CN114722649A - Method, device and equipment for continuous variable modeling of meandering stream and storage medium - Google Patents
Method, device and equipment for continuous variable modeling of meandering stream and storage medium Download PDFInfo
- Publication number
- CN114722649A CN114722649A CN202110009594.2A CN202110009594A CN114722649A CN 114722649 A CN114722649 A CN 114722649A CN 202110009594 A CN202110009594 A CN 202110009594A CN 114722649 A CN114722649 A CN 114722649A
- Authority
- CN
- China
- Prior art keywords
- training
- data
- pattern
- grid
- grid node
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000000034 method Methods 0.000 title claims abstract description 65
- 238000012549 training Methods 0.000 claims abstract description 236
- 238000004088 simulation Methods 0.000 claims abstract description 77
- 230000008021 deposition Effects 0.000 claims abstract description 6
- 238000004590 computer program Methods 0.000 claims description 9
- 238000012935 Averaging Methods 0.000 claims description 2
- 230000007547 defect Effects 0.000 abstract description 4
- 230000002349 favourable effect Effects 0.000 abstract 1
- 230000008569 process Effects 0.000 description 17
- 238000010586 diagram Methods 0.000 description 10
- 238000004422 calculation algorithm Methods 0.000 description 9
- 230000006870 function Effects 0.000 description 5
- 238000011160 research Methods 0.000 description 3
- 230000008859 change Effects 0.000 description 2
- 238000005516 engineering process Methods 0.000 description 2
- 230000006872 improvement Effects 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 238000009825 accumulation Methods 0.000 description 1
- 230000009471 action Effects 0.000 description 1
- 230000005540 biological transmission Effects 0.000 description 1
- 238000007635 classification algorithm Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 238000005315 distribution function Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000005065 mining Methods 0.000 description 1
- 230000000877 morphologic effect Effects 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 238000005192 partition Methods 0.000 description 1
- 238000003909 pattern recognition Methods 0.000 description 1
- 230000000704 physical effect Effects 0.000 description 1
- 230000003068 static effect Effects 0.000 description 1
- 238000011426 transformation method Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
- G06F30/20—Design optimisation, verification or simulation
- G06F30/23—Design optimisation, verification or simulation using finite element methods [FEM] or finite difference methods [FDM]
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
- G06F30/20—Design optimisation, verification or simulation
- G06F30/27—Design optimisation, verification or simulation using machine learning, e.g. artificial intelligence, neural networks, support vector machines [SVM] or training a model
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
- G06Q50/02—Agriculture; Fishing; Forestry; Mining
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Evolutionary Computation (AREA)
- General Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- Geometry (AREA)
- Computer Hardware Design (AREA)
- Business, Economics & Management (AREA)
- Health & Medical Sciences (AREA)
- Life Sciences & Earth Sciences (AREA)
- Marketing (AREA)
- Primary Health Care (AREA)
- Strategic Management (AREA)
- Tourism & Hospitality (AREA)
- General Health & Medical Sciences (AREA)
- General Business, Economics & Management (AREA)
- Economics (AREA)
- Human Resources & Organizations (AREA)
- Artificial Intelligence (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Mining & Mineral Resources (AREA)
- Medical Informatics (AREA)
- Software Systems (AREA)
- Marine Sciences & Fisheries (AREA)
- Animal Husbandry (AREA)
- Agronomy & Crop Science (AREA)
- Image Processing (AREA)
Abstract
The embodiment of the invention provides a method, a device, equipment and a storage medium for continuous variable modeling of a meandering stream, wherein the method comprises the following steps: establishing a training image of a continuous variable of the meandering stream according to the simulation grid; scanning the continuous variable training image of the meandering stream through a data template to obtain a plurality of training modes; and scanning the grid nodes by a data sample plate aiming at each empty grid node in the simulation grid, obtaining the training mode of the grid node under the constraint of the soft data trend, determining the training mode corresponding to the training mode of the grid node in a plurality of training modes of the training image of the meandering river, introducing the determined training mode into the grid node, and finishing the simulation of the grid node. The scheme can represent a complex space structure and a geometric form and can reflect the geological deposition rule of the meandering stream; the method is realized under the soft data trend constraint, and is further favorable for overcoming the defect of poor continuity of the riverway of non-trend constraint multi-point geostatistics.
Description
Technical Field
The invention relates to the technical field of oil-gas exploration, in particular to a method, a device, equipment and a storage medium for continuous variable modeling of a meandering river.
Background
With the continuous improvement of oil-gas exploration and development degree and the continuous expansion of research fields, the space structure and the geometric form of an exploration target geologic body are increasingly complex, and the requirement on modeling of the geologic body is higher and higher. Taking the meandering stream as an example, the meandering stream has high degree of curvature of its planar form, complex internal structure, and the lateral direction of the river continuously moves, forming a bank deposit on the concave bank of the river, and during a flood event or under the action of diagenesis, the side accumulation layer is developed inside the bank, and the bank is divided into a plurality of crescent side volumes, resulting in complex physical property spatial distribution and residual oil distribution mode, and great modeling difficulty.
The traditional two-point geostatistical modeling method can only express the correlation between two points based on a variation function, and is difficult to characterize a complex spatial structure and a geometric form (such as a curved river). In the prior art, a new multipoint geostatistical random simulation method based on image elements is provided. Multipoint geostatistics uses a "training image" instead of a variogram to express multipoint-to-multipoint correlations. The training images are digitized images that can represent the actual reservoir structure, geometry, and distribution patterns, which can reflect a priori geological concepts and depositional patterns. The conditional probability distribution function is obtained by scanning a training image for a given data event. Since then, foreign researchers have proposed many new multi-point algorithms or improved algorithms, such as Snesim, simbat, Growth-sim, etc. However, most of the algorithms can only be used for discrete variable simulation and are not suitable for continuous variables.
In addition, during the oil and gas geological research process, deterministic trend surface or volume data can be obtained through seismic attributes or inversion. The data is used for reservoir modeling, can greatly improve the certainty of modeling, is a common thought of a conventional modeling method, and is less applied to multi-point geological modeling.
Therefore, a method capable of representing complex spatial structures and geometric forms and being suitable for continuous variable multi-point geological modeling is not provided in the prior art.
Disclosure of Invention
The embodiment of the invention provides a method for modeling a continuous variable of a meandering stream, which aims to solve the technical problems that a complex spatial structure and a geometric form cannot be represented and the continuity is poor in continuous variable multipoint geological modeling in the prior art. The method comprises the following steps:
establishing a simulation grid according to the logging data;
establishing a continuous meandering river training image according to geological knowledge and a deposition rule;
setting a plurality of data filters, wherein the data filters are a series of weighted values related to a data sample plate, and scanning the training image through the data sample plate to obtain a plurality of training patterns of the training image, wherein each training pattern is the sum of the score values of the plurality of data filters when the data sample plate is scanned each time;
and scanning the grid nodes by the data sample plate aiming at each empty grid node in the simulation grid, obtaining the training mode of the grid node under the constraint of the soft data trend, determining the training mode corresponding to the training mode of the grid node in a plurality of training modes of the training image, introducing the determined training mode into the grid node, and finishing the simulation of the grid node.
The embodiment of the invention also provides a continuous variable modeling device of the meandering stream, which is used for solving the technical problems that the continuous variable multi-point geological modeling in the prior art cannot represent a complex spatial structure and a geometric form and has poor continuity. The device includes:
the simulation grid establishing module is used for establishing a simulation grid according to the logging data;
the training image establishing module is used for establishing a continuous variable training image of the meandering stream according to geological knowledge and a deposition rule;
a training image scanning module, configured to set multiple data filters, where a data filter is a series of weight values related to a data template, and scan the training image through the data template to obtain multiple training patterns of the training image, where each training pattern is a sum of score values of multiple data filters during each scanning of the data template;
and the grid node simulation module is used for scanning the grid node through the data sample plate aiming at each empty grid node in the simulation grid, obtaining a training mode of the grid node under the constraint of soft data trend, determining a training mode corresponding to the training mode of the grid node in a plurality of training modes of the training image, introducing the determined training mode into the grid node, and finishing the simulation of the grid node.
The embodiment of the invention also provides computer equipment which comprises a memory, a processor and a computer program which is stored on the memory and can be operated on the processor, wherein the processor realizes the random meandering stream continuous variable modeling method when executing the computer program so as to solve the technical problems that the continuous variable multi-point geological modeling in the prior art cannot represent complex space structures and geometric forms and is poor in continuity.
The embodiment of the invention also provides a computer readable storage medium, which stores a computer program for executing the random meandering stream continuous variable modeling method, so as to solve the technical problems that the continuous variable multi-point geological modeling in the prior art cannot represent complex spatial structures and geometric forms and is poor in continuity.
In the embodiment of the invention, a simulation grid is established based on logging data, namely, well point data is sampled into the grid to be used as hard data of the simulation grid, then a data sample plate is formed by arranging a plurality of data filters aiming at hollow grid nodes of the simulation grid, then a data sample plate is used for scanning a continuous variable training image of a meandering stream to obtain a plurality of training modes, then the hollow grid nodes are scanned through the data sample plate, the training mode of the hollow grid nodes is obtained under the constraint of soft data trend, and finally the training mode corresponding to the training mode of the hollow grid nodes in the plurality of training modes is introduced into the hollow grid nodes, so that the simulation of the grid nodes is completed, and the next grid node is simulated according to a random path until the simulation of all the hollow grid nodes is completed. Because the simulation grid is established based on the logging data, the empty grid nodes in the simulation grid are simulated in a mode of scanning a training image through a data template to obtain a training mode; because the training image considers the distance of the space variable and the position relation, the establishment of a continuous variable multi-point geostatistical model under the guidance of a geological model is facilitated, and compared with the traditional geostatistical algorithm, the method can represent a complex space structure and a geometric form and can better reflect the geological rule of the meandering river; meanwhile, in the process of simulating the nodes, the method is realized under the soft data trend constraint, and the defect of poor continuity of the riverway of multi-point geostatistics without trend constraint is overcome.
Drawings
The accompanying drawings, which are included to provide a further understanding of the invention and are incorporated in and constitute a part of this application, illustrate embodiment(s) of the invention and together with the description serve to explain the principles of the invention. In the drawings:
fig. 1 is a flowchart of a continuous variable modeling method for a meandering stream according to an embodiment of the present invention;
FIG. 2 is a schematic illustration of raw well log data provided by an embodiment of the present invention;
FIG. 3 is a schematic diagram of a meandering stream porosity training image provided by an embodiment of the present invention;
FIG. 4 is a schematic diagram of a process for obtaining a filter score by a two-dimensional filter for well log data according to an embodiment of the present invention;
FIG. 5 is a schematic diagram of a two-step partition of a two-dimensional filter according to an embodiment of the present invention;
FIG. 6 is a diagram illustrating a result of a trend-free multi-point simulation according to an embodiment of the present invention;
FIG. 7 is a diagram of a soft data constrainer according to an embodiment of the present invention;
FIG. 8 is a diagram illustrating multi-point simulation results under a trend constraint according to an embodiment of the present invention;
FIG. 9 is a schematic diagram of two-point geostatistical simulation results provided by embodiments of the present invention;
fig. 10 is a flowchart for implementing the above-mentioned variable modeling method for the continuous type of the meandering stream according to the embodiment of the present invention;
FIG. 11 is a block diagram of a computer device according to an embodiment of the present invention;
fig. 12 is a structural block diagram of a meandering stream continuous variable modeling apparatus according to an embodiment of the present invention.
Detailed Description
In order to make the objects, technical solutions and advantages of the present invention more apparent, the present invention will be described in further detail with reference to the following embodiments and accompanying drawings. The exemplary embodiments and descriptions of the present invention are provided to explain the present invention, but not to limit the present invention.
In an embodiment of the present invention, a method for modeling a continuous variable of a meandering stream is provided, as shown in fig. 1, the method includes:
step 102: establishing a simulation grid according to the logging data;
step 104: establishing a continuous variable training image of the meandering stream according to geological knowledge and a deposition rule;
step 106: setting a plurality of data filters, wherein the data filters are a series of weight values related to a data sample plate, and scanning the meandering river training image through the data sample plate to obtain a plurality of training modes of the meandering river training image, wherein each training mode is the sum of the score values of the plurality of data filters when the data sample plate is scanned each time;
step 108: and scanning the grid nodes by the data sample plate aiming at each empty grid node in the simulation grid, obtaining the training mode of the grid node under the constraint of the soft data trend, determining the training mode corresponding to the training mode of the grid node in a plurality of training modes of the training image of the meandering stream, and introducing the determined training mode into the grid node to finish the simulation of the grid node.
As can be seen from the flow shown in fig. 1, in the embodiment of the present invention, a simulation grid is established based on logging data, that is, well point data is sampled into the grid to serve as hard data of the simulation grid, a plurality of data filters are arranged for hollow grid nodes of the simulation grid to form a data sample, a training image of a meandering river is scanned by using the data sample to obtain a plurality of training patterns, the hollow grid nodes are scanned by using the data sample, a training pattern of the hollow grid nodes is obtained under the constraint of a soft data trend, and finally, a training pattern corresponding to the training pattern of the hollow grid nodes in the plurality of training patterns is introduced into the hollow grid nodes, so as to complete simulation of the grid nodes, and a next grid node is simulated according to a random path until simulation of all the hollow grid nodes is completed. Because the simulation grid is established based on the logging data, the empty grid nodes in the simulation grid are simulated in a mode of scanning a training image through a data template to obtain a training mode; because the training image considers the distance of the space variable and the position relation, the establishment of a continuous variable multi-point geostatistical model under the guidance of a geological model is facilitated, and compared with the traditional geostatistical algorithm, the method can represent a complex space structure and a geometric form and can better reflect the geological rule of the meandering river; meanwhile, in the process of simulating the nodes, the method is realized under the soft data trend constraint, and the defect of poor continuity of the riverway of multi-point geostatistics without trend constraint is overcome.
In specific implementation, as shown in fig. 10, before the meandering stream continuous variable modeling is implemented, modeling data needs to be prepared, that is, the measured well data may be marked on the nearest grid node to establish a simulation grid. For example, according to the size of the work area, a suitable grid step length is selected, a simulation grid is established, well point data is sampled into the grid through grid coarsening to serve as hard data of the simulation grid, and the hard data is shown in fig. 2.
In specific implementation, as shown in fig. 10, after the simulation grid is established, the continuous meandering stream training image can be established based on the simulation grid, and the process of establishing the continuous meandering stream training image is not specifically limited in the present application and can be implemented by using the existing method. For example, the continuous meandering stream training image can be established by methods such as outcrop modeling or modern sedimentary exploration, fine geological research of a tight well pattern area, core observation and analysis, seismic information mining and the like. The training image of the continuous meandering river established by the modern sedimentary image transformation method is shown in fig. 3.
In specific implementation, after the continuous type meandering stream training image is established, in order to enable the continuous type variable modeling of the meandering stream to reflect the geological rule of the meandering stream, in this embodiment, a data template with a plurality of data filters is proposed to scan the meandering stream training image and the grid nodes in the process of simulating the grid nodes, for example, a plurality of data filters are arranged, the plurality of data filters are a series of weight values related to the data template, and the size of the data filters is consistent with that of the data template, so that the method is conveniently applied to u-shaped meandering stream training image and the grid nodes0A training mode of the center, wherein each node of the data filter corresponds to a relative data template center u0Is offset vector hiI.e. the data filter can be defined as:
{f(hi)；i＝1…J}
TJ＝{u0；hi,i＝1…J}
wherein u is0As a central node of the data template, f (h)i) Is a data filter, i is a grid variable, J is the maximum grid number of the data filter, TJFor searching templates, hi＝(x,y,z)iIs an offset vector, x, y, z are grid coordinates, integers;
then scanning the meandering stream training image through the data sample plate to obtain a plurality of training modes of the meandering stream training image, wherein each training mode is the sum of score values of a plurality of data filters when the data sample plate is scanned each time;
in particular, the data templates mayDefining training patterns by scanning a training image, wherein each training pattern is the sum of score values of a plurality of data filters during each scanning of the data template, as shown in fig. 4, extracting a conditional data event of the training image by the data filter in the scanning process, and then calculating the score value S of the data filter by the following formulaT(u)：
Wherein S isT(u) is the filter score, pat (u + h)i) Is the mode mesh node value; j ═ nx×ny×nzIs the maximum grid data, nx、ny、nzThe number of grids in three directions.
Since a single data filter is not sufficient to capture the information carried by the training pattern, a train of multiple data filters needs to be provided to obtain the various information contained in the training pattern. These data filters form a vector consisting of the sum of the scores of each training pattern
As shown in figure 5 of the drawings,
wherein k is 1, …, F, k is the number of filters;
thus, the dimension of the training pattern is from J to nx×ny×nzReducing to F. For continuous type training images, the F-filter can be used directly to construct the training pattern.
In one embodiment, any combination of an averaging filter, a gradient filter, and a curvature filter may be provided in each direction of the data template.
Specifically, the data filter grid is arranged to be consistent with the data template in the X/Y/Z directions. Suppose the number of nodes in X direction of data template is ni(niIs an odd number),
mi＝(ni-1)/2
αi＝-mi,…,m
average filter:
gradient filter:
curvature filter:
miis half of the number of nodes in a certain direction of the data template, alphaiIs the distance from the center point of the sample plate; by permutation and combination, there are 2 or 3 combinations of data filters for two-dimensional data, i.e. 6 filter choices, and 9 filter choices for three-dimensional data.
In specific implementation, for each empty grid node in the simulation grid, the process of obtaining the training pattern of the grid node by scanning the grid node by the data template is similar to the process of obtaining the training pattern by scanning the training image by the data template, and the process is not repeated again, so that the training pattern corresponding to the training pattern of the grid node can be introduced into the grid node, that is, the data of the training pattern corresponding to the training pattern of the grid node is filled into the grid node to be used as a known data point, so as to complete the simulation of the grid node, and then the next grid node is simulated according to a random path until all the nodes are simulated completely.
In specific implementation, in order to improve the matching of the training patterns, in this embodiment, determining a training pattern corresponding to the training pattern of the mesh node from among the plurality of training patterns of the training image of the meandering river includes:
classifying a plurality of training patterns of the training image of the meandering stream by segmenting a score value space of the training patterns, forming a pattern prototype by the training patterns of the same class, and obtaining a plurality of pattern prototypes, wherein the pattern prototype is the average of all the training patterns in the pattern prototype;
determining a pattern prototype corresponding to the training pattern of the grid node;
and determining any training pattern in the determined pattern prototype as a training pattern corresponding to the training pattern of the grid node.
Specifically, similar training patterns have similar F.K scores. Thus, as shown in FIG. 10, the training patterns can be classified by dividing the score value space, and similar training patterns are classified into a class, i.e., a pattern prototype. A pattern prototype may be defined as the average of all training patterns falling within a certain training pattern class. The pattern prototype and the filter data template are the same size.
For continuous training images, prototype values obtained from a prototype model may be passed through the data template TJTo calculate.
prot(hi) Is the prototype value of a certain classification, hiFor in-data templates TJThe ith offset, c is the number of training images in the classification of prototype model, uj(i-1, …, c) is the center mesh of the particular training pattern.
Specifically, for example, a two-step method may be used to classify the training patterns, as shown in fig. 5, so as to obtain the maximum CPU efficiency.
First, a fast classification algorithm is used to classify all training patterns into a coarser set of patterns, called the parent class (i.e., pattern prototypes). Each parent category represents his own prototype of the pattern. The parent category may include many training patterns, which require further refinement, and similar classification methods may be used to further subdivide the parent category into multiple sub-categories. Each subcategory may have its own prototype of the pattern.
In specific implementation, in the process of determining the pattern prototype corresponding to the training pattern of the mesh node, a minimum distance algorithm may be adopted, for example, the distance between the training pattern of the mesh node and each pattern prototype is calculated; and determining the mode prototype with the minimum distance as the mode prototype corresponding to the training mode of the grid node.
Specifically, the pattern prototype corresponding to the training pattern of the grid node may be calculated by the following minimum distance algorithm formula:
l (dev, prot) is a distance function; dev (u)0+hi) Data template for the region to be simulated is obtained at u0+hiA node value of a location; prot (u)0) Is the score value of the prototype model in a certain classification; w is the weight value of each grid; j is the grid number of the data template; u. of0A center mesh that is a prototype of the pattern; h isiIs the offset of u from the center grid. w is amThe determination of the values is divided into four types of determination: m 1, original hard data; m 2, simulated data; m is 3, reference data; m is 4, soft data; w is amFour types of weight values; n is a radical ofmThe number of the type of grids in the data event; w1+W2+W3+W41, and W1＞W2＞W3,W4
Specifically, after the training patterns are classified to create pattern prototypes (parent categories and child categories), simulation can be started. And (3) accessing the simulation grid G along a random path, wherein at each empty grid node u, the data template T and a data filter template with the same size are used for extracting a conditional data event dev (u), obtaining a training mode of the grid node, and further finding a mode prototype which is closest to the training mode of the grid node through a minimum distance algorithm. If there are sub-categories for the pattern prototype, then a training pattern is randomly selected from the sub-categories to be introduced into the mesh nodes of simulation mesh G. The referenced part is frozen as hard data and then a simulation of the next grid node is performed.
Specifically, a series of linear data filters are arranged, training patterns are classified within the range of the linear data filters, and then the training patterns are matched, so that the purpose of reducing dimensionality can be achieved.
In specific implementation, in order to improve the continuity of simulation grid nodes in the process of obtaining the training mode of the grid nodes by scanning the grid nodes through the data templates so as to further improve the continuity of continuous variable modeling of the meandering stream, the embodiment proposes to obtain the training mode of the grid nodes under the soft data trend constraint, specifically, as shown in fig. 10, when the grid nodes are scanned by the data templates to extract conditional data events, the data templates are used to scan the soft data to extract the conditional data events, and the conditional data events extracted from the soft data are used to replace unmarked grids in the range of the data templates until u is used to scan the grid nodes to extract the conditional data events0And filling all unmarked grids in the range of the data template as the center, obtaining the training mode of the grid node, introducing the training mode corresponding to the training mode of the grid node into the grid node, namely filling the data of the training mode corresponding to the training mode of the grid node into the grid node as a known data point to complete the simulation of the grid node, and simulating the next grid node according to a random path until all the nodes are simulated.
Specifically, if the grid node simulation process is not completed under the soft data trend constraint, the obtained trend-free multi-point simulation result is shown in fig. 6, the soft data is shown in fig. 7 as an example, the grid node simulation process is completed under the soft data trend constraint, the obtained trend-constrained multi-point simulation result is shown in fig. 8, and it can be known by comparison that, under the soft data trend constraint, pattern recognition and matching are performed only by depending on the geological patterns contained in the training images, so that multiple solutions exist, which leads to poor continuity of the meandering stream channel, and even results that partially do not conform to the geological rules. Under the soft data trend constraint, the stability of the simulation result is improved, the river channel shape is natural, the continuity is good, and the geological knowledge is well met.
In particular implementations, the soft data volume may be a spatial trend volume.
In specific implementation, for each empty grid node in the simulation grid, the training image and the grid node are scanned by adopting the data template, and then the data of the training pattern of the training image corresponding to the training pattern of the grid node is filled into the grid node to be used as a known data point so as to complete the simulation of the grid node, and then the next grid node is simulated according to a random path until all the nodes are simulated. The grid node simulation process is similar to the process of selecting pictures from a stack of similar pictures to finish a jigsaw puzzle game, the prior Snesim method stores all training modes in a search tree, and the continuous variable modeling method of the meandering stream just stores central grid points in a memory, so the requirement on a random access memory can be greatly reduced.
Specifically, the porosity model of the meandering stream established by the continuous variable modeling method of the meandering stream is shown in fig. 8, and the porosity model of the meandering stream established by the existing two-point geostatistical method is shown in fig. 9, and the comparison shows that: the traditional two-point geostatistics method can only consider the correlation between two points in space based on a variation function, and is difficult to accurately represent a complex space structure and reproduce the geometric form of a complex target. The method for modeling the continuous variable of the meandering stream applies the training images to replace variation functions, and has the advantages that the distances of space variables are considered, the position relation of the space variables is considered, complex geometric bodies can be expressed, the position relation and the morphological characteristics of the side beaches, the breach fans, the river channels and the lateral volumes of the meandering stream can be well shown, and the change of the porosity of each microphase unit is represented.
In this embodiment, a computer device is provided, as shown in fig. 11, and includes a memory 1102, a processor 1104, and a computer program stored on the memory and executable on the processor, and the processor implements any of the above-mentioned meandering river continuous variable modeling methods when executing the computer program.
In particular, the computer device may be a computer terminal, a server or a similar computing device.
In the present embodiment, there is provided a computer-readable storage medium storing a computer program for executing any of the above-described variable modeling methods of the meandering river continuous type.
In particular, computer-readable storage media, including both non-transitory and non-transitory, removable and non-removable media, may implement information storage by any method or technology. The information may be computer readable instructions, data structures, modules of a program, or other data. Examples of computer-readable 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 technology, compact disc read only memory (CD-ROM), Digital Versatile Discs (DVD) or other optical storage, magnetic cassettes, magnetic tape magnetic disk storage or other magnetic storage devices, or any other non-transmission medium that can be used to store information that can be accessed by a computing device. As defined herein, a computer readable storage medium does not include a transitory computer readable medium such as a modulated data signal and a carrier wave.
Based on the same inventive concept, the embodiment of the invention also provides a continuous variable modeling device for the meandering stream, as described in the following embodiments. The problem solving principle of the meandering stream continuous variable modeling device is similar to that of the meandering stream continuous variable modeling method, so that the implementation of the meandering stream continuous variable modeling device can be referred to that of the meandering stream continuous variable modeling method, and repeated parts are not repeated. As used hereinafter, the term "unit" or "module" may be a combination of software and/or hardware that implements a predetermined function. Although the means described in the embodiments below are preferably implemented in software, an implementation in hardware, or a combination of software and hardware is also possible and contemplated.
Fig. 12 is a block diagram showing a structure of a meandering river continuous variable modeling apparatus according to an embodiment of the present invention, and as shown in fig. 12, the apparatus includes:
a simulation grid establishing module 1202 for establishing a simulation grid according to the logging data;
a training image establishing module 1204, configured to establish a continuous variable training image of the meandering stream according to the simulation grid;
a training image scanning module 1206, configured to set a plurality of data filters, where a data filter is a series of weight values related to a data template, and scan the training image through the data template to obtain a plurality of training patterns of the training image, where each training pattern is a sum of score values of the plurality of data filters at each scanning of the data template;
a grid node simulation module 1208, configured to scan the grid node through the data template for each empty grid node in the simulation grid, obtain a training pattern of the grid node under soft data trend constraint, determine, from among multiple training patterns of the training image, a training pattern corresponding to the training pattern of the grid node, refer the determined training pattern to the grid node, and complete simulation of the grid node.
In an embodiment, the training image scanning module is further configured to scan the grid node by using the data template to extract a conditional data event, scan the soft data by using the data template to extract the conditional data event, and replace an unmarked grid within the range of the data template by using the conditional data event extracted from the soft data to obtain the training pattern of the grid node.
In one embodiment, the grid node simulation module comprises:
the classification unit is used for classifying a plurality of training patterns of the training image of the meandering stream by segmenting a score value space of the training patterns, and the training patterns of the same class form a pattern prototype to obtain a plurality of pattern prototypes, wherein the pattern prototype is the average of all the training patterns in the pattern prototype;
a pattern prototype determining unit for determining a pattern prototype corresponding to the training pattern of the mesh node;
and a training pattern determining unit for determining any one of the determined pattern prototypes as a training pattern corresponding to the training pattern of the mesh node.
In an embodiment, the pattern prototype determining unit is specifically configured to calculate distances between the training patterns of the mesh nodes and the respective pattern prototypes; and determining the mode prototype with the minimum distance as the mode prototype corresponding to the training mode of the grid node.
In one embodiment, the training image scanning module is further configured to set any combination of an average filter, a gradient filter, and a curvature filter in each direction of the data template.
The embodiment of the invention realizes the following technical effects: the simulation method includes the steps that simulation grids are established based on logging data, namely well point data are sampled into the grids and serve as hard data of the simulation grids, then a data sample plate is formed by arranging a plurality of data filters aiming at hollow grid nodes of the simulation grids, a plurality of training modes of a training image of the meandering river are scanned by the data sample plate, the hollow grid nodes are scanned by the data sample plate, the training modes of the hollow grid nodes are obtained under the constraint of soft data trend, finally the training modes corresponding to the training modes of the hollow grid nodes in the training modes are introduced into the hollow grid nodes, the simulation of the grid nodes is completed, the next grid node is simulated according to a random path until all the hollow grid nodes are simulated. Because the simulation grid is established based on the logging data, the empty grid nodes in the simulation grid are simulated in a mode of scanning a training image through a data template to obtain a training mode; because the training image considers the distance of the space variable and the position relation, the establishment of a continuous variable multi-point geostatistical model under the guidance of a geological model is facilitated, and compared with the traditional geostatistical algorithm, the method can represent a complex space structure and a geometric form and can better reflect the geological rule of the meandering river; meanwhile, in the process of simulating the nodes, the method is realized under the soft data trend constraint, and the defect of poor continuity of the riverway of multi-point geostatistics without trend constraint is overcome.
It will be apparent to those skilled in the art that the modules or steps of the embodiments of the invention described above may be implemented by a general purpose computing device, they may be centralized on a single computing device or distributed across a network of multiple computing devices, and alternatively, they may be implemented by program code executable by a computing device, such that they may be stored in a storage device and executed by a computing device, and in some cases, the steps shown or described may be performed in an order different than that described herein, or they may be separately fabricated into individual integrated circuit modules, or multiple ones of them may be fabricated into a single integrated circuit module. Thus, embodiments of the invention are not limited to any specific combination of hardware and software.
The above description is only a preferred embodiment of the present invention, and is not intended to limit the present invention, and various modifications and changes may be made to the embodiment of the present invention by those skilled in the art. Any modification, equivalent replacement, or improvement made within the spirit and principle of the present invention should be included in the protection scope of the present invention.
Claims (12)
1. A continuous variable modeling method for a meandering stream is characterized by comprising the following steps:
establishing a simulation grid according to the logging data;
establishing a continuous variable training image of the meandering stream according to geological knowledge and a deposition rule;
setting a plurality of data filters, wherein the data filters are a series of weighted values related to a data sample plate, and scanning the training image through the data sample plate to obtain a plurality of training patterns of the training image, wherein each training pattern is the sum of the score values of the plurality of data filters when the data sample plate is scanned each time;
and scanning the grid nodes by the data sample plate aiming at each empty grid node in the simulation grid, obtaining the training mode of the grid node under the constraint of the soft data trend, determining the training mode corresponding to the training mode of the grid node in a plurality of training modes of the training image, introducing the determined training mode into the grid node, and finishing the simulation of the grid node.
2. The method as claimed in claim 1, wherein the step of obtaining the training pattern of the grid node under the soft data trend constraint by scanning the grid node through the data template comprises:
and when the data sample plate is used for scanning the grid nodes to extract the conditional data events, the data sample plate is used for scanning the soft data to extract the conditional data events, and the conditional data events extracted from the soft data are used for replacing the unmarked grids in the range of the data sample plate to obtain the training mode of the grid nodes.
3. The method of continuous variable modeling of a meandering stream as claimed in claim 1, wherein determining a training pattern corresponding to the training pattern of the mesh node among a plurality of training patterns of the training image comprises:
classifying a plurality of training patterns of the training image by segmenting a score value space of the training patterns, forming a pattern prototype by the training patterns of the same class, and obtaining a plurality of pattern prototypes, wherein the pattern prototype is the average of all the training patterns in the pattern prototype;
determining a pattern prototype corresponding to the training pattern of the grid node;
and determining any training pattern in the determined pattern prototype as a training pattern corresponding to the training pattern of the grid node.
4. The method of claim 3, wherein determining a pattern prototype corresponding to the training pattern of the grid node comprises:
calculating the distance between the training pattern of the grid node and each pattern prototype;
and determining the mode prototype with the minimum distance as the mode prototype corresponding to the training mode of the grid node.
5. The method according to any one of claims 1 to 4, wherein providing a plurality of data filters comprises:
in each direction of the data template, any combination of an averaging filter, a gradient filter, and a curvature filter is provided.
6. A continuous variable modeling device for a meandering river, comprising:
the simulation grid establishing module is used for establishing a simulation grid according to the logging data;
the training image establishing module is used for establishing a training image of the continuous type variable of the meandering stream according to geological knowledge and a deposition rule;
the training image scanning module is used for setting a plurality of data filters, the data filters are a series of weight values related to the data sample plate, the training images are scanned through the data sample plate, and a plurality of training modes of the training images are obtained, wherein each training mode is the sum of the score values of the plurality of data filters when the data sample plate is scanned each time;
and the grid node simulation module is used for scanning the grid node through the data sample plate aiming at each empty grid node in the simulation grid, obtaining a training mode of the grid node under the constraint of soft data trend, determining a training mode corresponding to the training mode of the grid node in a plurality of training modes of the training image, introducing the determined training mode into the grid node, and finishing the simulation of the grid node.
7. The device of claim 6, wherein the training image scanning module is further configured to scan the soft data with the data template to extract the conditional data event when the grid node is scanned with the data template to extract the conditional data event, and replace the unmarked grid in the range of the data template with the conditional data event extracted from the soft data to obtain the training pattern of the grid node.
8. The meandering river continuous variable modeling apparatus of claim 6, wherein the grid node simulation module comprises:
the classification unit is used for classifying the training patterns of the training image by segmenting the score value space of the training patterns, and the training patterns of the same class form a pattern prototype to obtain a plurality of pattern prototypes, wherein the pattern prototype is the average of all the training patterns in the pattern prototype;
a pattern prototype determining unit for determining a pattern prototype corresponding to the training pattern of the mesh node;
and a training pattern determining unit for determining any one of the determined pattern prototypes as a training pattern corresponding to the training pattern of the mesh node.
9. The meandering stream continuous variable modeling apparatus according to claim 8, wherein said pattern prototype determining unit is specifically configured to calculate distances between the training patterns of the mesh nodes and the respective pattern prototypes; and determining the mode prototype with the minimum distance as the mode prototype corresponding to the training mode of the grid node.
10. The device according to any one of claims 6 to 9, wherein the training image scanning module is further configured to set any combination of an average filter, a gradient filter, and a curvature filter in each direction of the data template.
11. A computer device comprising a memory, a processor and a computer program stored on the memory and executable on the processor, wherein the processor implements the method of the meandering river continuous variable modeling of any one of claims 1 to 5 when executing the computer program.
12. A computer-readable storage medium, characterized in that the computer-readable storage medium stores a computer program for executing the meandering river continuous type variable modeling method according to any one of claims 1 to 5.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110009594.2A CN114722649A (en) | 2021-01-05 | 2021-01-05 | Method, device and equipment for continuous variable modeling of meandering stream and storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110009594.2A CN114722649A (en) | 2021-01-05 | 2021-01-05 | Method, device and equipment for continuous variable modeling of meandering stream and storage medium |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN114722649A true CN114722649A (en) | 2022-07-08 |
Family
ID=82234516
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110009594.2A Pending CN114722649A (en) | 2021-01-05 | 2021-01-05 | Method, device and equipment for continuous variable modeling of meandering stream and storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114722649A (en) |
-
2021
- 2021-01-05 CN CN202110009594.2A patent/CN114722649A/en active Pending
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| WO2021147529A1 (en) | Multipoint geostatistical pre-stack inversion method based on updated theory of permanence of probability ratio | |
| Arpat | Sequential simulation with patterns | |
| Zhang et al. | Filter-based classification of training image patterns for spatial simulation | |
| US9140821B2 (en) | Ordered multipoint geostatistics simulation using non-symmetric search mask | |
| US8311779B2 (en) | Multipoint geostatistics method using branch runlength compression and local grid transformation | |
| AU2013296421B2 (en) | Hybrid method of combining multipoint statistic and object-based methods for creating reservoir property models | |
| AU2013296416B2 (en) | Method for editing a multi-point facies simulation | |
| US9279905B2 (en) | Method for extracting a thumbnail image from a training image so as to constrain the multipoint geostatistical modeling of the subsoil | |
| CN109086437A (en) | A kind of image search method merging Faster-RCNN and Wasserstein self-encoding encoder | |
| CN102057368B (en) | Distribution of properties in a 3D volumetric model using a maximum continuity field | |
| Wu et al. | Fast FILTERSIM simulation with score-based distance | |
| WO2022011893A1 (en) | Reservoir-based modeling method and device for pore network model | |
| CN105989375A (en) | Classifier, classification device and classification method for classifying handwritten character images | |
| CN110927793B (en) | Reservoir prediction method and system based on sequential random fuzzy simulation | |
| CN112861890B (en) | Reservoir evaluation model construction method and reservoir identification method | |
| CN113963123B (en) | Geological modeling method and device integrating deep learning and multipoint statistics | |
| Zuo et al. | Accelerating simulation for the multiple-point statistics algorithm using vector quantization | |
| Strebelle | Multiple-point statistics simulation models: pretty pictures or decision-making tools? | |
| CN114722649A (en) | Method, device and equipment for continuous variable modeling of meandering stream and storage medium | |
| CN112818551A (en) | Novel algorithm for effectively improving accuracy of geographic detector model | |
| CN108986217B (en) | Multipoint geostatistical modeling method based on pattern vector distance | |
| EP3579026B1 (en) | Method of characterising a subsurface region using multiple point statistics | |
| Gourgaris et al. | A density based k-means initialization scheme | |
| Kanai et al. | Improvement of Accuracy for Estimating Permeability Distribution Using Deep Learning | |
| Ge et al. | New classification method for remotely sensed imagery via multiple-point simulation: experiment and assessment |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |