameters ( =0.76, =-0.24), there is a significant difference between the histogram of the well to be calibrated and the histogram of the standard well (see Figure 3 The corresponding Wassestan distance also has a significant area (see...). Figure 4 This indicates that the density curve of this well has not undergone multi-well consistency processing and cannot be used for seismic inversion or well logging interpretation in subsequent studies, requiring further correction.

[0062] (3) Iterative optimization process.

[0063] Iterations 1-10: Initial Wassstein distance: 0.156; Temperature decreased from 1.0 to 0.63; Acceptance rate 85%, exploring the parameter space; A better solution was found in the 8th iteration: =0.85, =0.15, the Wasserstein distance drops to 0.121.

[0064] Iterations 11-30: Temperature decreased from 0.63 to 0.18; Acceptance rate gradually decreased to 45%, entering a refined search; The current optimum was found in the 25th iteration. =0.92, = -0.08, the Wasserstein distance drops to 0.067.

[0065] Iterations 31-50: Temperature decreased from 0.18 to 0.03; Acceptance rate remained around 30%, local optimization; Breakthrough in iteration 42: =0.96, =-0.05, the Wasserstein distance drops to 0.041.

[0066] Iterations 51-65: When the temperature drops below 0.01°C, the acceptance rate decreases to 15%; the convergence condition is met in iteration 59. =0.98, =-0.03, Wasserstein distance=0.029; the Wasserstein distance of 0.029 is less than the distribution matching threshold of 0.03, the algorithm automatically terminates and the optimization is completed.

[0067] After reaching the convergence threshold, the optimization parameters were determined. =0.98, =-0.03), at this time the Wasserstein distance is 0.029, and the difference between the corrected isobaric histogram of the well to be corrected and the isobaric histogram of the standard well is not significant (see Figure 5 ), and the distribution of equal-width histograms under the initial parameter conditions (see Figure 3 ), cumulative distribution function (CDF) and Wasserstein distance (see Figure 4 This presents a stark contrast; the cumulative distribution functions of the well to be corrected and the standard well show very little difference after correction, and the area enclosed by the Wasserstein distance also becomes very small after correction (see...). Figure 6 This indicates that the optimized density logging curve of the well to be calibrated matches the standard well well well.

[0068] Figure 7 The consistency correction results of density curves (i.e., logging curves) of two example wells (i.e., two wells to be corrected) provided in this application embodiment are shown. Here, GR represents the natural gamma curve, DEN represents the density curve, DEN_NOR_P3W_D represents the corrected density curve (i.e., the corrected logging curve), and P3W1 represents the layer symbol. Comparing the density curves of the two example wells with their corresponding corrected density curves reveals that the corrected density curves within the same layer range are more comparable; the numerical range is more reasonable, especially... Figure 7 In the middle right well, the density curve was locally too small or too large. After correction, the density curve maintained good consistency with the density curve of the adjacent well. Overall, the consistency of the density curves of the two wells after correction was significantly improved compared with that before correction. The correction results of this method can be used for seismic inversion or well logging interpretation in subsequent studies.

[0069] (3) Optimization results.

[0070] Optimal parameters: =0.98, =-0.03; Minimum Wasserstein distance: 0.029; Actual number of iterations: 59.

[0071] (4) Verification of correction effect.

[0072] The distribution patterns of the corrected logging curves of each well to be corrected are highly consistent with those of the standard wells, and the main statistical characteristics match well: the mean deviation decreased from 0.28 to 0.02; the standard deviation decreased from 0.15 to 0.03; and the similarity of the distribution pattern increased by 70%.

[0073] This implementation case demonstrates that the application uses Wassstein distance to comprehensively consider the distribution morphology characteristics, avoiding insufficient matching caused by relying solely on statistics; it employs simulated annealing optimization algorithm to automatically find the optimal consistency mapping parameters, improving correction accuracy and efficiency, and effectively completing the probabilistic morphological standardization correction of the distribution of physical property logging data.

[0074] The beneficial effects of the multi-well consistency correction method based on Wassstein distance proposed in this application are mainly reflected in the following aspects: First, the calibration accuracy is significantly improved. By introducing the Wasserstein distance, the differences in distribution patterns between the standard well and the well to be calibrated are fully considered, avoiding the insufficient matching caused by relying solely on statistics, and achieving more accurate distribution matching. Second, by using the annealing algorithm, the parameters of the consistency mapping are dynamically adjusted with the goal of minimizing the Wasserstein distance. This achieves adaptive parameter adjustment, which is highly automated, reduces manual intervention, and thus improves the accuracy of the corrected logging curves. Third, computational efficiency is optimized by controlling the iteration process by determining whether the Wasserstein distance between the well to be calibrated and the standard well reaches the distribution matching threshold at the current iteration number, thus avoiding unnecessary iterations and reducing computation time by more than 30% while ensuring accuracy.

[0075] Based on the same inventive concept, this application also provides a multi-well consistency correction system based on Wassstein distance. The solution provided by this system is similar to the solution described in the above method. Therefore, the specific limitations of one or more embodiments of the multi-well consistency correction system based on Wassstein distance provided below can be found in the limitations of the multi-well consistency correction method based on Wassstein distance described above, and will not be repeated here.

[0076] In one exemplary embodiment, such as Figure 8 As shown, a multi-well consistency correction system based on Wassstein distance is provided, comprising: Module 1 is used to acquire logging curves of standard wells and logging curves of multiple wells to be calibrated.

[0077] Module 2 is used to construct an equal-width histogram of the standard well based on the logging curves of the standard well.

[0078] Optimization module 3 is used to iteratively optimize the parameters of the consistency mapping function based on the equal width histogram of the standard well and the logging curves of multiple wells to be calibrated, using the empirical distribution function method and with the objective of minimizing the Wasserstein distance between each well to be calibrated and the standard well, by employing the annealing algorithm, to obtain the optimal parameters corresponding to each well to be calibrated.

[0079] The calibration model generation module 4 is used to input the optimal parameters corresponding to each well to be calibrated into the consistency mapping function to obtain the optimal matching calibration model corresponding to each well to be calibrated.

[0080] The calibration module 5 is used to input the logging curves of each well to be calibrated into the corresponding optimal matching calibration model to obtain the calibrated logging curves of each well to be calibrated, thus completing the consistency calibration of multiple wells.

[0081] In one exemplary embodiment, a computer device is provided, which may be a server or a terminal, and its internal structure diagram may be as follows. Figure 9 As shown, this computer device includes a processor, memory, input / output (I / O) interfaces, and a communication interface. The processor, memory, and I / O interfaces are connected via a system bus, and the communication interface is also connected to the system bus via the I / O interfaces. The processor provides computational and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system, computer programs, and a database. The internal memory provides the environment for the operating system and computer programs in the non-volatile storage media to run. The database stores logging curves of standard wells and logging curves of multiple wells to be calibrated. The I / O interfaces are used for exchanging information between the processor and external devices. The communication interface is used for communication with external terminals via a network connection. When the computer program is executed by the processor, it implements a multi-well consistency calibration method based on Wasserstein distance.

[0082] Those skilled in the art will understand that Figure 9 The structures shown are merely block diagrams of some structures related to the present application and do not constitute a limitation on the computer device to which the present application is applied. Specific computer devices may include more or fewer components than shown in the figures, or combine certain components, or have different component arrangements. In an exemplary embodiment, a computer device is provided, including a memory and a processor. The memory stores a computer program, and the processor executes the computer program to implement the steps in the above-described method embodiments.

[0083] Those skilled in the art will understand that all or part of the processes in the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium. When executed, the computer program can include the processes of the embodiments described above. Any references to memory, databases, or other media used in the embodiments provided in this application can include at least one of non-volatile and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetic random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. By way of illustration and not limitation, RAM can take many forms, such as Static Random Access Memory (SRAM) or Dynamic Random Access Memory (DRAM).

[0084] The databases involved in the embodiments provided in this application may include at least one type of relational database and non-relational database. Non-relational databases may include, but are not limited to, blockchain-based distributed databases. The processors involved in the embodiments provided in this application may be general-purpose processors, central processing units, graphics processing units, digital signal processors, programmable logic devices, etc., and are not limited to these.

[0085] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0086] This document uses specific examples to illustrate the principles and implementation methods of this application. The descriptions of the above embodiments are only for the purpose of helping to understand the methods and core ideas of this application. Furthermore, those skilled in the art will recognize that, based on the ideas of this application, there will be changes in the specific implementation methods and application scope. Therefore, the content of this specification should not be construed as a limitation of this application.

Claims

1. A multi-well consistency correction method based on Wassstein distance, characterized in that, The multi-well consistency correction method based on Wassstein distance includes: Obtain logging curves from standard wells and logging curves from multiple wells to be calibrated; Based on the logging curves of standard wells, construct an equal-width histogram of the standard wells; Based on the equal width histogram of the standard well and the logging curves of multiple wells to be calibrated, and using the empirical distribution function method, with the goal of minimizing the Wasserstein distance between each well to be calibrated and the standard well, the parameters of the consistency mapping function are iteratively optimized using the annealing algorithm to obtain the optimal parameters for each well to be calibrated. The optimal parameters corresponding to each well to be corrected are input into the consistency mapping function to obtain the optimal matching correction model corresponding to each well to be corrected. The logging curves of each well to be calibrated are input into the corresponding optimal matching calibration model to obtain the calibrated logging curves of each well, thus completing the multi-well consistency calibration.

2. The multi-well consistency correction method based on Wassstein distance according to claim 1, characterized in that, For any well to be calibrated, based on the isowidth histogram of the standard well and the logging curve of the well to be calibrated, and using the empirical distribution function method, with the objective of minimizing the Wasserstein distance between the well to be calibrated and the standard well, the parameters of the consistency mapping function are iteratively optimized using the annealing algorithm to obtain the optimal parameters corresponding to the well to be calibrated. Specifically, these parameters include: Determine the parameters of the consistency mapping function at the current iteration number to obtain the current parameters; if the current iteration number is the initial iteration number, determine the initial parameters of the consistency mapping function based on the equal-width histogram of the standard well. The current parameters are input into the consistency mapping function to obtain the current matching correction model; Input the logging curve of the well to be calibrated into the current matching calibration model to obtain the current calibrated logging curve of the well to be calibrated. Construct the current equal-width histogram of the well to be corrected based on the current corrected logging curve; The empirical distribution function method is used to calculate the isowidth histogram of the standard well and the current isowidth histogram of the well to be corrected, and the Wasserstein distance between the well to be corrected and the standard well is obtained at the current iteration number. The first parameter is obtained based on the Wasserstein distance between the well to be calibrated and the standard well at the current iteration number and the Wasserstein distance between the well to be calibrated and the standard well at the previous iteration number; the first parameter is the current parameter or the parameter at the previous iteration number; Determine whether the Wasserstein distance between the well to be calibrated and the standard well corresponding to the first parameter reaches the distribution matching threshold; If so, then the first parameter is determined as the optimal parameter; If not, then perform the update process; the update process includes: determining the current annealing temperature based on the annealing temperature of the previous iteration and the current iteration; obtaining the updated parameters based on the current annealing temperature and the first parameter, and using the updated parameters as the parameters of the consistency mapping function for the next iteration, updating the iteration number, and returning to determine the parameters of the consistency mapping function for the current iteration, thus obtaining the current parameters.

3. The multi-well consistency correction method based on Wassstein distance according to claim 2, characterized in that, If the current iteration number is the initial iteration number, the initial parameters of the consistency mapping function are determined based on the equal-width histogram of the standard wells, specifically including: Based on the equal width histogram of the standard well, a standard numerical range is determined; the standard numerical range is the numerical range of the logging curve of the standard well. The initial parameters of the consistency mapping function are determined by using the standard numerical range as a constraint on the numerical range to be corrected; the numerical range to be corrected is the numerical range of the logging curve after the well transformation.

4. The multi-well consistency correction method based on Wassstein distance according to claim 2, characterized in that, The first parameter is obtained based on the Wasserstein distance between the well to be calibrated and the standard well in the current iteration and the Wasserstein distance between the well to be calibrated and the standard well in the previous iteration. Specifically, it includes: Determine whether the Wasserstein distance between the well to be calibrated and the standard well in the current iteration number is less than the Wasserstein distance between the well to be calibrated and the standard well in the previous iteration number, and obtain the first judgment result; If the first determination result is yes, then the current parameter is determined as the first parameter; If the first judgment result is negative, then the Metropolis acceptance criterion is used to determine whether to accept the current parameter, thus obtaining the first parameter.

5. The multi-well consistency correction method based on Wassstein distance according to claim 4, characterized in that, Based on the Metropolis acceptance criterion, a decision is made on whether to accept the current parameter, resulting in the first parameter, which includes: Based on the Metropolis acceptance criteria, a second judgment result is obtained by deciding whether to accept the current parameter. If the second determination result is yes, then the current parameter is determined as the first parameter; If the second judgment result is negative, then the parameter under the previous iteration number is determined as the first parameter.

6. The multi-well consistency correction method based on Wassstein distance according to claim 1, characterized in that, Obtain logging curves from a standard well and logging curves from multiple wells to be calibrated, specifically including: One of the multiple target wells is designated as the standard well, and the remaining wells are designated as wells to be calibrated. Obtain initial logging curves for multiple target wells; Data cleaning was performed on the initial logging data of multiple target wells to obtain the logging curves of the standard wells and the logging curves of multiple wells to be calibrated.

7. The multi-well consistency correction method based on Wassstein distance according to claim 1, characterized in that, The expression for the optimal matching correction model is: ； in, The corrected logging curves for the well to be corrected; The slope parameter; For normalization; The logging curves of the well to be calibrated; This is the intercept parameter.

8. The multi-well consistency correction method based on Wassstein distance according to claim 1, characterized in that, The formula for calculating the Wassstein distance is: ； in, The Wasserstein distance between the well to be calibrated and the standard well; Let be the cumulative distribution function of the well to be corrected; This is the cumulative distribution function for a standard well; The corrected logging curves for the well to be corrected; The corrected logging curves for a standard well; The integral variable represents the range of values ​​for the well logging curve.

9. A multi-well consistency correction system based on Wassstein distance, characterized in that, The multi-well consistency correction system based on Wasselstein distance includes: The acquisition module is used to acquire the logging curves of a standard well and the logging curves of multiple wells to be calibrated; The module is used to construct an isowidth histogram of a standard well based on its logging curves. The optimization module is used to iteratively optimize the parameters of the consistency mapping function based on the equal width histogram of the standard well and the logging curves of multiple wells to be calibrated, using the empirical distribution function method and with the goal of minimizing the Wasserstein distance between each well to be calibrated and the standard well, and to obtain the optimal parameters for each well to be calibrated. The calibration model generation module is used to input the optimal parameters corresponding to each well to be calibrated into the consistency mapping function to obtain the optimal matching calibration model corresponding to each well to be calibrated. The calibration module is used to input the logging curves of each well to be calibrated into the corresponding optimal matching calibration model to obtain the calibrated logging curves of each well and complete the consistency calibration of multiple wells.

10. A computer device, comprising: A memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that the processor executes the computer program to implement the multi-well consistency correction method based on Wassstein distance as described in any one of claims 1-8.

Images