CN113985493B - Intelligent modeling method for underground multi-information constrained isochronous stratum grillwork - Google Patents
Intelligent modeling method for underground multi-information constrained isochronous stratum grillwork Download PDFInfo
- Publication number
- CN113985493B CN113985493B CN202111302831.0A CN202111302831A CN113985493B CN 113985493 B CN113985493 B CN 113985493B CN 202111302831 A CN202111302831 A CN 202111302831A CN 113985493 B CN113985493 B CN 113985493B
- Authority
- CN
- China
- Prior art keywords
- well
- samples
- training
- sample
- carrying
- 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.)
- Active
Links
- 238000000034 method Methods 0.000 title claims abstract description 52
- 238000004458 analytical method Methods 0.000 claims abstract description 4
- 238000012549 training Methods 0.000 claims description 56
- 238000013528 artificial neural network Methods 0.000 claims description 31
- 238000004519 manufacturing process Methods 0.000 claims description 13
- 238000004364 calculation method Methods 0.000 claims description 7
- 238000012937 correction Methods 0.000 claims description 6
- 239000000463 material Substances 0.000 claims description 6
- 238000005457 optimization Methods 0.000 claims description 3
- 238000013473 artificial intelligence Methods 0.000 abstract description 4
- 238000003062 neural network model Methods 0.000 abstract description 4
- 230000008021 deposition Effects 0.000 abstract 1
- 238000005516 engineering process Methods 0.000 description 6
- 239000004576 sand Substances 0.000 description 5
- 238000010276 construction Methods 0.000 description 4
- 230000009286 beneficial effect Effects 0.000 description 3
- 238000011161 development Methods 0.000 description 3
- 238000001514 detection method Methods 0.000 description 2
- 238000009826 distribution Methods 0.000 description 2
- 230000000694 effects Effects 0.000 description 2
- 238000012545 processing Methods 0.000 description 2
- 230000033764 rhythmic process Effects 0.000 description 2
- 230000015572 biosynthetic process Effects 0.000 description 1
- 238000004891 communication Methods 0.000 description 1
- 238000005520 cutting process Methods 0.000 description 1
- 238000013434 data augmentation Methods 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000000605 extraction Methods 0.000 description 1
- 238000001914 filtration Methods 0.000 description 1
- 230000006870 function Effects 0.000 description 1
- 230000006698 induction Effects 0.000 description 1
- 238000007781 pre-processing Methods 0.000 description 1
- 238000002360 preparation method Methods 0.000 description 1
- 238000003672 processing method Methods 0.000 description 1
- 238000005070 sampling Methods 0.000 description 1
- 238000000638 solvent extraction Methods 0.000 description 1
- 238000012795 verification Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01V—GEOPHYSICS; GRAVITATIONAL MEASUREMENTS; DETECTING MASSES OR OBJECTS; TAGS
- G01V11/00—Prospecting or detecting by methods combining techniques covered by two or more of main groups G01V1/00 - G01V9/00
Landscapes
- Physics & Mathematics (AREA)
- Life Sciences & Earth Sciences (AREA)
- General Life Sciences & Earth Sciences (AREA)
- General Physics & Mathematics (AREA)
- Geophysics (AREA)
- Geophysics And Detection Of Objects (AREA)
Abstract
The invention provides an intelligent modeling method for an underground multi-information constrained isochronous stratum grid, which is characterized in that high-precision small-layer division results are obtained by utilizing an artificial intelligence method through comprehensive geological information and logging information, and inter-well horizons conforming to geological laws are obtained by means of seismic information based on the results, so that the small-layer point-to-surface division process is realized, and finally, the high-precision isochronous stratum grid model can be obtained. The invention can accelerate the convergence of the neural network model, refine the deposition evolution analysis of complex and other surfaces; the working efficiency can be greatly improved through an automatic procedure under the condition of more wells, and the intelligent small-layer division precision can be effectively improved through the method.
Description
Technical Field
The invention belongs to the technical field of geological data processing, and particularly relates to an intelligent modeling method for an underground multi-information constrained isochronous stratigraphic framework.
Background
Traditional isochronous stratigraphic grid construction is based on fine stratigraphic horizons, and the horizon sources of small strata are mainly 2 methods:
determining large horizon division (oil group interface) by well-seismic combination, subdividing small strata based on logging and sand information, and finally adjusting the division result by means of seismic data reflection characteristics; for example, references 1, huo Chunliang, gu Li, zhao Chunming, weipeng, yang Qinggong. Reservoir fine modeling based on integrated seismic, logging and geology [ J ]. Petro-journal, 2007 (06): 66-71. Propose "gyratory contrast, hierarchical control" thinking: (1) well-shock bonding determines the oil group interface and establishes a "coarse" isochronal stratigraphic framework. (2) By utilizing the characteristic of high vertical resolution of logging data, a traditional equal elevation comparison method and a logging curve similarity comparison method are adopted, and well point fine layering is carried out by taking a sand body as a unit; based on well earthquake calibration, taking earthquake inversion data as a background to make a well-connecting section; and then adjusting the comparison layer position according to whether the comparison result between wells is consistent with the seismic reflection characteristics or not, and establishing a fine stratum frame.
The division of the small layers is obtained by interpolation based on the oil group layer and the rotation characteristic information. For example, reference 2, gaobao Yu, sun Lichun, hu Guangyi, zhang Yuan. River phase reservoir geological modeling method based on single sand body discusses [ J ]. Chinese offshore oil and gas, 2008 (01): 34-37. Using well point information, using layer sequence stratigraphy principle to divide stratum layer system into 4 oil groups; then, carrying out well calibration on seismic data through the time-depth relation of a single well (such as VSP logging and the like) to establish the corresponding relation between the seismic reflection characteristics and logging curves, and explaining each layer sequence interface with isochronous meaning on the seismic reflection section, wherein the isochronous interfaces at least need to be mid-term reference plane convolution interfaces corresponding to high-resolution layer sequence stratigraphy; and finally, carrying out secondary gyratory division and comparison according to the gyratory superposition mode and formation development characteristics in the work area, interpolating small-layer layers by using the top surface trend constraint of 4 oil groups, and generating a construction grid model by processing the relation between faults and strata in the form of corner grids. In the construction modeling process, the construction grid is fully ensured to reflect the stratum mode, and the longitudinal grids with the same serial number have equal significance.
The small layer division method in reference 1 is difficult to determine trend of inter-well horizons, if few well destination areas are encountered, no effective horizon information exists between wells, and according to the method, accurate isochronous small layers are difficult to establish (the longitudinal resolution scale of the earthquake is far greater than that of the logging, so on the small layer scale, small layer division cannot be performed only by the earthquake reflection characteristics); otherwise, if a multi-well destination area is encountered, the method needs to divide the manual horizon for each well, the time and labor cost is high, the doping of human subjective factors is more, and the dividing precision is affected.
In reference 2, large-scale horizon division (oil layer group) is directly performed based on seismic data by means of logging information, small layers are obtained according to interpolation, the result is obviously inaccurate, and because the distribution trend (fluctuation in depth) of the large layers and the small layers on a geological plane has strong inconsistency due to factors such as stratum lithology, geological structure and the like, even wrong small layer division results can occur, manual correction is needed at the moment, and the working cost is greatly increased.
Therefore, the fine horizon obtained in the traditional modeling process cannot well obtain the small horizon dividing result, or can be completed only by needing a large amount of manpower, material resources and time cost, and the efficiency is low; meanwhile, no good technology has yet been available for obtaining inter-well horizon distribution, namely how to transform each well horizon into the two-dimensional surface of the whole target work area in a point-to-surface manner, the current processing method is generally to divide sand group horizons with larger scale based on well data or seismic data, and then establish an isochronous stratum grid model through interpolation and linear equal division technologies. However, the accuracy and geological compliance of the isochronal stratigraphic framework are based on accurate division of the actual small strata, and the above problems are to be solved.
With respect to the prior art of intelligent horizon partitioning, literature: shang Fuhua, li Jincheng, field, cao Maojun, du Ruishan. Computer technology and development based on improved BP neural network stratigraphic method [ J ]. Computer technology 2020,30 (09): 148-153. Authors perform stratigraphic classification whose essential technology is lithology classification using neural networks based on log curves. And selecting the three-layer BP neural network which is the most basic, and taking lithology categories as the basis of stratum division. The sampling points of the abrupt change of lithology are layering points of the small layers. Thus, identification of lithology is an important point of work. Meanwhile, the author selects an L-M algorithm to improve and optimize the training effect of the neural network; the authors select natural potential curves and natural gamma curves as main curves for lithology division, and refer to the characteristics of resistivity, acoustic time difference, density and neutron curves to divide small layers. The authors select improved BP neural networks, which are a basic neural network of a type which is relatively mature in development at present, but have a plurality of defects: 1. when dealing with some complicated nonlinear problems, the training of the network is easy to fall into local minima, so that the training is failed, and a converged neural network model is not obtained; 2. the convergence speed is low, namely the algorithm efficiency is low; 3. the structural selection of the neural network is not guided by a set of system theory, and is selected by expert experience, so that the generalization capability of the neural network is insufficient for different target work areas, different geological conditions and different horizon divisions; 4. the network has high dependence on samples and high data quality requirement.
Liu Yingjie an intelligent stratum comparison technical method and application [ D ]. Yan Shanda science, 2013. A Gaussian model is utilized to obtain a more accurate layering result through credibility calculation and control of layering interface selection; after lithology intervals are determined, parameters such as natural gamma, natural potential, acoustic time difference, sand layer thickness, rhythm and the like of each interval are extracted according to logging attribute characteristics, and the parameters are taken as characteristic attributes of each small layer, and a Probability Neural Network (PNN) is adopted for carrying out inter-well stratum comparison and connecting. The data preprocessing process of the method is complex: the logging curve rhythm type and the characteristic vector table are summarized through logging attribute feature induction, namely, the parameters of the image are converted into numerical values to be recognized by a computer, and in addition, signals are filtered through filtering treatment of a natural gamma curve (GR), a sound wave time difference curve (AC) and a natural potential curve (SP), so that logging data are square-wave. The establishment of the learning sample refers to a standard well, and the verification of the neural network performance only selects 4 wells, so that the applicability of the neural network can not be verified for an oilfield work area, and the use of too little data set can not provide effective and comprehensive geological information for the subsequent modeling work through horizon division widely and accurately.
Therefore, the existing intelligent horizon dividing method is generally only aimed at single well information, geological information and seismic information around a target well are not considered while horizons are divided, at most, constraint correction is carried out by using the two types of information after a preliminary dividing result is obtained, and the artificial intelligence has some common challenges in the field of geophysical interpretation: data set size, data quality, model convergence speed, final application effect, etc.
Disclosure of Invention
Aiming at the technical problems, the invention provides an intelligent modeling method for an underground multi-information constrained isochronous stratigraphic framework, which takes logging and geological information as main materials, adds geological constraint in a data set manufacturing mode while carrying out small-layer intelligent division based on a logging curve, and enables the geological information to participate in the whole division process so as to obtain a result which accords with the geological information. Based on fine small layer division, the problem of inter-well layer space division is solved by means of seismic inversion information and a neural network, and a high-precision isochronous stratum grid model is established. The whole method can effectively improve the working procedure efficiency through an artificial intelligence technology, and the manufacture of the data set containing geological constraints can break the limitation of the neural network on the size of the data set, the convergence speed of the model and the like.
The specific technical scheme is as follows:
the intelligent modeling method for the underground multi-information constrained isochronous stratum grillage is characterized by comprising the following steps of:
s1, analyzing a target work area according to expert experience, selecting lithology-sensitive logging curves, determining that the type of the curves serving as a neural network data set is C, the number of sample points of each curve is L, determining N wells serving as a neural network training set according to the number of existing logging information, and determining a prediction set sample, namely M blind wells needing to be divided into small layers;
s2, carrying out geological analysis on a work area, and determining an object source direction alpha and a vertical object source direction beta, wherein beta-alpha=90 ° Dividing the entire work zone log into 4 parts numbered (1) - (4) based on α and β;
s3, manufacturing a neural network training data set according to the divided work areas;
the manufacturing flow is as follows: determining that one well connecting sample consists of Q wells, wherein each well has C well logging curves, respectively manufacturing 25% of the number of two sets of training set samples according to the direction of a material source and a vertical material source, and carrying out the rest 50% of data samples in a random combination mode;
s3.1, following the object source direction: numbering N the well training samples of the (1) (3) region i ，i＝1,2,3…,N 1 The number of samples is N 1 The method comprises the steps of carrying out a first treatment on the surface of the Numbering N the well training samples of the (2) (4) region i ，i＝1,2,3…,N 2 The number of samples is N 2 The method comprises the steps of carrying out a first treatment on the surface of the Wherein n=N 1 +N 2 Generating a well sequence number W constituting a sample according to formula (1) k ，k＝1,2…,Q；
I a ＝[N j *random(0～1)+0.5] (1)
[]Representing a round-up, j=1, 2, where N is the random combination of the well samples taken from the (1) (3) region when j=1 j ＝N 1 When j=2, taking (2) (4) area samples for random combination, N j ＝N 2 ；
S3.2, vertical object source direction: numbering N the well training samples of the (1) (2) regions i ，i＝1,2,3…,N 3 The number of samples is N 3 The method comprises the steps of carrying out a first treatment on the surface of the Numbering N the well training samples of the (3) (4) region i ，i＝1,2,3…,N 4 The number of samples is N 4 The method comprises the steps of carrying out a first treatment on the surface of the Where n=n 3 +N 4 Generating a well sequence number W constituting a sample according to formula (2) k ，k＝1,2..,Q；
I v ＝[N j *random(0～1)+0.5] (2)
[]Representing a round-up, j=3, 4, where N is the random combination of the well samples taken from the (1) (2) region when j=3 j ＝N 3 The method comprises the steps of carrying out a first treatment on the surface of the When j=4, taking (3) (4) area samples for random combination, N j ＝N 4 ；
S3.3, random combination: uniformly numbering N for all training samples in work area i I=1, 2,3 …, N; generating a well sequence number W constituting a sample according to formula (3) k ，k＝1,2..,Q；
I r ＝[N*random(0～1)+0.5] (3)
[] Representing the upward rounding, wherein N is the total number of training samples;
and forming a neural network training set according to the three modes.
S4, numbering the predicted samples M j J=1, 2,3 … M, number N for all training samples i I=1, 2,3 …, N, the well number W is generated according to equation (4) k And M is as follows j One sample was composed, k=1, 2., Q-1; namely, the training set takes a predicted sample which is formed by 1 well in the prediction set and 1 well in the Q-1 well;
I p ＝[N*random(0～1)+0.5] (4)
[] Representing the upward rounding, wherein N is the number of training samples;
finally, forming a final neural network data set according to the training set obtained in the step S3 and the generated prediction set;
s5, constructing a feature pyramid network FPN, and performing model training by using the training set in the S4;
s6, carrying out batch small-layer division on a prediction set, namely blind wells, in the target work area by using the model obtained in the S5;
s7, taking the horizon dividing result obtained in the S6 as a boundary in a target work area, carrying out random optimization inversion on a target well section based on seismic data to obtain a wave impedance inversion initial model, taking each sample point in the model as a pseudo well, selecting the pseudo well to manufacture a new sample set for training according to a random combination method in the S3, and predicting pseudo wells with other unknown horizon information based on a convergence model to further obtain a small two-dimensional layer of the whole work area, carrying out error calculation on the horizon according to manual layering of a real logging, and carrying out interpolation on the basis of well point error calculation to obtain a correction result;
s8, repeating the step S7, and continuously optimizing and correcting the intelligent stratum division result by taking the well horizon as a constraint condition to obtain a two-dimensional horizon division result which finally accords with a geological rule;
s9, obtaining a final isochronous stratigraphic grid model based on the result obtained in the S8 as horizon constraint.
The technical scheme of the invention has the beneficial effects that:
1. the data set manufacturing method in the step S2 has the beneficial effects that: the neural network model convergence can be accelerated; meanwhile, the method discards the traditional neural network learning mode of single well information feature extraction, converts one-dimensional logging data into two-dimensional data through a well connection data form, improves the learning efficiency of the FPN network, can better save geological space features between wells, simultaneously adds geological information into a dataset to serve as constraint conditions, penetrates through the whole intelligent small-layer division process, and is beneficial to finally obtaining horizon division results which are more in line with real geological information; under the condition of fewer wells, the problem that the neural network model cannot be trained due to the lack of data can be solved by data augmentation through the production of the continuous well data set, and the working efficiency can be greatly improved through an automatic process under the condition of more wells.
2. The FPN network is selected to carry out the task of small-layer classification, the types of the small layers can be identified and the positions of layering lines can be accurately detected, the network structure and the multi-scale detection function can be used for reserving more communication information of the layers in the geological space, the running efficiency of the model and the detection performance of small-scale targets can be improved, and compared with the method for selecting a common neural network to realize the task of simple layer classification, the method can effectively improve the intelligent small-layer classification precision.
3. The method comprises the steps of obtaining high-precision small-layer division results by utilizing an artificial intelligence method through comprehensive geological information and logging information, obtaining inter-well horizons conforming to geological laws by means of seismic information based on the results, and finally obtaining a high-precision isochronous stratum grid model through the point-to-surface division process of the small layers.
Drawings
FIG. 1 is a plot division result of an embodiment;
FIG. 2 is a well-tie sample structure of an embodiment;
FIG. 3 is a blind well intelligent small-layer division result of an embodiment;
FIG. 4 is a wave impedance inversion initial model of an embodiment;
FIG. 5 illustrates horizon correction according to an embodiment.
Detailed Description
The specific technical scheme of the invention is described by combining the embodiments.
An intelligent modeling method for an underground multi-information constrained isochronous stratum grid comprises the following steps:
s1, analyzing a target work area according to expert experience, selecting a logging curve sensitive to lithology such as sand shale, determining that the type of the curve serving as a neural network data set is C, the number of sample points of each curve is L, determining N wells serving as a neural network training set according to the number of existing logging information, and determining that a prediction set sample is the M blind wells needing to be divided into small layers;
s2, pairingGeological analysis is carried out on the work area, and the object source direction alpha and the vertical object source direction beta, beta-alpha=90 are determined ° Dividing the whole work area well logging into 4 parts numbered (1) - (4) based on alpha and beta, wherein the solid line is a cutting boundary line, and the direction indicated by an arrow is the direction of an object source as shown in fig. 1;
s3, according to the above-mentioned division work area, carry on the training dataset of neural network to make, its preparation flow is: it is determined that one well connecting sample is composed of Q wells, and C log curves (in step S1) are determined for each well, then one sample is as shown in fig. 2, 25% of the number of two sets of training set samples are respectively manufactured according to the direction of the object source and the direction of the vertical object source, and the rest 50% of the data samples are performed by means of random combination.
S3.1, following the object source direction: for the well training samples in the region (1) (3) (number of samples is N 1 ) Numbering N is carried out i (i＝1,2,3…,N 1 ) For the well training samples in the (2) (4) region (number of samples N 2 ) Numbering N is carried out i (i＝1,2,3…,N 2 ) Where n=n 1 +N 2 Generating a well sequence number W constituting a sample according to formula (1) k (k＝1,2…,Q)；
I a ＝[N j *random(0～1)+0.5] (1)
[]Representing a round-up, j=1, 2, where N is the random combination of the well samples taken from the (1) (3) region when j=1 j ＝N 1 When j=2, taking (2) (4) area samples for random combination, N j ＝N 2 ；
S3.2, vertical object source direction: for the well training samples in the region (1) (2) (number of samples is N 3 ) Numbering N is carried out i (i＝1,2,3…,N 3 ) For the well training samples in the region (3) (4) (number of samples is N 4 ) Numbering N is carried out i (i＝1,2,3…,N 4 ) Where n=n 3 +N 4 Generating a well sequence number W constituting a sample according to formula (2) k (k＝1,2..,Q)；
I v ＝[N j *random(0～1)+0.5] (2)
[]Representing a round-up, j=3, 4, where j=3, taking the well samples of the (1) (2) region for random combinationTime N j ＝N 3 When j=4, taking (3) (4) area samples for random combination, N j ＝N 4 ；
S3.3, random combination: uniformly numbering N for all training samples in work area i (i=1, 2,3 …, N) generating a well number W constituting one sample according to formula (3) k (k＝1,2..,Q)；
I r ＝[N*random(0～1)+0.5] (3)
[] Representing the upward rounding, wherein N is the total number of training samples;
and forming a neural network training set according to the three modes.
S4, numbering the predicted samples M j (j=1, 2,3 … M), number N for all training samples i (i=1, 2,3 …, N), generating a well number W according to equation (4) k (k=1, 2., Q-1) and M j Forming a sample (namely, taking a predicted sample from a training set, wherein the predicted sample is formed by a total of 1 well and 1 well in a predicted set;
I p ＝[N*random(0～1)+0.5] (4)
[] Representing the upward rounding, wherein N is the number of training samples;
finally, forming a final neural network data set according to the training set obtained in the step S3 and the generated prediction set;
s5, constructing a Feature Pyramid Network (FPN), and performing model training by using the training set in the S4;
s6, carrying out batch small-layer division on a prediction set, namely blind wells, in the target work area by using the model obtained in the S5, as shown in FIG. 3;
s7, taking the horizon dividing result obtained in the S6 as a boundary in a target work area, carrying out random optimization inversion on the target well section based on seismic data to obtain a wave impedance inversion initial model, as shown in FIG. 4, taking each sample point in the model as a pseudo well, selecting the pseudo well to manufacture a new sample set according to a random combination method in the S3 for training, and predicting the pseudo wells with other unknown horizon information based on a convergence model, so as to obtain a small two-dimensional layer of the whole work area, carrying out error calculation on the horizon according to manual layering of a real well logging, and carrying out interpolation on the basis of well point error calculation to obtain a correction result, as shown in FIG. 5;
s8, repeating the step S7, and continuously optimizing and correcting the intelligent stratum division result by taking the well horizon as a constraint condition to obtain a two-dimensional horizon division result which finally accords with a geological rule;
s9, obtaining a final isochronous stratigraphic grid model based on the result obtained in the S8 as horizon constraint.
Claims (3)
1. The intelligent modeling method for the underground multi-information constrained isochronous stratum grillage is characterized by comprising the following steps of:
s1, analyzing a target work area according to expert experience, selecting lithology-sensitive logging curves, determining that the type of the curves serving as a neural network data set is C, the number of sample points of each curve is L, determining N wells serving as a neural network training set according to the number of existing logging information, and determining a prediction set sample, namely M blind wells needing to be divided into small layers;
s2, carrying out geological analysis on the work area, determining an object source direction alpha and a vertical object source direction beta, wherein beta-alpha=90 DEG, and dividing the whole work area well logging into 4 parts numbered (1) - (4) based on alpha and beta;
s3, manufacturing a neural network training data set according to the divided work areas;
s4, numbering predicted samples, and numbering all training samples to generate a sample predicted sample; forming a final neural network data set according to the training set obtained in the step S3 and the generated prediction set;
s5, constructing a feature pyramid network FPN, and performing model training by using the training set in the S4;
s6, carrying out batch small-layer division on a prediction set, namely blind wells, in the target work area by using the model obtained in the S5;
s7, taking the horizon dividing result obtained in the S6 as a boundary in a target work area, carrying out random optimization inversion on a target well section based on seismic data to obtain a wave impedance inversion initial model, taking each sample point in the model as a pseudo well, selecting the pseudo well to manufacture a new sample set for training according to a random combination method in the S3, and predicting pseudo wells with other unknown horizon information based on a convergence model to further obtain a small two-dimensional layer of the whole work area, carrying out error calculation on the horizon according to manual layering of a real logging, and carrying out interpolation on the basis of well point error calculation to obtain a correction result;
s8, repeating the step S7, and continuously optimizing and correcting the intelligent stratum division result by taking the well horizon as a constraint condition to obtain a two-dimensional horizon division result which finally accords with a geological rule;
s9, obtaining a final isochronous stratigraphic grid model based on the result obtained in the S8 as horizon constraint.
2. The intelligent modeling method for the underground multi-information constrained isochronous stratigraphic framework according to claim 1, wherein the neural network training data set is created in the step S3;
the manufacturing flow is as follows: determining that one well connecting sample consists of Q wells, wherein each well has C well logging curves, respectively manufacturing 25% of the number of two sets of training set samples according to the direction of a material source and a vertical material source, and carrying out the rest 50% of data samples in a random combination mode;
s3.1, following the object source direction: numbering N the well training samples of the (1) (3) region i ，i＝1,2,3…,N 1 The number of samples is N 1 The method comprises the steps of carrying out a first treatment on the surface of the Numbering N the well training samples of the (2) (4) region i ，i＝1,2,3…,N 2 The number of samples is N 2 The method comprises the steps of carrying out a first treatment on the surface of the Where n=n 1 +N 2 Generating a well sequence number W constituting a sample according to formula (1) k ，k＝1,2…,Q；
I a ＝[N j *random(0～1)+0.5] (1)
[]Representing a round-up, j=1, 2, where N is the random combination of the well samples taken from the (1) (3) region when j=1 j ＝N 1 When j=2, taking (2) (4) area samples for random combination, N j ＝N 2 ；
S3.2, vertical object source direction: numbering N the well training samples of the (1) (2) regions i ，i＝1,2,3…,N 3 The number of samples is N 3 The method comprises the steps of carrying out a first treatment on the surface of the Numbering N the well training samples of the (3) (4) region i ，i＝1,2,3…,N 4 The number of samples is N 4 The method comprises the steps of carrying out a first treatment on the surface of the Where n=n 3 +N 4 Generating a well sequence number W constituting a sample according to formula (2) k ，k＝1,2..,Q；
I v ＝[N j *random(0～1)+0.5] (2)
[]Representing a round-up, j=3, 4, where N is the random combination of the well samples taken from the (1) (2) region when j=3 j ＝N 3 The method comprises the steps of carrying out a first treatment on the surface of the When j=4, taking (3) (4) area samples for random combination, N j ＝N 4 ；
S3.3, random combination: uniformly numbering N for all training samples in work area i I=1, 2,3 …, N; generating a well sequence number W constituting a sample according to formula (3) k ，k＝1,2..,Q；
I r ＝[N*random(0～1)+0.5] (3)
[] Representing the upward rounding, wherein N is the total number of training samples;
and forming a neural network training set according to the three modes.
3. The method for intelligent modeling of an underground multi-information constrained isochronous stratigraphic framework according to claim 1, wherein S4 specifically comprises: numbering predicted samples M j J=1, 2,3 … M, number N for all training samples i I=1, 2,3 …, N, the well number W is generated according to equation (4) k And M is as follows j One sample was composed, k=1, 2., Q-1; namely, the training set takes a predicted sample which is formed by 1 well in the prediction set and 1 well in the Q-1 well;
I p ＝[N*random(0～1)+0.5] (4)
[] Represents the rounding up, N is the number of training samples.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111302831.0A CN113985493B (en) | 2021-11-05 | 2021-11-05 | Intelligent modeling method for underground multi-information constrained isochronous stratum grillwork |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111302831.0A CN113985493B (en) | 2021-11-05 | 2021-11-05 | Intelligent modeling method for underground multi-information constrained isochronous stratum grillwork |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN113985493A CN113985493A (en) | 2022-01-28 |
| CN113985493B true CN113985493B (en) | 2023-11-03 |
Family
ID=79746579
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202111302831.0A Active CN113985493B (en) | 2021-11-05 | 2021-11-05 | Intelligent modeling method for underground multi-information constrained isochronous stratum grillwork |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113985493B (en) |
Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104991274A (en) * | 2015-07-03 | 2015-10-21 | 中国石油大学(华东) | Single-trap level favorable region optimal selection method under multi-geological factor quantitative constraints |
| CN105182444A (en) * | 2015-09-09 | 2015-12-23 | 中国海洋石油总公司 | High resolution sequence stratigraphic framework constraint geostatistical inversion method |
| CN105373648A (en) * | 2015-10-09 | 2016-03-02 | 西南石油大学 | Modeling method for sand shale interbed type sedimentary body reservoir architecture |
| CN109388817A (en) * | 2017-08-04 | 2019-02-26 | 中国石油化工股份有限公司 | A kind of Reservoir Fracture three-dimensional modeling method |
| CN110838175A (en) * | 2019-11-07 | 2020-02-25 | 成都北方石油勘探开发技术有限公司 | Geological model building method for gas injection oil reservoir development |
| CN111596364A (en) * | 2020-05-20 | 2020-08-28 | 中国海洋石油集团有限公司 | Seismic sedimentation microphase combination analysis method based on high-precision sequence stratigraphic framework |
| CN112505753A (en) * | 2020-10-23 | 2021-03-16 | 中国石油集团工程咨询有限责任公司 | Seismic-driven high-precision sequence grid model construction method |
| CN113313825A (en) * | 2021-05-17 | 2021-08-27 | 中国石油天然气集团有限公司 | Three-dimensional geological modeling method and device for oil reservoir |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20160349389A1 (en) * | 2015-05-29 | 2016-12-01 | Cgg Services Sa | Method for developing a geomechanical model based on seismic data, well logs and sem analysis of horizontal and vertical drill cuttings |
-
2021
- 2021-11-05 CN CN202111302831.0A patent/CN113985493B/en active Active
Patent Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104991274A (en) * | 2015-07-03 | 2015-10-21 | 中国石油大学(华东) | Single-trap level favorable region optimal selection method under multi-geological factor quantitative constraints |
| CN105182444A (en) * | 2015-09-09 | 2015-12-23 | 中国海洋石油总公司 | High resolution sequence stratigraphic framework constraint geostatistical inversion method |
| CN105373648A (en) * | 2015-10-09 | 2016-03-02 | 西南石油大学 | Modeling method for sand shale interbed type sedimentary body reservoir architecture |
| CN109388817A (en) * | 2017-08-04 | 2019-02-26 | 中国石油化工股份有限公司 | A kind of Reservoir Fracture three-dimensional modeling method |
| CN110838175A (en) * | 2019-11-07 | 2020-02-25 | 成都北方石油勘探开发技术有限公司 | Geological model building method for gas injection oil reservoir development |
| CN111596364A (en) * | 2020-05-20 | 2020-08-28 | 中国海洋石油集团有限公司 | Seismic sedimentation microphase combination analysis method based on high-precision sequence stratigraphic framework |
| CN112505753A (en) * | 2020-10-23 | 2021-03-16 | 中国石油集团工程咨询有限责任公司 | Seismic-driven high-precision sequence grid model construction method |
| CN113313825A (en) * | 2021-05-17 | 2021-08-27 | 中国石油天然气集团有限公司 | Three-dimensional geological modeling method and device for oil reservoir |
Non-Patent Citations (2)
| Title |
|---|
| 井震联合分级控制小层对比方法;刘伟方;于兴河;孙勤华;滕团余;潘建国;;石油物探(06);全文 * |
| 高分辨率层序格架约束的地震反演及应用;曹彤;郭少斌;;科技导报(09);全文 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN113985493A (en) | 2022-01-28 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108957549B (en) | Braided river sediment heterogeneous compact sandstone gas reservoir geological modeling method | |
| US11016214B2 (en) | Dolomite reservoir prediction method and system based on well and seismic combination, and storage medium | |
| CN106405651B (en) | Full waveform inversion initial velocity model construction method based on logging matching | |
| CN107783187B (en) | Method for establishing three-dimensional velocity field by combining logging velocity and seismic velocity | |
| CN109884710B (en) | Micro-logging tomography method aiming at excitation well depth design | |
| CN111983691B (en) | Reservoir prediction method and software system with multi-model fusion | |
| CN102176052B (en) | Hierarchical sequence analysis method oriented to generation of three-dimensional hierarchical grids | |
| CN102901985B (en) | A kind of Depth Domain interval velocity modification method being applicable to relief surface | |
| CN113642698A (en) | Geophysical logging intelligent interpretation method, system and storage medium | |
| CN111983683B (en) | Prediction method and system for lake-facies limestone reservoir under low-well condition | |
| CN116305850B (en) | Stratum thermal conductivity prediction method based on random forest model | |
| CN106569272A (en) | Earthquake attribute fusion method based on data property space ascending dimension | |
| CN110428497B (en) | Braided river training image generation method | |
| CN109975872B (en) | Method for directly compiling balance profile by using Corel DRAW software | |
| CN113985493B (en) | Intelligent modeling method for underground multi-information constrained isochronous stratum grillwork | |
| CN111290043A (en) | Calculation method of sedimentary stratum pseudo-acoustic logging data | |
| CN114442170A (en) | True earth surface velocity fusion modeling method for double complex regions | |
| CN111273344A (en) | Chromatographic inversion method based on continuous-to-refracted wave and processing terminal | |
| CN117236172B (en) | Deep learning low-frequency model construction method and system for igneous rock | |
| CN115469361B (en) | Clastic rock stratum three-dimensional geological modeling method | |
| Correia | Integration of reservoir characterization with history matching guided by pilot wells: Application to the Norne Field | |
| CN116699678A (en) | Speed modeling method and device for multi-information fusion | |
| CN117270043A (en) | Method for identifying beach dam sand | |
| CN115963569A (en) | Method and device for establishing phase-control initial model in phase-control inversion | |
| CN118068426A (en) | RBF inversion method under special reservoir layer sequence constraint |
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 | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |