Three-dimensional oil reservoir history fitting method and system based on discrete cosine transform

Abstract

The invention discloses a three-dimensional oil reservoir history fitting method and system based on discrete cosine transform. The method comprises the steps that a permeability field is converted into M1 frequency domains from a spatial domain through discrete cosine transform; disturbing the M1 frequency domains to generate N prior geologic models; the calculation result is analyzed through a Pareto correlation analysis method, and first M2 frequency domains with the maximum correlation with the target function are obtained; analyzing the simulation data of the prior geologic model and the actual data to obtain error data of the simulation data, and judging whether the error data exceeds a preset error threshold value or not; and if the error data exceeds the threshold value, determining a to-be-adjusted parameter from M2 frequency domains based on the error data, and adjusting the to-be-adjusted parameter through an ES-MDA assimilation method, according to the technical scheme of the invention, the method can solve a problem that the automatic historical fitting of the oil reservoir cannot be intelligently and accurately completed in the prior art.

Description

Technical Field

[0001] The invention relates to the field of artificial intelligence technology, and in particular to a three-dimensional reservoir history fitting method and system based on discrete cosine transform. Background Art

[0002] In the field of reservoir numerical simulation, accurate history matching is of vital importance to reservoir dynamic prediction; however, the existing conventional commercial reservoir numerical simulation software mainly adopts the method of adjusting permeability overall or by region when performing history matching; this method has many disadvantages. On the one hand, its adjustment process is highly subjective and lacks a clear zoning basis; on the other hand, in most cases, the adjusted permeability does not match the porosity well and cannot correctly reflect the relationship between the original formation parameters, which leads to deviations in the predicted data.

[0003] In terms of reservoir history matching, most of the existing related research is limited to two-dimensional field parameterization and channel phases, and there are serious deficiencies in dealing with three-dimensional complex non-homogeneous multi-layer reservoirs; even the latest research still remains in two-dimensional field parameterization and channel phases, with limited practical application value, and three-dimensional complex conventional sandstone and carbonate rock real reservoir history matching. Summary of the invention

[0004] In view of this, the present invention proposes a three-dimensional reservoir history fitting method and system based on discrete cosine transform, which can scientifically convert the permeability field, accurately screen the key frequency domain, realize data compression, greatly improve the calculation efficiency, effectively adjust parameters, and improve the accuracy of reservoir history fitting.

[0005] To achieve the above object, the present invention provides the following technical solutions: A three-dimensional reservoir history matching method based on discrete cosine transform, comprising: The permeability field is converted from the spatial domain to M1 frequency domains by discrete cosine transform; Using the Monte Carlo method to perturb the M1 frequency domains, generating N priori geological models; Perform numerical simulation calculations on the N generated priori geological models to obtain calculation results corresponding to each priori geological model; The calculation results are analyzed by a Pareto correlation analysis method to obtain the first M2 frequency domains with the greatest correlation with the objective function; Analyze the simulated data and actual data of the priori geological model to obtain error data of the simulated data, and determine whether the error data exceeds a preset error threshold; If exceeded, determining the parameters to be adjusted from the M2 frequency domains based on the error data, and adjusting the parameters to be adjusted by the ES-MDA assimilation method until the error data of the generated a posteriori geological model does not exceed the preset error threshold; If it does not exceed, it is determined that the 3D reservoir history matching is completed.

[0006] On the basis of the above technical solution, the present invention can also be improved as follows: Furthermore, the step of converting the permeability field from the spatial domain to M1 frequency domains by discrete cosine transform comprises: The coefficients after discrete cosine transformation are calculated by formula (1); Formula (1); In the formula, is the coefficient after discrete cosine transform, , , is the normalization coefficient, , , are the number of grid points in the x, y, and z directions, respectively. is the function value in the original three-dimensional space field, and i, j, k are the spatial indexes in the x, y, and z directions respectively.

[0007] Furthermore, the use of the Monte Carlo method to perturb the M1 frequency domains to generate N priori geological models includes: Each frequency domain is multiplied by the corresponding discrete cosine transformed coefficient and then weighted summed to obtain N prior geological models, where each frequency domain retains different permeability field characteristic information.

[0008] Furthermore, the calculation results are analyzed by the Pareto correlation analysis method to obtain the first M2 frequency domains with the greatest correlation with the objective function, including: The correlation coefficient between the two frequency domains X and Y is calculated by formula (2); Formula (2); In the formula, is the correlation coefficient between the two frequency domains X and Y, is the ith data point of the frequency domain X, is the i-th data point of frequency domain Y, is the average value of X in the frequency domain, is the average value of Y in the frequency domain, is the number of data points.

[0009] Further, the determining the parameters to be adjusted from the M2 frequency domains based on the error data includes: Calculate the posterior geological model parameter values ​​using formula (3); Formula (3); In the formula, is the posterior geological model parameter value, is the prior geological model parameter value, is the regional correlation coefficient, is the covariance coefficient between the prior data, is the variance coefficient of the actual data, is the actual data at the jth position, is the simulated data corresponding to the j-th position of the prior geological model.

[0010] Further, determining the parameters to be adjusted from the M2 frequency domains based on the error data, and adjusting the parameters to be adjusted by using the ES-MDA assimilation method, includes: Decomposing the error data in multiple dimensions, determining M3 frequency domains to be adjusted from the M2 frequency domains, and determining parameters corresponding to the M3 frequency domains to be adjusted as parameters to be adjusted; Determine the adjustment direction and parameter adjustment range of the parameter to be adjusted; The parameter to be adjusted is adjusted based on the adjustment direction and parameter adjustment amplitude range of the parameter to be adjusted.

[0011] Further, the parameter to be adjusted is adjusted by the ES-MDA assimilation method until the error data of the generated a posteriori geological model does not exceed a preset error threshold, including: After adjusting the parameters to be adjusted by the ES-MDA assimilation method, the simulated data and actual data of the posterior geological model are analyzed to obtain error data corresponding to the simulated data, and it is determined whether the error data exceeds a preset error threshold, and the process is iterated continuously until the error data is less than the preset error threshold.

[0012] A three-dimensional reservoir history matching system based on discrete cosine transform, comprising: A discrete cosine transform module, used to transform the permeability field from the spatial domain to M1 frequency domains by discrete cosine transform; A priori geological model generation module is used to perturb the M1 frequency domains using the Monte Carlo method to generate N priori geological models; A numerical simulation calculation module is used to perform numerical simulation calculations on the N generated priori geological models to obtain calculation results corresponding to each priori geological model; A correlation analysis module, used to analyze the calculation results by a Pareto correlation analysis method to obtain the first M2 frequency domains with the greatest correlation with the objective function; The error analysis module is used to analyze the simulated data of the priori geological model and the actual data to obtain the error data of the simulated data and determine whether the error data exceeds a preset error threshold; If exceeded, determining the parameters to be adjusted from the M2 frequency domains based on the error data, and adjusting the parameters to be adjusted by the ES-MDA assimilation method until the error data of the generated a posteriori geological model does not exceed the preset error threshold; If it does not exceed, it is determined that the 3D reservoir history matching is completed.

[0013] An electronic device comprises a memory, a processor and a computer program stored in the memory and running on the processor, wherein the processor implements the steps of the method when executing the computer program.

[0014] A non-transitory computer-readable storage medium stores a computer program, which implements the steps of the method when executed by a processor.

[0015] The present invention has the following advantages: The three-dimensional reservoir history fitting method based on discrete cosine transform in the present invention converts the permeability field into the frequency domain through discrete cosine transform, which can mine the deep characteristics of the data, just like organizing the data in an organized manner, and provides a good foundation for subsequent precise analysis.

[0016] The three-dimensional reservoir history fitting method based on discrete cosine transform in the present invention utilizes the Monte Carlo method to perturb the frequency domain to generate N prior geological models, fully considering multiple possibilities, making the prior model more diverse, more comprehensively covering the possible geological conditions of the reservoir, and avoiding the fitting deviation caused by a single model.

[0017] In the present invention, the three-dimensional reservoir history fitting method based on discrete cosine transform and Pareto correlation analysis can screen out the top M2 frequency domains with the greatest correlation with the objective function, accurately locate the factors that significantly affect the key indicators of the reservoir (such as oil production, water production, etc.), reduce unnecessary calculations, and improve analysis efficiency.

[0018] The three-dimensional reservoir history fitting method based on discrete cosine transform in the present invention can continuously narrow the gap between simulation and reality by comparing the errors between simulation data and actual data, determining the parameters to be adjusted based on the error data and adjusting them using the ES-MDA assimilation method. After iterative optimization, it is ensured that the generated a posteriori geological model can highly accurately reflect the actual situation of the reservoir, providing a reliable basis for reservoir development decisions.

[0019] The three-dimensional reservoir history fitting method based on discrete cosine transform in the present invention constitutes a complete and efficient process from data conversion, prior model generation, correlation analysis to error adjustment. The various links in the whole process are closely coordinated, which can effectively improve the speed and quality of reservoir history fitting and promote more scientific and reasonable reservoir development planning. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] For the purpose of illustration and not limitation, the present invention is now described in conjunction with the embodiments of the present invention and the accompanying drawings, in which: Figure 1 It is a schematic flow chart of a three-dimensional reservoir history fitting method based on discrete cosine transform in an embodiment of the present invention; Figure 2 It is a schematic diagram of the main components of the three-dimensional reservoir history matching system based on discrete cosine transform in an embodiment of the present invention; Figure 3 It is a schematic diagram of the correlation between the multipliers in different frequency domains and the historical oil and water production in an embodiment of the present invention; Figure 4 This is a schematic diagram of the physical structure of the electronic device provided by the present invention. DETAILED DESCRIPTION

[0021] In order to enable those skilled in the art to better understand the solution of the present invention, the technical solution in the implementation of the present invention will be clearly and completely described below in conjunction with the drawings in the implementation of the present invention. Obviously, the described implementation is only a part of the implementation of the present invention, not all of the implementations. Based on the implementation of the present invention, all other implementations obtained by ordinary technicians in this field without creative work should fall within the scope of protection of the present invention.

[0022] It should be noted that the terms "first", "second", etc. in the specification of the present invention and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchanged where appropriate, so as to describe the embodiments of the present invention described herein. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device that includes a series of steps or units is not necessarily limited to those steps or units clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0023] It should be noted that, in the absence of conflict, the embodiments of the present invention and the features in the embodiments can be combined with each other. The embodiments of the present invention will be described in detail below with reference to the accompanying drawings.

[0024] Figure 1FIG. 1 is a flow chart of a three-dimensional reservoir history fitting method based on discrete cosine transform in an embodiment of the present invention. Figure 1 As shown, the three-dimensional reservoir history matching method based on discrete cosine transform provided by the embodiment of the present invention includes the following steps S101 to S107.

[0025] S101, convert the permeability field from the spatial domain to M1 frequency domains by discrete cosine transform.

[0026] Specifically: First, obtain the permeability field data of the oil reservoir. This data is a three-dimensional space field that represents the permeability distribution of the oil reservoir at different spatial locations. The data structure of this three-dimensional space field can be imagined as a large cube composed of countless tiny cubes. Each tiny cube (corresponding to a spatial location) has a specific permeability value corresponding to it, and these values ​​together constitute the permeability field of the oil reservoir.

[0027] Next, the coefficients after discrete cosine transformation are calculated by formula (1); Formula (1); In the formula, is the coefficient after discrete cosine transform; , , are normalization coefficients, and their function is to ensure that the transformed data is within the appropriate numerical range; , , are the number of grid points in the x, y, and z directions, respectively. In the three-dimensional space of the permeability field, a certain number of grid points are divided in each direction. These grid points are used to discretize the space and then perform mathematical calculations; is the function value in the original three-dimensional space field, that is, the permeability value corresponding to each spatial position; i, j, k are the spatial indexes in the x, y, and z directions, respectively, through which each data point in the three-dimensional space field can be located.

[0028] The coefficients after discrete cosine transformation often have very important characteristics. One of the most notable characteristics is energy concentration. After discrete cosine transformation, the energy of the data is not evenly distributed in all coefficients, but concentrated in low-frequency coefficients. In other words, low-frequency coefficients carry most of the energy. Based on this characteristic, we can adopt a data compression strategy to achieve data compression by retaining low-frequency coefficients and discarding high-frequency coefficients. This compression method can greatly reduce the amount of data without losing too much key information.

[0029] Moreover, low-frequency coefficients have another important advantage in the frequency domain, which is that it is easier to extract data features. In the frequency domain, the data features represented by low-frequency coefficients are often related to some macroscopic characteristics of the reservoir, such as the overall permeability trend of the reservoir, the direction of the main seepage channel, etc. By analyzing and processing low-frequency coefficients, key information in the reservoir permeability field can be more efficiently mined, providing strong data support for subsequent reservoir numerical simulation, development plan formulation, and other work.

[0030] S102, using the Monte Carlo method to perturb M1 frequency domains to generate N priori geological models.

[0031] Specifically, for each frequency domain, a set of random coefficients is generated, such as Figure 3 As shown, these random coefficients obey a specific probability distribution (such as uniform distribution or normal distribution), and their value ranges are determined according to reservoir characteristics and actual experience.

[0032] Each frequency domain is multiplied by a corresponding random coefficient and then weighted summed to obtain N prior geological models, where each frequency domain retains different permeability field characteristic information.

[0033] Assume the frequency domain is , the random coefficient is , then the prior geological model It can be expressed as: ; In the formula, is the random coefficient corresponding to the ith frequency domain in the jth prior geological model.

[0034] By repeatedly repeating the above random coefficient generation and weighted summation process, N different prior geological models are generated. The N different prior geological models have different disturbance characteristics in the frequency domain, reflecting the possible different geological structures of the reservoir.

[0035] S103, performing numerical simulation calculations on the generated N priori geological models to obtain calculation results corresponding to each priori geological model.

[0036] Specifically, for each prior geological model, it is used as input and the reservoir numerical simulation software is used for simulation calculation. In the numerical simulation process, it is necessary to consider the fluid flow equations in the reservoir, such as the mass conservation equation and the momentum conservation equation.

[0037] During the simulation, the boundary conditions and initial conditions of the reservoir are set. The boundary conditions include the boundary type of the reservoir (such as closed boundary, constant pressure boundary, etc.) and the physical quantity values ​​on the boundary (such as pressure, flow, etc.), while the initial conditions are the physical quantity distribution of the reservoir at the initial moment (such as initial pressure, initial saturation, etc.).

[0038] The simulation software calculates the dynamic changes of the reservoir over a period of time by solving the fluid flow equation. The calculation results include the changes in physical quantities such as pressure, saturation, flow rate, etc. at different locations in the reservoir over time.

[0039] For each priori geological model, the above numerical simulation process is repeated to obtain N groups of different calculation results, which reflect the dynamic behavior of the reservoir under different priori geological models.

[0040] S104, analyzing the calculation results by using a Pareto correlation analysis method to obtain the first M2 frequency domains with the greatest correlation with the objective function.

[0041] Specifically, the correlation coefficient between the two frequency domains X and Y is calculated by formula (2); Formula (2); In the formula, is the correlation coefficient between the two frequency domains X and Y, is the i-th data point of the frequency domain X, and these data points constitute the data set of the frequency domain x under specific circumstances; is the i-th data point of frequency domain Y, and these data points constitute the data set of frequency domain Y; is the average value of X in the frequency domain, is the average value of Y in the frequency domain, is the number of data points, indicating the total number of data points used in calculating the correlation coefficient.

[0042] The value range is between -1 and 1. =1, indicating that the two frequency domains are completely positively correlated. =-1, indicating that the two frequency domains are completely negatively correlated. =0, indicating that the two frequency domains are unrelated.

[0043] In actual operation, we need to perform such correlation coefficient calculation for each pair of frequency domains. For example, suppose there are M1 frequency domains and we need to calculate In this way, a correlation coefficient matrix is ​​obtained, in which each element represents the correlation between a pair of frequency domains.

[0044] Then, for each frequency domain and the frequency domain corresponding to the objective function (assuming that the objective function can also be expressed in the frequency domain), the correlation coefficient between them is calculated. The objective function is usually a physical quantity related to reservoir production, such as oil production, water production, pressure, etc. Through this calculation, the degree of correlation between each frequency domain and the objective function can be obtained.

[0045] Finally, the frequency domains are sorted according to the calculated correlation coefficients, and the first M2 frequency domains with the largest correlation are selected. The first M2 frequency domains have the most important influence on the objective function. In the subsequent reservoir model adjustment and optimization process, the first M2 frequency domains are focused on to more accurately perform history matching and production prediction for the reservoir.

[0046] S105, analyzing the simulated data of the priori geological model and the actual data to obtain error data of the simulated data, and determining whether the error data exceeds a preset error threshold.

[0047] Specifically, actual production data of the oil reservoir is collected, and the actual production data includes the measured values ​​of physical quantities such as actual oil production, water production, and pressure of the oil well over time.

[0048] For each simulation result of the prior geological model, it is compared with the corresponding actual production data. For example, the simulated oil production is compared with the actual oil production, the simulated pressure is compared with the actual pressure, etc.

[0049] The calculated error data is compared with the preset error threshold. The error threshold is determined according to the production requirements of the reservoir and actual experience. For example, the error threshold of oil production can be set to 5 cubic meters per day, the error threshold of pressure can be set to 0.5 MPa, etc.

[0050] S106, if exceeded, determine the parameters to be adjusted from M2 frequency domains based on the error data, and adjust the parameters to be adjusted by ES-MDA assimilation method until the error data of the generated a posteriori geological model does not exceed the preset error threshold.

[0051] Specifically, the posterior geological model parameter values ​​are calculated using formula (3); Formula (3); In the formula, is the posterior geological model parameter value, is the prior geological model parameter value, which represents the model parameter before this adjustment and is the basis for the adjustment; is the regional correlation coefficient, which reflects the correlation between the prior model and the measured data in a specific area, and plays an important role in determining the amplitude of parameter adjustment; It is the covariance coefficient between the prior data, which describes the correlation of the prior data itself and affects the judgment of the reliability of the prior model; is the variance coefficient of the actual data, which reflects the degree of dispersion of the actual data and helps to evaluate the fluctuation of the actual data; is the actual data at the jth position, which reflects the real situation of the reservoir and is the basis for adjusting the model; is the simulation data corresponding to the jth position of the prior geological model, and is the predicted value of the prior geological model for the reservoir conditions.

[0052] When the error data exceeds the preset error threshold, we first perform a multi-dimensional decomposition of the error data. This decomposition can analyze the source of the error from multiple perspectives, such as the error distribution at different spatial locations, different time points, or different physical quantities (such as pressure, permeability, etc.).

[0053] This decomposition enables us to gain deeper insight into how errors are generated and propagated across the M2 frequency domains.

[0054] According to the distribution of the error data in the frequency domain, M3 frequency domains to be adjusted are determined from the M2 frequency domains. These frequency domains to be adjusted are the frequency domains that contribute more to the error.

[0055] Then, the parameters corresponding to the M3 frequency domains to be adjusted are determined as the parameters to be adjusted. These parameters may include amplitude, phase, etc. in the frequency domain, and their inaccuracy may be the main reason for the error.

[0056] By analyzing the relationship between the error data and the parameters to be adjusted, the adjustment direction of each parameter to be adjusted is determined. For example, if the permeability simulation value corresponding to a certain frequency domain is too high, resulting in a large error, then the parameters related to the permeability in the frequency domain need to be adjusted in the direction of reduction.

[0057] The determination of this adjustment direction requires comprehensive consideration of the physical properties of the reservoir and actual production data to ensure the rationality of the adjustment.

[0058] The adjustment range of each parameter to be adjusted is determined based on the size of the error, the geological complexity of the reservoir, and the uncertainty of the prior model. If the error is large and the reservoir geology is complex, a larger adjustment range may be allowed, but excessive adjustment should also be avoided to prevent model instability.

[0059] The determination of the adjustment range can be achieved by establishing an error-adjustment range relationship model, which can be constructed based on historical data and reservoir engineering experience.

[0060] Adjust the parameters to be adjusted according to the determined adjustment direction and amplitude range. During the adjustment process, ensure that the parameter value is always within a reasonable physical range, for example, the permeability cannot be negative, and the saturation must be between 0 and 1.

[0061] After each adjustment, the posterior geological model is regenerated and numerical simulation calculations are performed to evaluate whether the adjusted model is closer to the actual situation.

[0062] After adjusting the parameters to be adjusted by the ES-MDA assimilation method, we need to reanalyze the simulated data and actual data of the posterior geological model. By calculating the error data between the two, we can determine whether the error data exceeds the preset error threshold.

[0063] If the error data still exceeds the preset error threshold, then it is necessary to repeat the above-mentioned parameter determination, adjustment direction and amplitude range determination, and parameter adjustment process for continuous iteration.

[0064] In each iteration, the ES-MDA assimilation method recalculates the posterior model parameters based on the new prior information (adjusted parameters and models) and actual data, so that the model gradually converges to a state where the error data is less than the preset error threshold, thereby obtaining an accurate posterior geological model and completing the three-dimensional reservoir history fitting.

[0065] S107: If it is not exceeded, it is determined that the three-dimensional reservoir history fitting is completed.

[0066] Specifically, when the error data does not exceed the preset error threshold, it indicates that the current posterior geological model can better fit the actual production situation of the reservoir.

[0067] At this point, it can be considered that the 3D reservoir history fitting process has been successfully completed, and the obtained geological model can be used for subsequent development and production prediction of the reservoir, such as predicting the future production of oil wells and the trend of reservoir pressure changes.

[0068] Figure 2 FIG. 1 is a schematic diagram of the main components of a three-dimensional reservoir history matching system based on discrete cosine transform in an embodiment of the present invention. Figure 2 As shown, the three-dimensional reservoir history fitting system 1 based on discrete cosine transform provided by the embodiment of the present invention includes a discrete cosine transform module 10, a priori geological model generation module 20, a numerical simulation calculation module 30, a correlation analysis module 40, and an error analysis module 50.

[0069] The discrete cosine transform module 10 is used to transform the permeability field from the spatial domain to M1 frequency domains through discrete cosine transform; the coefficients after discrete cosine transform are calculated by formula (1).

[0070] The prior geological model generation module 20 is used to perturb the M1 frequency domains using the Monte Carlo method to generate N prior geological models; for each frequency domain, a set of random coefficients is generated, each frequency domain is multiplied by the corresponding random coefficient, and then weighted summation is performed to obtain N prior geological models, wherein each frequency domain retains different permeability field characteristic information.

[0071] The numerical simulation calculation module 30 is used to perform numerical simulation calculations on the N generated priori geological models to obtain calculation results corresponding to each priori geological model; The correlation analysis module 40 is used to analyze the above calculation results by using the Pareto correlation analysis method to obtain the first M2 frequency domains with the greatest correlation with the objective function; and calculate the correlation coefficient between the two frequency domains X and Y by using formula (2).

[0072] The error analysis module 50 is used to analyze the simulation data and actual data of the prior geological model to obtain the error data of the simulation data, and determine whether the error data exceeds the preset error threshold; if exceeded, determine the parameter to be adjusted from the M2 frequency domains based on the error data, and adjust the parameter to be adjusted by the ES-MDA assimilation method until the error data of the generated a posteriori geological model does not exceed the preset error threshold; if not exceeded, it is determined that the three-dimensional reservoir history fitting has been completed; calculate the a posteriori geological model parameter value by formula (3); perform multi-dimensional decomposition on the error data, determine M3 frequency domains to be adjusted from the M2 frequency domains, and determine the parameters corresponding to the M3 frequency domains to be adjusted as the parameters to be adjusted; determine the adjustment direction and parameter adjustment amplitude range of the parameter to be adjusted; adjust the parameter to be adjusted based on the adjustment direction and parameter adjustment amplitude range of the parameter to be adjusted; and adjust the parameter to be adjusted by the ES-MDA assimilation method. After the MDA assimilation method adjusts the parameters to be adjusted, the simulated data and actual data of the posterior geological model are analyzed to obtain error data corresponding to the simulated data, and it is determined whether the error data exceeds a preset error threshold, and the method is iterated continuously until the error data is less than the preset error threshold.

[0073] Figure 4 A schematic diagram of the physical structure of an electronic device provided in an embodiment of the present invention, such as Figure 4 As shown, the electronic device 60 includes: a processor 601 (processor), a memory 602 (memory) and a bus 603; The processor 601 and the memory 602 communicate with each other via the bus 603; The processor 601 is used to call the program instructions in the memory 602 to execute the methods provided by the above-mentioned method embodiments, for example, including: converting the permeability field from the spatial domain to M1 frequency domains by discrete cosine transform; perturbing the M1 frequency domains by using the Monte Carlo method to generate N priori geological models; performing numerical simulation calculations on the generated N priori geological models to obtain calculation results corresponding to each priori geological model; analyzing the calculation results by using the Pareto correlation analysis method to obtain the first M2 frequency domains with the greatest correlation with the objective function; analyzing the simulated data and actual data of the priori geological model to obtain error data of the simulated data, and judging whether the error data exceeds a preset error threshold; if so, determining the parameters to be adjusted from the M2 frequency domains based on the error data, and adjusting the parameters to be adjusted by using the ES-MDA assimilation method until the error data of the generated a posteriori geological model does not exceed the preset error threshold; if not, judging that the three-dimensional reservoir history fitting has been completed.

[0074] The present embodiment provides a non-transitory computer-readable storage medium, which stores computer instructions. The computer instructions enable a computer to execute the methods provided by the above-mentioned method embodiments, for example, including: converting the permeability field from the spatial domain to M1 frequency domains by discrete cosine transform; perturbing the M1 frequency domains by Monte Carlo method to generate N priori geological models; performing numerical simulation calculations on the generated N priori geological models to obtain calculation results corresponding to each priori geological model; analyzing the calculation results by Pareto correlation analysis method to obtain the first M2 frequency domains with the greatest correlation with the objective function; analyzing the simulated data and actual data of the priori geological model to obtain error data of the simulated data, and judging whether the error data exceeds a preset error threshold; if so, determining the parameters to be adjusted from the M2 frequency domains based on the error data, and adjusting the parameters to be adjusted by ES-MDA assimilation method until the error data of the generated a posteriori geological model does not exceed the preset error threshold; if not, judging that the three-dimensional reservoir history fitting has been completed.

[0075] A person skilled in the art can understand that all or part of the steps of implementing the above method embodiment can be completed by hardware related to program instructions, and the aforementioned program can be stored in a computer-readable storage medium. When the program is executed, it executes the steps of the above method embodiment; and the aforementioned storage medium includes: ROM, RAM, magnetic disk or optical disk, etc., various storage media that can store program codes.

[0076] The above specific implementations do not constitute a limitation on the protection scope of the present invention. It should be understood by those skilled in the art that various modifications, combinations, sub-combinations and substitutions may occur depending on design requirements and other factors. Any modification, equivalent substitution and improvement made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. A three-dimensional reservoir history fitting method based on discrete cosine transform, characterized in that: include: The permeability field is converted from the spatial domain to M1 frequency domains by discrete cosine transform; Using the Monte Carlo method to perturb the M1 frequency domains, generating N priori geological models; Perform numerical simulation calculations on the N generated priori geological models to obtain calculation results corresponding to each priori geological model; The calculation results are analyzed by a Pareto correlation analysis method to obtain the first M2 frequency domains with the greatest correlation with the objective function; Analyze the simulated data and actual data of the priori geological model to obtain error data of the simulated data, and determine whether the error data exceeds a preset error threshold; If exceeded, determining the parameters to be adjusted from the M2 frequency domains based on the error data, and adjusting the parameters to be adjusted by the ES-MDA assimilation method until the error data of the generated a posteriori geological model does not exceed the preset error threshold; If it does not exceed, it is determined that the 3D reservoir history matching is completed.

2. The three-dimensional reservoir history fitting method based on discrete cosine transform according to claim 1 is characterized in that: The permeability field is converted from the spatial domain to M1 frequency domains by discrete cosine transform, comprising: The coefficients after discrete cosine transformation are calculated by formula (1); Formula (1): In the formula, is the coefficient after discrete cosine transform, , , is the normalization coefficient, , , are the number of grid points in the x, y, and z directions, respectively. is the function value in the original three-dimensional space field, and i, j, k are the spatial indexes in the x, y, and z directions respectively.

3. The three-dimensional reservoir history fitting method based on discrete cosine transform according to claim 1 is characterized in that: The method of using the Monte Carlo method to perturb the M1 frequency domains to generate N priori geological models includes: For each frequency domain, a set of random coefficients is generated, each frequency domain is multiplied by the corresponding random coefficient, and then weighted summation is performed to obtain N prior geological models, where each frequency domain retains different permeability field characteristic information.

4. The three-dimensional reservoir history matching method based on discrete cosine transform according to claim 1 is characterized in that: The calculation results are analyzed by the Pareto correlation analysis method to obtain the first M2 frequency domains with the greatest correlation with the objective function, including: The correlation coefficient between the two frequency domains X and Y is calculated by formula (2); Formula (2): In the formula, is the correlation coefficient between the two frequency domains X and Y, is the ith data point of the frequency domain X, is the i-th data point of frequency domain Y, is the average value of X in the frequency domain, is the average value of Y in the frequency domain, is the number of data points.

5. The three-dimensional reservoir history matching method based on discrete cosine transform according to claim 1, characterized in that: The determining the parameters to be adjusted from the M2 frequency domains based on the error data includes: Calculate the posterior geological model parameter values ​​using formula (3); Formula (3): In the formula, is the posterior geological model parameter value, is the prior geological model parameter value, is the regional correlation coefficient, is the covariance coefficient between the prior data, is the variance coefficient of the actual data, is the actual data at the jth position, is the simulated data corresponding to the prior geological model at the jth position.

6. The three-dimensional reservoir history matching method based on discrete cosine transform according to claim 1, characterized in that: The step of determining the parameters to be adjusted from the M2 frequency domains based on the error data, and adjusting the parameters to be adjusted by using the ES-MDA assimilation method, comprises: Decomposing the error data in multiple dimensions, determining M3 frequency domains to be adjusted from the M2 frequency domains, and determining parameters corresponding to the M3 frequency domains to be adjusted as parameters to be adjusted; Determine the adjustment direction and parameter adjustment range of the parameter to be adjusted; The parameter to be adjusted is adjusted based on the adjustment direction and parameter adjustment amplitude range of the parameter to be adjusted.

7. The three-dimensional reservoir history matching method based on discrete cosine transform according to claim 6 is characterized in that: The adjusting the parameters to be adjusted by the ES-MDA assimilation method until the error data of the generated a posteriori geological model does not exceed a preset error threshold includes: After adjusting the parameters to be adjusted by the ES-MDA assimilation method, the simulated data and actual data of the posterior geological model are analyzed to obtain error data corresponding to the simulated data, and it is determined whether the error data exceeds a preset error threshold, and the process is iterated continuously until the error data is less than the preset error threshold.

8. A system for three-dimensional reservoir history matching based on discrete cosine transform, characterized in that: include: A discrete cosine transform module, used to transform the permeability field from the spatial domain to M1 frequency domains by discrete cosine transform; A priori geological model generation module is used to perturb the M1 frequency domains using the Monte Carlo method to generate N priori geological models; A numerical simulation calculation module is used to perform numerical simulation calculations on the N generated priori geological models to obtain calculation results corresponding to each priori geological model; A correlation analysis module, used to analyze the calculation results by a Pareto correlation analysis method to obtain the first M2 frequency domains with the greatest correlation with the objective function; The error analysis module is used to analyze the simulated data and actual data of the priori geological model to obtain error data of the simulated data and determine whether the error data exceeds a preset error threshold; If exceeded, determining the parameters to be adjusted from the M2 frequency domains based on the error data, and adjusting the parameters to be adjusted by the ES-MDA assimilation method until the error data of the generated a posteriori geological model does not exceed the preset error threshold; If it does not exceed, it is determined that the 3D reservoir history matching is completed.

9. An electronic device comprising a memory, a processor, and a computer program stored in the memory and running on the processor, characterized in that: When the processor executes the computer program, the steps of the method according to any one of claims 1 to 7 are implemented.

10. A non-transitory computer readable medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 7 are implemented.

Images