Geological model attribute value correction method and device, storage medium and processor

Abstract

The application discloses a geological model attribute value correction method and device, relates to the technical field of geological model local reconstruction, and mainly aims at improving the quality of a geological model. The main technical scheme of the application is as follows: a target region with a fitting error exceeding a preset threshold in an initial geological model is determined by a preset region determination method. The measured geological attribute values of the target region and preset range nodes are taken as hard constraints, a preset variation function is combined, a random simulation algorithm is adopted to generate multiple sets of target region candidate attribute value distributions, and a preliminary optimization scheme is screened out according to the geological features of the target region. The geological attribute values of the nodes without measured values in the target region of the initial model are replaced by the simulation values corresponding to the preliminary optimization scheme, so that a preliminary correction geological model is obtained. If there is a new target region with an error exceeding the threshold in the preliminary correction geological model, the node simulation values are adjusted according to an error fine-tuning mechanism until the error reaches the standard; if not, the preliminary correction model is directly determined as an optimal geological model.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of local reconstruction of geological models, and in particular to a geological model attribute value correction method and device. BACKGROUND

[0002] A geological model is a digital representation of the subsurface geological structure and attributes (such as permeability, porosity, etc.), which provides basic data for reservoir numerical simulation and is used to predict the fluid distribution and flow rules of the subsurface reservoir. In the model construction process, specific geological attribute values are assigned to each region, which directly reflect the physical properties of the subsurface rock and are key parameters for reservoir numerical simulation. Accurate attribute assignment can construct a geological model that is closer to reality, thereby improving the accuracy of fluid flow simulation and providing a reliable basis for oil exploration and development.

[0003] In the history matching stage of reservoir numerical simulation, it is often found that there is a significant deviation between the actual measured values (such as production and pressure monitoring values in actual production data) in certain regions of the geological model and the simulation results. This deviation is usually due to the difference between the geological attribute values (such as permeability, porosity, etc.) in the corresponding region of the model and the true geological conditions. In order to improve the simulation accuracy, the attribute values of such regions need to be corrected. The current mainstream correction method mainly uses a uniform assignment strategy, such as directly setting the attribute values of the target region to the average value of the surrounding regions or simply applying the parameters of the adjacent regions.

[0004] However, the current uniform assignment method for the modified region of the model ignores the gradual change trend of the attribute values in space and does not fully consider the spatial continuity characteristics of the geological attributes. At the same time, the attribute values of the unmodified region are distributed according to the natural geological laws and have a reasonable spatial variation trend. This leads to the fact that the modified region and the unmodified region cannot effectively connect in terms of attribute variation trend, thereby inevitably causing discontinuity. This discontinuity destroys the natural transition characteristics of the attribute values in the geological space, affects the overall quality of the geological model, and further reduces the accuracy of the reservoir numerical simulation work. SUMMARY

[0005] In view of the above problems, the present application provides a geological model attribute value correction method and device, the main purpose of which is to improve the quality of the geological model.

[0006] To solve the above technical problems, the present application proposes the following solutions:

[0007] In a first aspect, the present application provides a geological model attribute value correction method, which comprises:

[0008] determining a target region in the initial geological model whose fitting error exceeds a preset threshold according to a preset region determination method;

[0009] The measured geologic attribute values of the target region and nodes within a preset range of the target region are taken as hard constraints, a preset variation function is combined, and a plurality of groups of candidate attribute value distribution situations of the target region are generated by using a random simulation algorithm, wherein each group of candidate attribute value distribution situations is generated by using different random seeds;

[0010] According to the geologic features of the target region, a preliminary candidate attribute value distribution situation is selected from the plurality of groups of candidate attribute value distribution situations;

[0011] The geologic attribute values of target nodes in the target region of the initial geologic model that do not have measured geologic attribute values are replaced by corresponding simulated attribute values in the preliminary candidate attribute value distribution situation, so as to obtain a preliminary corrected geologic model;

[0012] According to the preset region determination method, it is determined whether there is a new target region in the preliminary corrected geologic model that still has a fitting error exceeding a preset threshold;

[0013] If there is, the simulated attribute values of the target nodes in the new target region are adjusted according to an error fine-tuning mechanism until the fitting error meets the standard, so as to obtain an optimal geologic model;

[0014] If there is not, the preliminary corrected geologic model is determined as the optimal geologic model.

[0015] In a second aspect, the present application provides an attribute value correction device of a geologic model, and the device comprises:

[0016] A first region determination unit is configured to determine a target region in an initial geologic model that has a fitting error exceeding a preset threshold according to a preset region determination method;

[0017] A distribution simulation unit is configured to take the measured geologic attribute values of the target region and nodes within a preset range of the target region as hard constraints, combine a preset variation function, and generate a plurality of groups of candidate attribute value distribution situations of the target region by using a random simulation algorithm, wherein each group of candidate attribute value distribution situations is generated by using different random seeds;

[0018] A distribution screening unit is configured to select a preliminary candidate attribute value distribution situation from the plurality of groups of candidate attribute value distribution situations obtained by the distribution simulation unit according to the geologic features of the target region;

[0019] An attribute replacement unit is configured to replace the geologic attribute values of target nodes in the target region of the initial geologic model that do not have measured geologic attribute values by corresponding simulated attribute values in the preliminary candidate attribute value distribution situation selected by the distribution screening unit, so as to obtain a preliminary corrected geologic model;

[0020] The second region determining unit is configured to determine, according to the preset region determining method, whether there is a new target region with a fitting error exceeding a preset threshold in the preliminary corrected geological model obtained after the attribute replacing unit replaces the attribute values.

[0021] The model adjusting unit is configured to adjust the simulation attribute values of the target nodes in the new target region according to an error fine-tuning mechanism until the fitting error meets the standard, and obtain an optimal geological model, if the second region determining unit determines that there is the new target region.

[0022] The model determining unit is configured to determine the preliminary corrected geological model as the optimal geological model, if the second region determining unit determines that there is no new target region.

[0023] In order to achieve the above-mentioned purpose, according to a third aspect of the present application, a storage medium is provided, which comprises a stored program, wherein the storage medium controls a device in which the storage medium is located to execute the attribute value correction method of the geological model according to the first aspect when the program is running.

[0024] In order to achieve the above-mentioned purpose, according to a fourth aspect of the present application, a processor is provided, which is used to run a program, wherein the processor executes the attribute value correction method of the geological model according to the first aspect when the program is running.

[0025] According to the technical scheme, the method and device for correcting attribute values of a geological model are provided, first, a target region with a fitting error exceeding a preset threshold in an initial geological model is identified according to a preset region determination method. Then, measured geological attribute values corresponding to nodes in the target region and a preset range around the target region are taken as a hard constraint condition, a preset variation function representing a spatial autocorrelation structure is combined, and a random simulation algorithm is used to generate multiple sets of candidate attribute value distribution in the target region, each set of distribution being generated by a different random seed. Since the random simulation process takes the variation function as a structural control basis and always relies on measured data in the neighborhood, each set of attribute distribution generated is not only strictly matched with well point measured values, but also naturally presents structural continuity defined by the preset variation function in space, effectively reflecting the geological complexity of a real reservoir. Further, based on the geological features of the target region, a preliminarily selected candidate attribute value distribution is selected from the multiple sets of candidate attribute value distribution, and the target nodes without measured geological attribute values in the target region are updated according to the preliminarily selected candidate attribute value distribution, to obtain a preliminarily corrected geological model. To improve the model accuracy, the preset region determination method is applied again to determine whether there is a new target region with a fitting error exceeding a preset threshold; if there is, the attribute values of the target nodes in the new target region are fine-tuned based on an error fine-tuning mechanism until the fitting error meets the standard, and finally an optimal geological model is obtained. Of course, if there is no new target region, the preliminarily corrected geological model can be determined as the optimal geological model. Compared with the way of uniform assignment or linear uniform adjustment of the modification region in the prior art, the present application establishes a spatial statistical correlation between the nodes to be assigned and the measured nodes by the variation function, ensures local data matching, and ensures smooth transition of the attribute field inside the modification region and at the junction with the unmodified region, avoiding attribute mutation or discontinuity caused by one-size-fits-all modification of attribute values, and significantly improving the spatial consistency of the geological model and the reliability of numerical simulation.

[0026] The above description is only a summary of the technical scheme of the present application. In order to more clearly understand the technical means of the present application, the contents of the specification can be implemented, and in order to make the above and other purposes, features and advantages of the present application more obvious and easy to understand, the specific embodiments of the present application are described below. BRIEF DESCRIPTION OF DRAWINGS

[0027] Various other advantages and benefits will become apparent to those of ordinary skill in the art upon reading the following detailed description of the preferred embodiments. The accompanying drawings are included to provide a better understanding of the preferred embodiments, and are not to be considered as limitations on the present application. Moreover, in the entire drawings, the same reference numerals represent the same components. In the drawings:

[0028] Figure 1 A flowchart of a method for correcting attribute values of a geological model provided by an embodiment of the present application is shown;

[0029] Figure 2 Fig. 2 shows a flow chart of another method for modifying attribute values of a geological model according to an embodiment of the present application;

[0030] Figure 3 Fig. 3 shows a block diagram of a device for modifying attribute values of a geological model according to an embodiment of the present application;

[0031] Figure 4 Fig. 4 shows a block diagram of another device for modifying attribute values of a geological model according to an embodiment of the present application. DETAILED DESCRIPTION

[0032] Exemplary embodiments of the present disclosure will be described more fully hereinafter with reference to the accompanying drawings. While exemplary embodiments of the present disclosure are shown in the drawings, it is to be understood that the present disclosure can be embodied in various forms without being limited by the embodiments set forth herein. Rather, these embodiments are provided so that this disclosure will be thorough and complete, and will fully convey the scope of the present disclosure to those skilled in the art.

[0033] In view of the problem of attribute mutation and spatial discontinuity between the modified region and the surrounding unmodified region caused by uniform assignment or simple interpolation in the prior art method for modifying attribute values of a geological model, through a large amount of creative labor, the traditional method of homogenizing or linearizing the error region is abandoned, and the reconstruction of the attribute of the target region is instead regarded as an uncertainty modeling process subject to the dual constraints of the measured geological attribute values and the geological spatial structure.

[0034] Specifically, the measured geological attribute values of the target region and the nodes within the preset range thereof are hard constraints, and a plurality of spatial attribute distribution schemes conforming to geological laws, i.e., a plurality of candidate attribute value distribution conditions, are generated by a random simulation algorithm in combination with a preset variogram reflecting the heterogeneous characteristics of the reservoir; then, the most reasonable candidate attribute value is selected according to the geological features of the target region, such as regional sedimentation or structural features, to complete the preliminary correction; finally, the accuracy of the model is ensured through iterative error detection and local fine-tuning.

[0035] The method ensures the spatial correlation and gradual change trend of the attribute values between the nodes through the variogram, so that the modified node attribute naturally continues the geological change pattern of the surrounding region, fundamentally avoiding the attribute discontinuity problem caused by the one-size-fits-all assignment, and significantly improving the spatial consistency and numerical simulation reliability of the geological model.

[0036] Next, a method for modifying attribute values of a geological model according to the present application will be described in detail. Figure 1 A method for modifying attribute values of a geological model according to the present application will be described in detail, and the specific execution steps thereof are as shown in Fig. 1, including: Figure 1

[0037] ​101. Determine the target region in which the fitting error of the initial geological model exceeds the preset threshold according to the preset region determination method.

[0038] In the present application, first, an initial geological model built by a geological modeling personnel is received; subsequently, a reservoir numerical simulation personnel carries out reservoir numerical simulation based on the model. In the simulation process, the numerical simulator independently solves the percolation equation for each grid cell (usually taking its geometric center as a node), outputs the simulation measurement value corresponding to the node, including the simulation pressure or simulation production (such as the output of oil, water, and gas).

[0039] It should be noted that the geological model is a three-dimensional digital representation of the underground reservoir, including spatial geometric structures (such as layer position, fault) and static geological properties (such as porosity, permeability, lithofacies) corresponding to each spatial position. The model is discretized into a finite number of grid cells, each grid representing a small region of a certain volume underground, and having a unique spatial coordinate and assigned static properties.

[0040] The node here refers to the grid cell or its geometric center, which is the basic unit of numerical calculation. The well point refers to the spatial position of the actual well underground, usually represented in the form of well trajectory. In numerical simulation, the trajectory of a well usually passes through multiple grid cells. Although these grids are all associated with the same well, since each grid is an independent geological volume unit, its static properties (such as porosity, permeability) are different, thus truly reflecting the heterogeneity of the reservoir in the vertical and horizontal directions.

[0041] Although the attribute values and simulation results of each grid cell are independent, in the simulation output stage, the simulation measurement values of all grid cells associated with the same well need to be summarized to obtain the well-level simulation measurement values.

[0042] Correspondingly, the node error in the present application can be defined as follows: first, the actual measurement values of the well are assigned or mapped to each node participating in the production of the well to obtain the actual measurement values corresponding to each node, and then the error index of each node is obtained according to the actual measurement values and simulation measurement values of each node, which is used for subsequent model correction or history matching optimization.

[0043] This is because the actual measurement can only obtain well-level data, while the model parameters exist at the grid level. In order to scientifically use limited observation data to correct the entire geological model, the well-level actual measurement values must be reasonably mapped back to the relevant grid.

[0044] Subsequently, when determining the target region in which the fitting error exceeds the preset threshold, the actual situation that some grid nodes have no corresponding actual well point data needs to be considered. In this regard, the present embodiment supports the following two processing methods:

[0045] The first method directly calculates the error between the simulated and actual measured values ​​of nodes that have corresponding well point actual measured values, and identifies the target area accordingly.

[0046] The second approach extends to all nodes: for nodes without well data, the estimated measurement value can be inferred from nearby or related nodes with actual well data through spatial interpolation, geostatistical methods, or physical correlation models. Then, the error between the simulated value and the inferred value of the node is calculated. Finally, the target area is determined by combining the errors of nodes with and without well data.

[0047] Both of the above methods are applicable in this embodiment. There is no limitation on which one to use. The choice can be made flexibly based on factors such as actual data coverage, modeling accuracy requirements, and computational efficiency.

[0048] The reason for determining the target area based on the error between measured and simulated values, and then adjusting the attribute values ​​corresponding to the grid nodes, is that the accuracy of reservoir numerical simulation is highly dependent on the spatial distribution of subsurface geological properties (such as porosity and permeability), which often have uncertainties in the early stages of modeling. Actual measured values ​​(such as pressure and production at well points) are the only reliable observational basis reflecting the dynamic response of the real reservoir. By comparing the simulation results with the measured data, areas where the model deviates significantly from the actual reservoir behavior can be identified—that is, target areas where the fitting error significantly exceeds the preset threshold.

[0049] By specifically modifying the attribute values ​​of the grid nodes in these areas, the geological model's ability to characterize the heterogeneity of real reservoirs can be effectively improved, thus enhancing the reliability of subsequent simulations and predictions.

[0050] 102. The measured geological attribute values ​​of the target area and the nodes within the preset range of the target area are used as hard constraints. Combined with the preset variation function, the distribution of multiple sets of candidate attribute values ​​of the target area is generated by using a random simulation algorithm.

[0051] In this step, target nodes in the target area that do not have measured geological attribute values ​​are identified. Then, access paths between these target nodes are defined to determine the access order. Following these access paths, based on the measured geological attribute values ​​of the target area and its nodes within a preset range as hard constraints, attribute simulation operations are sequentially performed on each target node using a preset variation function to obtain simulated attribute values ​​for each node. Finally, based on the simulated attribute values ​​of each target node, a set of candidate attribute value distributions for the target area is obtained. Subsequently, the above steps are repeated using different random seeds to generate multiple sets of candidate attribute value distributions for the target area.

[0052] It should be noted that the random seed refers to an integer value used to initialize a pseudo-random number generator, which determines the sequence of sampling results in the random simulation process. Different seeds generate different attribute implementations. In each round of generating candidate attribute distribution, the random simulation algorithm used can be the same (such as using sequential Gaussian simulation multiple times), or different (such as alternating sequential Gaussian simulation and multiple-point geostatistical simulation). Regardless of whether the algorithm is the same, as long as the algorithm includes a random sampling link (almost all geological random simulation algorithms do), a random seed is still needed to control the randomness and repeatability of the simulation, and multiple sets of attribute distribution are generated by changing the seed.

[0053] 103. Selecting a preliminary candidate attribute value distribution from the multiple sets of candidate attribute value distributions according to the geological characteristics of the target region.

[0054] 104. Replacing the geological attribute values of the target nodes in the target region of the initial geological model that do not have measured geological attribute values with the corresponding simulated attribute values in the preliminary candidate attribute value distribution to obtain a preliminary corrected geological model.

[0055] 105. Determining whether there is still a new target region in the preliminary corrected geological model where the fitting error exceeds the preset threshold according to the preset region determination method.

[0056] 106. If there is, adjusting the simulated attribute values of the target nodes in the new target region according to the error fine-tuning mechanism until the fitting error meets the standard to obtain an optimal geological model.

[0057] 107. If there is not, determining the preliminary corrected geological model as the optimal geological model.

[0058] In step 103, first, the multiple sets of candidate attribute value distributions generated in step 102 are obtained; then, these distributions are screened in combination with the known geological characteristics of the target region. The geological characteristics include but are not limited to: sedimentary facies types, sequence structures, sand body connectivity, fault blocking relationships, lithofacies combination rules, and regional permeability-porosity empirical relationships. The purpose of screening is to retain the implementation that not only conforms to the geological rules but also is consistent with the dynamic response trend.

[0059] In step 104, the initial geological model is updated locally based on the preliminary candidate attribute value distribution selected in step 103. Specifically: all grid nodes in the target region that do not have measured geological attribute values (i.e., target nodes) are identified, the attributes of the nodes with measured values are kept unchanged, and the simulated attribute values (such as porosity and permeability) in the preliminary candidate attribute value distribution corresponding to the spatial position are assigned to the above-mentioned target nodes.

[0060] If multiple sets of preliminary candidate attribute value distributions are outputted in step 103, one of them can be selected for replacement, or multiple preliminary corrected geological models can be constructed respectively for subsequent comparative analysis.

[0061] After the attribute value replacement is completed, in step 105, it is determined whether there are new target regions with fitting errors exceeding the preset threshold in each preliminary corrected geological model according to the preset region determination method defined in step 101.

[0062] If multiple preliminary corrected models are constructed in step 104, the determination process needs to be performed on each model respectively. If one or more of the models no longer contain new target regions, the models are preferentially selected as the optimal geological model and step 107 is entered.

[0063] If all the models still contain new target regions, the model with the minimum overall fitting error is selected as the specified preliminary corrected geological model, and step 106 is performed for the next round of iterative correction.

[0064] If only one preliminary corrected geological model is generated in step 104, it is determined whether to jump to step 106 (if there are new target regions) or step 107 (if there are no new target regions) according to the model.

[0065] It should be noted that the new target region described herein and the target region of the previous round can have various spatial relationships: they can be completely the same, or they can be in a containing relationship (e.g., the new target region is a sub-region of the target region), or they can be partially overlapping or completely independent different regions, depending on the spatial evolution characteristics of the model correction error.

[0066] Based on the above Figure 1As can be seen from the implementation mode, the application provides a geological model attribute value correction method, first, a target region with a fitting error exceeding a preset threshold in an initial geological model is identified according to a preset region determination method. Then, measured geological attribute values corresponding to nodes in the target region and a preset range around the target region are taken as a hard constraint condition, a preset variogram representing a spatial autocorrelation structure is combined, and a random simulation algorithm is used to generate multiple sets of candidate attribute value distribution conditions in the target region, and each set of distribution is generated through a different random seed. Since the random simulation process takes the variogram as a structural control basis and always relies on measured data in the neighborhood, each set of attribute distribution generated is not only strictly matched with the well point measured value, but also naturally presents a structural continuity defined by the preset variogram in space, effectively reflecting the geological complexity of the real reservoir. Further, based on the geological features of the target region, a preliminarily selected candidate attribute value distribution condition is selected from the multiple sets of candidate attribute value distribution, and the target nodes without measured geological attribute values in the target region are updated according to the preliminarily selected candidate attribute value distribution condition, to obtain a preliminarily corrected geological model. In order to improve the model accuracy, the preset region determination method is applied again to determine whether there is a new target region with a fitting error exceeding a preset threshold; if there is, the attribute values of the target nodes in the new target region are fine-tuned based on an error fine-tuning mechanism until the fitting error meets the standard, and finally an optimal geological model is obtained. Of course, if there is no new target region, the preliminarily corrected geological model can be determined as the optimal geological model. Compared with the way of uniform assignment or linear uniform adjustment of the modification region in the prior art, the application establishes a spatial statistical correlation between the nodes to be assigned and the measured nodes through the variogram, ensures local data matching, and ensures smooth transition of the attribute field inside the modified region and at the junction with the unmodified region, avoiding attribute mutation or discontinuity caused by one-size-fits-all modification of attribute values, and significantly improving the spatial consistency of the geological model and the reliability of numerical simulation.

[0067] Further, as a refinement and extension of the embodiment shown in Figure 1 , the embodiment of the application further provides another geological model attribute value correction method, as shown in Figure 2 , the specific steps are as follows:

[0068] 201. Determine a target region with a fitting error exceeding a preset threshold in an initial geological model according to a preset region determination method.

[0069] In this embodiment, numerical simulation of the reservoir can be performed according to the initial geological model to obtain simulated measurement values of each grid node in the model, and the simulated measurement values include simulated production values or simulated pressure values. Then, the grid nodes are divided into two categories: one category is nodes crossed by well tracks (referred to as first nodes), and the other category is nodes not crossed by well tracks (referred to as second nodes).

[0070] It should be noted that the actual measurement values (such as the total production of a single well or the bottom hole pressure) are for the whole well, and do not directly belong to a certain node. In order to be used for node-level error calculation, the whole actual measurement value of the well needs to be allocated to each first node through which the well trajectory passes. For example, the total production of the well can be reasonably allocated to these nodes according to the proportion of the contribution of each first node to the well production (determined by the flow characteristics in the simulation); the bottom hole pressure is usually allocated to the first node corresponding to the main producing layer, or allocated to multiple related nodes according to multi-layer test data. In this way, each first node obtains a corresponding node-level actual measurement value.

[0071] Then, according to the difference between the node-level actual measurement value of the first node and its simulated measurement value, the error value of each first node is calculated. For the second node (i.e. the node not passed through by the well trajectory), since there is no direct actual measurement value, it needs to be estimated by means of spatial correlation: using a preset variation function, the spatial correlation weight between each second node and the surrounding first nodes is calculated; then according to these weights and the actual measurement values of the first nodes, the estimated measurement value of the second node is obtained by interpolation; then it is compared with the simulated measurement value of the second node to obtain the error value of the second node.

[0072] Finally, by synthesizing the error values of all first nodes and second nodes, the area whose error significantly exceeds the preset threshold is identified as the target area. The specific implementation steps are as follows:

[0073] First, according to the error values of the first nodes and the second nodes respectively, the nodes whose error values exceed the preset threshold in the initial geological model are screened out and marked as error-exceeding nodes. Then, the error-exceeding nodes that are spatially adjacent to each other (i.e. share a face, edge or corner) are clustered to form one or more connected domains, and each connected domain constitutes a preliminary target sub-area.

[0074] Further, if there is an isolated error-exceeding node that is not clustered, and the distance between the node and any preliminary target sub-area does not exceed the preset adjacency threshold, the isolated node is merged into the corresponding sub-area to avoid missing local abnormal areas.

[0075] After the above processing is completed, a set of expanded preliminary target sub-areas is obtained. At this time, the number and spatial distribution of the sub-areas need to be considered: if there is only one preliminary target sub-area, it is directly determined as the target area; if there are multiple preliminary target sub-areas, the spatial distance between each sub-area is further calculated, and for the sub-areas whose distance is less than or equal to the preset merging threshold, they are merged into a larger target area; the sub-areas that are far apart are kept independently.

[0076] Therefore, the final determined target region can be one or more, depending on the spatial aggregation characteristics of the error distribution. It should be noted that in this step, whether the adjacent threshold used to determine whether the isolated node is included in the sub-region, or the distance standard based on which the multiple preliminary target sub-regions are determined to be merged, can be determined based on the preset variogram function.

[0077] Specifically, in the adjacent threshold and merging threshold steps of target region determination, the core role of the variogram function is to quantify the critical distance of spatial correlation, and to convert the spatial variation rule of geological attributes into a quantitative standard for regional division. The specific application is as follows:

[0078] I. Determination of the adjacent threshold: based on the range of the variogram function, the range of the variogram function refers to the spatial distance when the variogram value tends to be stable, at which point the spatial correlation between nodes is basically eliminated. In this step, the adjacent threshold is set to the range of the variogram function. The logic is as follows: if the distance between the isolated error exceeding node and the preliminary target sub-region is less than the range, it means that the spatial correlation between them is strong, so the isolated node is merged into the sub-region; if the distance is greater than the range, the spatial correlation is weak, and the isolated node is kept separately.

[0079] II. Determination of the merging threshold: also based on the range of the variogram function, for the merging judgment of multiple preliminary target sub-regions, the merging threshold also uses the range of the variogram function. The logic is as follows: if the distance between two preliminary target sub-regions is less than the range, it means that the spatial correlation between them is strong, so they are merged; if the distance is greater than the range, the spatial correlation is weak, which means that it is an independent abnormal region, and each is kept separately. The essence of the above preset variogram function is to convert "spatial continuity of geological attributes" into "quantitative rules of error region division".

[0080] The formula of the variogram function is as follows:

[0081] (Formula 1)

[0082] Among them, the variogram function is an important tool for describing the spatial correlation of geological attributes, which reflects the variation degree and direction characteristics of attribute values in space.

[0083] In the formula, where is the variogram value, which reflects the average degree of attribute difference between two nodes with a distance of h; the smaller γ(h) is, the stronger the spatial correlation is, is the spatial distance, representing the interval between two nodes in three-dimensional space, is the attribute value at position is the attribute value at position is the number of attribute sample pairs with a distance of h.

[0084] The variogram is a core tool for describing the spatial variation of geological properties. It reflects the trend of how the difference in a certain property value (such as porosity or permeability) changes with distance between different locations.

[0085] It is important to clarify that when calculating the spatial correlation weight between each second node and its surrounding first nodes using a pre-defined variogram function, the variogram function quantifies the spatial correlation between the second and first nodes. Specifically, for a second node without a measured value, the γ(h) value between it and its surrounding first nodes with measured values ​​is calculated using the variogram function; the smaller the γ(h) value (i.e., the closer the spatial distance h), the stronger the correlation between the two attributes. Therefore, when deriving the estimated measured value of the second node, a weighted sum of the actual measured values ​​of multiple first nodes can be used (the weights are determined by the γ(h) value), thereby achieving the transfer of measured values ​​from nodes with measured values ​​to nodes without measured values.

[0086] 202. The measured geological attribute values ​​of the target area and the nodes within the preset range of the target area are used as hard constraints. Combined with the preset variation function, the distribution of multiple sets of candidate attribute values ​​of the target area is generated by using a random simulation algorithm.

[0087] In this embodiment, firstly, all grid nodes within the target area that do not possess measured geological attribute values ​​(hereinafter referred to as target nodes) are identified. Then, an access path covering all target nodes is generated to determine the execution order of subsequent simulations. This access path can be generated in any of the following ways:

[0088] 1. Random Path: Randomly shuffle and sort the target nodes;

[0089] 2. Spatial scanning path: Traverse space according to rules such as rows, columns, or spirals;

[0090] 3. Uncertainty-based ranking: Prioritize simulating regions with large errors or little information.

[0091] Regardless of the method used, the path must ensure that each target node is visited only once in a fixed order to guarantee the repeatability of the simulation process. Subsequently, along this path, nodes with measured geological attribute values ​​within the target region and its preset neighborhood are used as hard constraints. Combined with a preset variogram function, conditional random simulation is performed on each target node sequentially to obtain its simulated attribute values ​​(such as porosity and permeability). Combining the simulation results of all target nodes forms a set of candidate attribute value distributions for the target region. Finally, by changing the random seed and repeating the entire process, multiple sets of candidate attribute distributions that satisfy the same geostatistical constraints but have different spatial configurations can be generated.

[0092] When the measured geological attribute values of the nodes in the target region and the preset range of the target region are hard constraints, and the attribute simulation operation is sequentially performed on each target node in combination with the preset variogram, the simulated attribute values of the target nodes are obtained, the following steps can be performed:

[0093] 1. Determine the first target node on the access path, and calculate the spatial correlation weight between the first target node and the nodes having the measured geological attribute values in the target region and the preset range of the target region by using the preset variogram;

[0094] 2. Calculate the attribute estimation value and the estimation variance of the first target node by the Kriging method based on the spatial correlation weight and the measured geological attribute values of the nodes;

[0095] 3. Construct the conditional probability distribution according to the attribute estimation value and the estimation variance, generate a uniform random number in the preset interval based on the current random seed, input the uniform random number as a quantile into the conditional probability distribution, and obtain the simulated attribute value of the first target node;

[0096] 4. For the remaining target nodes except the first target node in the access path, determine the known data of the remaining target nodes, the known data including the measured geological attribute values of the nodes in the target region and the preset range of the target region and the simulated attribute values of all the target nodes having completed simulation in the access path;

[0097] 5. Calculate the spatial correlation weight between the remaining target nodes and the nodes in the known data by using the preset variogram;

[0098] 6. Calculate the attribute estimation value and the estimation variance of the remaining target nodes by the Kriging method based on the spatial correlation weight and the attribute values of the nodes in the known data;

[0099] 7. Construct the conditional probability distribution according to the attribute estimation value and the estimation variance, generate a uniform random number in the preset interval based on the current random seed, input the uniform random number as a quantile into the conditional probability distribution, and obtain the simulated attribute value of the remaining target nodes.

[0100] Mechanism of the variogram:

[0101] In the whole simulation process, the preset variogram is the core basis of the Kriging algorithm. Its role is reflected in the following aspects: for any target node, the variogram outputs the corresponding variogram value γ(h) by calculating the lag distance h between the target node and the surrounding known nodes; in the Kriging algorithm, γ(h) is used to calculate the spatial correlation weight: the smaller γ(h) (i.e., the closer the distance or the stronger the directional correlation), the greater the weight; the estimation value obtained therefrom reflects the local trend, and the estimation variance reflects the uncertainty.

[0102] Finally, the conditional probability distribution constructed by combining the estimated value and the variance ensures that the extracted simulated attribute values not only conform to randomness (driven by the random seed) but also follow the geological spatial rules, i.e., "similar attributes of adjacent nodes and increasing differences of attributes of distant nodes".

[0103] Therefore, the variogram not only controls the spatial structure characteristics of the attribute, but also guarantees the geological rationality and statistical consistency of multiple candidate realizations.

[0104] 203. Selecting, according to the geological characteristics of the target region, a preliminary candidate attribute value distribution from among multiple candidate attribute value distributions.

[0105] 204. Replacing the geological attribute values of the target nodes in the target region of the initial geological model that do not have measured geological attribute values with the corresponding simulated attribute values in the preliminary candidate attribute value distribution to obtain a preliminary corrected geological model.

[0106] 205. Determining, according to a preset region determination method, whether there is still a new target region in the preliminary corrected geological model that has a fitting error exceeding a preset threshold.

[0107] The embodiments of steps 203 to 205 are the same as steps 103-105 and can achieve the same technical effects and solve the same technical problems, and thus will not be repeated here.

[0108] 206. If there is, adjusting the simulated attribute values of the target nodes in the new target region according to an error fine-tuning mechanism until the fitting error meets the standard to obtain an optimal geological model.

[0109] In step 206, first, a first node having an actual measured value and a second node not having an actual measured value in the new target region are determined. Then, according to a preset variogram, the spatial correlation weight of the second node and the first node is determined. Next, the first error of the first node is determined using the actual measured value and the simulated measured value of the first node, and the second error of the second node is calculated in combination with the spatial correlation weight and the first error.

[0110] Then, the original simulated attribute values of each first node in the new target region are adjusted according to the first error of the first node, and the original simulated attribute values of each second node in the new target region are adjusted according to the second error of the second node. This process continues until the errors of all nodes in the preliminary corrected geological model are less than the preset error threshold, thereby obtaining an optimal geological model.

[0111] It needs to be particularly pointed out that the present step is significantly different from the foregoing scheme. The foregoing scheme determines the estimated measurement value of the second node through the correlation weight between the first node and the second node, and then determines the error of the second node according to the estimated measurement value and the corresponding simulated measurement value. In the present step, however, the error of the second node is calculated according to the spatial correlation weight between the first node and the second node and the known error of the first node. The reason for this is that:

[0112] The model currently being processed has been preliminarily corrected, and the attribute distribution thereof is generally reasonable. At this time, the focus is no longer on reconstructing the equivalent observation value of the well node, but on directly using the existing error information for efficient correction. Since the error itself already contains the direction and size of the dynamic fitting deviation, by transmitting the error from the first node with measured data to the second node without measured data through spatial correlation, additional uncertainty caused by interpolation of the estimated measurement value can be avoided, making the adjustment more targeted and physically consistent.

[0113] In addition, another difference between the present step and the foregoing scheme is that the foregoing adjustment is based on the measured geological attribute value of the known node in the target region and the measured geological attribute value of the node within the preset range to simulate the simulated geological attribute value of the unknown node in the target region. In the present step, however, the attribute value is fine-tuned based on the error of the first node and the second node in the measurement value. The reason for this is that:

[0114] The goal of the present step is to fine-tune the existing model, rather than to regenerate the attribute field. Directly fine-tuning the attribute value according to the error can preserve the well-matched parts in the early correction results, only locally perturbing the residual deviation region, thereby improving the convergence efficiency and preventing overfitting or structural distortion.

[0115] Wherein, when adjusting the simulated attribute value of the first node according to the first error between the actual measurement value and the simulated measurement value of the first node, if the simulated production or pressure of the node is lower than the measured value, the key geological attribute value such as permeability or porosity is appropriately increased; otherwise, it is decreased. The adjustment amplitude is proportional to the error size, but only a small correction is made to avoid model oscillation.

[0116] For the second node, the original simulated geological attribute value (here, the original simulated geological attribute refers to the one corresponding to the preliminary corrected geological model) is fine-tuned in the same direction according to the positive and negative and size of the second error. After completing a round of adjustment, numerical reservoir simulation is performed again to update the simulated measurement value of each node and calculate the error again. If the error of a node still exceeds the preset threshold, the fine-tuning process is repeated until the errors of all nodes meet the accuracy requirements, and the optimal geological model is finally obtained. The entire process only makes local and gradual corrections to the existing attribute values, without regenerating the attribute field, thereby efficiently eliminating the residual fitting deviation while preserving the early correction results.

[0117] 207. If not, the preliminary corrected geological model is determined as the optimal geological model.

[0118] The implementation of step 207 is the same as step 107, and can achieve the same technical effects, solve the same technical problems, and thus will not be repeated here.

[0119] Further, as an implementation of the method described above Figure 1 , the embodiment of the present application also provides an attribute value correction device of a geological model, which is used to implement the method described above Figure 1 . The device embodiment corresponds to the foregoing method embodiment, and for the sake of easy reading, the details of the foregoing method embodiment will not be repeated one by one, but it should be clear that the device in the present embodiment can correspondingly implement all the contents in the foregoing method embodiment. As shown in Figure 3 , the device comprises:

[0120] The first area determination unit 301 is configured to determine a target area with a fitting error exceeding a preset threshold in the initial geological model according to a preset area determination method;

[0121] The distribution simulation unit 302 is configured to take the measured geological attribute values of the target area and the nodes within a preset range of the target area determined by the first area determination unit 301 as hard constraints, combine a preset variation function, and generate a plurality of groups of candidate attribute value distribution situations of the target area by using a random simulation algorithm, wherein each group of candidate attribute value distribution situations is generated by using different random seeds;

[0122] The distribution screening unit 303 is configured to screen a preliminary candidate attribute value distribution situation from the plurality of groups of candidate attribute value distribution situations obtained by the distribution simulation unit 302 according to the geological features of the target area;

[0123] The attribute replacement unit 304 is configured to replace the geological attribute values of the target nodes in the target area of the initial geological model that do not have measured geological attribute values with the corresponding simulated attribute values in the preliminary candidate attribute value distribution situation screened by the distribution screening unit 303, to obtain a preliminary corrected geological model;

[0124] The second area determination unit 305 is configured to determine, according to the preset area determination method, whether there is still a new target area with a fitting error exceeding a preset threshold in the preliminary corrected geological model obtained after the attribute replacement unit 304 replaces the attribute values;

[0125] The model adjusting unit 306 is configured to adjust the simulation attribute values of the target nodes in the new target region according to an error fine-tuning mechanism until the fitting error meets the standard, so as to obtain the optimal geologic model, if the result of the second region determining unit 305 is that the target region exists.

[0126] The model determining unit 307 is configured to determine the preliminary corrected geologic model as the optimal geologic model, if the result of the second region determining unit 305 is that the target region does not exist.

[0127] Further, as an implementation of the method shown in the foregoing Figure 2 , the embodiment of the present application further provides another attribute value modification device of a geologic model, which is configured to implement the method shown in the foregoing Figure 2 . The device embodiment corresponds to the foregoing method embodiment, and for the convenience of reading, the details in the foregoing method embodiment will not be described one by one, but it should be clear that the device in the embodiment can correspondingly implement all the contents in the foregoing method embodiment. As shown in the foregoing Figure 4 , the device comprises:

[0128] The first region determining unit 301 is configured to determine a target region in which the fitting error exceeds the preset threshold in the initial geologic model according to a preset region determining method;

[0129] The distribution simulation unit 302 is configured to take the measured geologic attribute values of the nodes in the target region and the preset range of the target region determined by the first region determining unit 301 as hard constraints, combine a preset variation function, and generate multiple sets of candidate attribute value distribution situations of the target region by using a random simulation algorithm, wherein each set of candidate attribute value distribution situations is generated by using different random seeds;

[0130] The distribution screening unit 303 is configured to screen a preliminary candidate attribute value distribution situation from the multiple sets of candidate attribute value distribution situations obtained by the distribution simulation unit 302 according to the geologic features of the target region;

[0131] The attribute replacing unit 304 is configured to replace the geologic attribute values of the target nodes in the target region of the initial geologic model that do not have measured geologic attribute values with the corresponding simulation attribute values in the preliminary candidate attribute value distribution situation screened by the distribution screening unit 303, so as to obtain a preliminary corrected geologic model;

[0132] The second region determining unit 305 is configured to determine whether there is a new target region in which the fitting error exceeds the preset threshold in the preliminary corrected geologic model obtained after the attribute replacing unit 304 replaces the attribute values according to the preset region determining method;

[0133] The model adjusting unit 306 is configured to adjust the simulation attribute values of the target nodes in the new target region according to an error fine-tuning mechanism until the fitting error meets a standard, so as to obtain the optimal geologic model, if the result of the second region determining unit 305 is that the new target region exists.

[0134] The model determining unit 307 is configured to determine the preliminary corrected geologic model as the optimal geologic model, if the result of the second region determining unit 305 is that the new target region does not exist.

[0135] In an optional implementation, the first region determining unit 301 comprises:

[0136] The simulation module 3011 is configured to perform numerical simulation on the initial geologic model to obtain simulation measurement values of the grid nodes in the model, wherein the simulation measurement values comprise simulation production values or simulation pressure values.

[0137] The node determining module 3012 is configured to determine first nodes having actual measurement values and second nodes not having actual measurement values in the grid nodes.

[0138] The first error determining module 3013 is configured to determine error values of the first nodes according to the actual measurement values of the first nodes and the simulation measurement values obtained by the simulation module 3011.

[0139] The estimated measurement module 3014 is configured to determine spatial correlation weights between the second nodes and the first nodes determined by the node determining module 3012 by using a preset variation function, and determine estimated measurement values of the second nodes according to the actual measurement values of the first nodes and the spatial correlation weights.

[0140] The second error determining module 3015 is configured to determine error values of the second nodes according to the estimated measurement values of the second nodes obtained by the estimated measurement module 3014 and the simulation measurement values obtained by the simulation module 3011.

[0141] The region determining module 3016 is configured to determine a target region according to the error values of the first nodes obtained by the first error determining module 3013 and the error values of the second nodes obtained by the second error determining module 3015.

[0142] In an optional implementation, the region determining module 3016 is specifically configured to:

[0143] determine error over-limit nodes in the initial geologic model whose error values exceed a preset threshold according to the error values of the first nodes and the error values of the second nodes;

[0144] cluster the error over-limit nodes adjacent to each other into a connected domain to form a preliminary target sub-region.

[0145] If there is an isolated error exceeding node, and the distance between the isolated error exceeding node and the preliminary target sub-region is not more than the preset adjacency threshold, the isolated error exceeding node is added to the preliminary target sub-region;

[0146] The target region is obtained according to the processed preliminary target sub-region.

[0147] In an optional implementation, when the target region is obtained according to the processed preliminary target sub-region, the region determination module 3016 is specifically configured to:

[0148] If there is one preliminary target sub-region, the preliminary target sub-region is determined as the target region;

[0149] If there are multiple preliminary target sub-regions, the multiple preliminary target sub-regions are merged according to the distances between the multiple preliminary target sub-regions to obtain the target region.

[0150] In an optional implementation, the distribution simulation unit 302 is specifically configured to:

[0151] determine the target nodes in the target region that do not have measured geological attribute values;

[0152] define access paths between the target nodes, the access paths being used to determine the access order of the target nodes;

[0153] perform attribute simulation operations on the target nodes in sequence according to the target region and the measured geological attribute values of the nodes within the preset range of the target region as hard constraints, and in combination with a preset variogram, to obtain simulated attribute values of the target nodes;

[0154] obtain a group of candidate attribute value distribution conditions of the target region according to the simulated attribute values of the target nodes;

[0155] generate multiple groups of candidate attribute value distribution conditions of the target region by repeatedly performing the above steps with different random seeds.

[0156] In an optional implementation, when the attribute simulation operations are performed on the target nodes in sequence according to the target region and the measured geological attribute values of the nodes within the preset range of the target region as hard constraints, and in combination with a preset variogram, to obtain simulated attribute values of the target nodes, the distribution simulation unit 302 is specifically configured to:

[0157] determine a first target node on the access path, and calculate the spatial correlation weight between the first target node and the nodes within the preset range of the target region that have measured geological attribute values by using the preset variogram;

[0158] Based on the spatial correlation weight and the measured geological attribute value of the node, an attribute estimation value and an estimation variance of the first target node are calculated by a Kriging method;

[0159] A conditional probability distribution is constructed according to the attribute estimation value and the estimation variance, and a uniform random number in a preset interval is generated based on a current random seed, the uniform random number is input as a quantile into the conditional probability distribution, and a simulated attribute value of the first target node is obtained;

[0160] For the remaining target nodes except the first target node in the access path, known data of the remaining target nodes are determined, the known data including the measured geological attribute values of the nodes in the target area and a preset range of the target area and the simulated attribute values of all the target nodes in the access path that have been simulated;

[0161] The spatial correlation weight between the remaining target nodes and the nodes in the known data is calculated by using the preset variogram function;

[0162] Based on the spatial correlation weight and the attribute value of the node in the known data, an attribute estimation value and an estimation variance of the remaining target node are calculated by a Kriging method;

[0163] A conditional probability distribution is constructed according to the attribute estimation value and the estimation variance, and a uniform random number in a preset interval is generated based on a current random seed, the uniform random number is input as a quantile into the conditional probability distribution, and a simulated attribute value of the first target node is obtained;

[0164] In an optional implementation, the model adjusting unit 306 is specifically configured to:

[0165] determine a first node having an actual measurement value and a second node not having an actual measurement value in the new target area;

[0166] determine a spatial correlation weight between the second node and the first node according to the preset variogram function;

[0167] determine a first error of the first node according to the actual measurement value and the simulated measurement value of the first node, and determine a second error of the second node according to the spatial correlation weight and the first error;

[0168] adjust the original simulated attribute value of each first node in the new target area according to the first error of the first node, and adjust the original simulated attribute value of each second node in the new target area according to the second error of the second node, until the error of all the nodes in the preliminary corrected geological model is less than a preset error threshold, and an optimal geological model is obtained.

[0169] Further, the embodiment of the present application also provides a storage medium for storing a computer program, wherein the computer program controls a device where the storage medium is arranged to execute the attribute value correction method of the geological model described in the above Figures 1-2

[0170] Further, the embodiment of the present application also provides a processor for running a program, wherein the program executes the attribute value correction method of the geological model described in the above Figures 1-2

[0171] In the above embodiment, the description of each embodiment has its own focus, and the part not described in detail in an embodiment can be referred to the related description of other embodiments.

[0172] It can be understood that the related features in the above method and device can be mutually referred. In addition, the "first", "second" and the like in the above embodiment are used to distinguish the embodiments, and do not represent the advantages and disadvantages of the embodiments.

[0173] Those skilled in the art can clearly understand that, for the convenience and brevity of the description, the specific working process of the system, device and unit described above can refer to the corresponding process in the foregoing method embodiment, which will not be repeated here.

[0174] The algorithms and displays provided herein are not inherently related to any particular computer, virtual system, or other apparatus. Various general purpose systems can be used with programs in accordance with the teachings herein, or it can prove convenient to construct more specialized apparatus to perform the required method steps. The required structure for a variety of these systems will be apparent from the description above. In addition, the present application is not intended to be limited to a particular programming language. It will be appreciated that there are many programming languages that can be used to implement the teachings herein, and any specific language can be chosen for use in this environment. The particular language used is, therefore, not a limitation of the present application.

[0175] In addition, the memory can include non-persistent memory in computer readable medium, random access memory (RAM) and / or non-volatile memory, such as read-only memory (ROM) or flash memory (flash RAM), and the memory includes at least one memory chip.

[0176] Those skilled in the art will appreciate that embodiments of the present application can be a method, a system or a computer program product. Therefore, the present application can take the form of an entirely hardware embodiment, an entirely software embodiment or an embodiment combining software and hardware aspects. Furthermore, the present application can take the form of a computer program product on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROMs, optical storage devices, etc.) embodying computer readable program code thereon.

[0177] ​​The computer program instructions can also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer-implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart block or blocks. Figure 1 one or more flowcharts and / or blocks in the flowcharts and / or combination thereof. Figure 1 one or more flowcharts and / or blocks in the flowcharts and / or combination thereof.

[0178] The computer program instructions can also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer-implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart block or blocks. Figure 1 one or more flowcharts and / or blocks in the flowcharts and / or combination thereof. Figure 1 one or more flowcharts and / or blocks in the flowcharts and / or combination thereof.

[0179] The computer program instructions can also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer-implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart block or blocks. Figure 1 one or more flowcharts and / or blocks in the flowcharts and / or combination thereof. Figure 1 one or more flowcharts and / or blocks in the flowcharts and / or combination thereof.

[0180] In one typical configuration, the computing device includes one or more processors (CPUs), input / output interfaces, network interfaces, and memory.

[0181] The memory can include non-persistent memory and / or volatile memory, such as a random access memory (RAM) including a cache area for the temporary storage of data. A

[0182] Computer-readable media includes permanent and non-permanent, movable and non-movable media that can be implemented by any method or technology to store information. The information can be computer-readable instructions, data structures, program modules or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technologies, compact disc read-only memory (CD-ROM), digital versatile disc (DVD) or other optical storage, magnetic cassette, magnetic tape disk storage or other magnetic storage devices, or any other non-transmission medium that can be used to store information accessible to a computing device. According to the definition herein, computer-readable media does not include transitory media such as modulated data signals and carriers.

[0183] It should also be noted that the terms "comprising", "containing", or any other variant thereof are intended to cover non-exclusive inclusions, such that a process, method, article or apparatus that comprises a list of elements does not only include those elements, but can also include other elements not expressly listed or inherent to such process, method, article or apparatus. Without more limitations, the element defined by the phrase "comprising a" does not exclude the presence of additional identical elements in the process, method, article or apparatus that includes the element.

[0184] Those skilled in the art will appreciate that embodiments of the present application can be provided as a method, a system or a computer program product. Therefore, the present application can take the form of an entirely hardware embodiment, an entirely software embodiment or an embodiment combining software and hardware aspects. Moreover, the present application can take the form of a computer program product implemented on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer usable program code.

[0185] The above is only an embodiment of the present application and is not intended to limit the present application. For those skilled in the art, the present application can have various modifications and changes. Any modification, equivalent replacement, improvement, etc. within the spirit and principle of the present application shall be included in the scope of claims of the present application.

Claims

1. A method of correcting attribute values of a geological model, characterized by, The method comprises: determining a target region in an initial geological model in which fitting error exceeds a preset threshold according to a preset region determination method; using a random simulation algorithm to generate a plurality of candidate attribute value distribution conditions of the target region by taking the measured geological attribute values of the nodes in the target region and the nodes within a preset range of the target region as hard constraints and combining a preset variation function, wherein each candidate attribute value distribution condition is generated by using a different random seed; screening a preliminary candidate attribute value distribution condition from the plurality of candidate attribute value distribution conditions according to the geological features of the target region; replacing the geological attribute values of the target nodes in the target region of the initial geological model that do not have measured geological attribute values with corresponding simulated attribute values in the preliminary candidate attribute value distribution condition to obtain a preliminary corrected geological model; determining whether there is still a new target region in which fitting error exceeds the preset threshold in the preliminary corrected geological model according to the preset region determination method; if there is, adjusting the simulated attribute values of the target nodes in the new target region according to an error fine-tuning mechanism until the fitting error meets the standard to obtain an optimal geological model; if there is not, determining the preliminary corrected geological model as the optimal geological model.

2. The method of claim 1, wherein, The method comprises: performing numerical simulation of an oil reservoir according to an initial geological model to obtain simulated measurement values of each grid node in the model, wherein the simulated measurement values include simulated production values or simulated pressure values; determining first nodes having actual measurement values and second nodes not having actual measurement values among the grid nodes; determining error values of the first nodes according to the actual measurement values and the simulated measurement values of the first nodes; determining spatial correlation weights between each second node and the first nodes by using a preset variation function, and determining estimated measurement values of each second node according to the actual measurement values of the first nodes and the spatial correlation weights; determining error values of each second node according to the estimated measurement values and the simulated measurement values of each second node; determining a target region according to the error values of the first nodes and the error values of the second nodes.

3. The method of claim 2, wherein, The method comprises: determining error-exceeding nodes in the initial geological model in which error values exceed a preset threshold according to the error values of the first nodes and the error values of the second nodes; clustering mutually adjacent error-exceeding nodes into a connected domain to form a preliminary target subregion; if there is an isolated error-exceeding node and the distance between the isolated error-exceeding node and the preliminary target subregion does not exceed a preset adjacency threshold, adding the isolated error-exceeding node to the preliminary target subregion; obtaining a target region according to the processed preliminary target subregion.

4. The method of claim 3, wherein, The method comprises: if there is a preliminary target subregion, determining the preliminary target subregion as the target region; If multiple preliminary target sub-regions exist, the multiple preliminary target sub-regions are merged according to distances between the multiple preliminary target sub-regions to obtain a target region.

5. The method of claim 1, wherein, The measured geological attribute values of the nodes in the target region and the preset range of the target region are taken as hard constraints, and a plurality of sets of candidate attribute value distribution conditions of the target region are generated by using a random simulation algorithm in combination with a preset variation function, including: determining target nodes in the target region that do not have measured geological attribute values; defining access paths between the target nodes, the access paths being used to determine access sequences of the target nodes; along the access paths, performing attribute simulation operations on the target nodes in sequence according to the measured geological attribute values of the nodes in the target region and the preset range of the target region in combination with the preset variation function to obtain simulated attribute values of the target nodes; obtaining a set of candidate attribute value distribution conditions of the target region according to the simulated attribute values of the target nodes; repeating the above steps through different random seeds to generate a plurality of sets of candidate attribute value distribution conditions of the target region.

6. The method of claim 5, wherein, The measured geological attribute values of the nodes in the target region and the preset range of the target region are taken as hard constraints, and attribute simulation operations are performed on the target nodes in sequence in combination with a preset variation function to obtain simulated attribute values of the target nodes, including: determining a first target node on the access path, and calculating spatial correlation weights between the first target node and the nodes in the target region and the preset range of the target region that have measured geological attribute values by using the preset variation function; calculating attribute estimation values and estimation variances of the first target node by Kriging method based on the spatial correlation weights and the measured geological attribute values of the nodes; constructing a conditional probability distribution according to the attribute estimation values and the estimation variances, generating a uniform random number in a preset interval based on a current random seed, inputting the uniform random number as a quantile into the conditional probability distribution to obtain a simulated attribute value of the first target node; for the remaining target nodes except the first target node on the access path, determining known data of the remaining target nodes, the known data including the measured geological attribute values of the nodes in the target region and the preset range of the target region and the simulated attribute values of all the target nodes on the access path that have completed simulation; calculating spatial correlation weights between the remaining target nodes and the nodes in the known data by using the preset variation function; calculating attribute estimation values and estimation variances of the remaining target nodes by Kriging method based on the spatial correlation weights and the attribute values of the nodes in the known data; constructing a conditional probability distribution according to the attribute estimation values and the estimation variances, generating a uniform random number in a preset interval based on a current random seed, inputting the uniform random number as a quantile into the conditional probability distribution to obtain a simulated attribute value of the remaining target node.

7. The method of claim 1, wherein, adjusting the simulated attribute values of the target nodes in the new target region according to an error fine-tuning mechanism until a fitting error is up to standard to obtain an optimal geological model, including: determining a first node having an actual measurement value and a second node not having an actual measurement value in the new target region; determining a spatial correlation weight of the second node and the first node according to the preset variogram; determining a first error of the first node according to the actual measurement value and the simulated measurement value of the first node, and determining a second error of the second node according to the spatial correlation weight and the first error; adjusting an original simulated attribute value of each first node in the new target region according to the first error of the first node, and adjusting an original simulated attribute value of each second node in the new target region according to the second error of the second node, until the error of all nodes in the preliminary corrected geological model is less than a preset error threshold, and an optimal geological model is obtained.

8. An attribute value correction device of a geological model characterized by comprising: The device comprises: a first region determining unit configured to determine a target region having a fitting error exceeding a preset threshold in an initial geological model according to a preset region determination method; a distribution simulation unit configured to generate a plurality of candidate attribute value distribution situations of the target region by using a random simulation algorithm, with the actual geological attribute values of the nodes in the target region and a preset range of the target region determined by the first region determining unit as hard constraints and a preset variogram combined, wherein each candidate attribute value distribution situation is generated by using a different random seed; a distribution screening unit configured to screen a preliminary candidate attribute value distribution situation from the plurality of candidate attribute value distribution situations obtained by the distribution simulation unit according to the geological features of the target region; an attribute replacing unit configured to replace the geological attribute values of the target nodes in the target region of the initial geological model not having actual geological attribute values with the corresponding simulated attribute values in the preliminary candidate attribute value distribution situation screened by the distribution screening unit, and obtain a preliminary corrected geological model; a second region determining unit configured to determine whether there is a new target region having a fitting error exceeding a preset threshold in the preliminary corrected geological model obtained after the attribute replacing unit replaces the attribute values according to the preset region determination method; a model adjusting unit configured to adjust the simulated attribute values of the target nodes in the new target region according to an error fine-tuning mechanism if the second region determining unit determines that there is the new target region, until the fitting error meets the standard, and obtain an optimal geological model; a model determining unit configured to determine the preliminary corrected geological model as the optimal geological model if the second region determining unit determines that there is no new target region.

9. A storage medium, characterized by The storage medium comprises a stored program, wherein the program controls the device where the storage medium is located to execute the attribute value correction method of the geological model according to any one of claims 1 to 7 when the program is running.

10. A processor, comprising: The processor is configured to run a program, wherein the program executes the attribute value correction method of the geological model according to any one of claims 1 to 7 when the program is running.

Images