CN-121999150-A - Three-dimensional stratum modeling method combining bilinear interpolation and multipoint statistical simulation

CN121999150ACN 121999150 ACN121999150 ACN 121999150ACN-121999150-A

Abstract

The application belongs to the field of three-dimensional geologic modeling, and particularly discloses a three-dimensional stratum modeling method combining bilinear interpolation and multipoint statistical simulation, which comprises the steps of obtaining two geologic sections A and B which are parallel and discretized according to geologic attributes; according to a plurality of preset different positions, generating a plurality of middle sections through bilinear interpolation based on A and B, dividing each generated middle section into a conditional data part and a priori probability part, executing multipoint statistic simulation on each middle section based on the conditional data part, the priori probability part and the sections A and B serving as training images, fusing a multipoint statistic simulation result after probability distribution with the priori probability part through a Bayesian probability fusion mechanism to obtain the final geological attribute of each position in the middle section, and synthesizing all the obtained middle sections into a three-dimensional geological model. The application can give consideration to the calculation efficiency and the real depicting ability of complex geological structures in geological modeling.

Inventors

CHEN QIYU
FANG HONGFENG
LIU GANG
CHEN DAJIA
Cui Zhesi
ZHANG CE

Assignees

中国地质大学（武汉）

Dates

Publication Date: 20260508
Application Date: 20251230

Claims (10)

1. A three-dimensional stratum modeling method combining bilinear interpolation and multipoint statistical simulation is characterized by comprising the following steps: S10, acquiring two parallel geological section data discretized according to geological attributes, and respectively marking the two parallel geological section data as a section A and a section B; S20, sequentially generating a plurality of middle sections between the section A and the section B through bilinear interpolation based on the section A and the section B according to a plurality of preset different positions; s30, dividing each generated middle section into a conditional data part and a prior probability part, wherein the geological attribute of the conditional data part belongs to a known geological attribute set, and the geological attribute of the prior probability part does not belong to the known geological attribute set; S40, for each middle section, performing multi-point statistical simulation based on a conditional data part, an priori probability part and the sections A and B serving as training images, wherein the multi-point statistical simulation comprises the steps of distributing probability for simulation results obtained by matching from the training images according to the space distance from the current middle section to the sections A and B, and fusing the multi-point statistical simulation results distributed with the probability with the priori probability part through a Bayesian probability fusion mechanism to obtain the final geological attribute of each position in the middle section; S50, synthesizing all the intermediate sections processed in the S40 into a three-dimensional geological model.
2. The method for modeling a three-dimensional formation combining bilinear interpolation and multipoint statistical simulation according to claim 1, wherein step S20 is specifically: For each pixel P (x, y, z) on the ith middle section C i to be generated, firstly, carrying out linear interpolation along the horizontal direction, and obtaining middle results f (R 1 ) and f (R 2 ) of two positions of R 1 and R 2 on the middle section C i according to attribute values of four corner points Q 11 ,Q 21 ,Q 12 ,Q 22 on the section A and the section B; Then, linear interpolation is carried out along the vertical direction of the section, and a final interpolation result f (P) at a target point P (x, y, z) of the middle section C i is calculated according to the intermediate results f (R 1 ) and f (R 2 ).
3. The method for modeling a three-dimensional stratum combining bilinear interpolation and multipoint statistical simulation according to claim 2, wherein the calculation formula of the horizontal direction linear interpolation calculation f (R 1 ) and f (R 2 ) is: the calculation formula of the vertical linear interpolation calculation f (P) is as follows: Wherein, the 、、 Geological attribute values at corner points Q 11 ,Q 21 ,Q 12 ,Q 22 respectively; respectively representing indexes of the cross section A, B in the x-axis direction of the three-dimensional space; = Representing indexes of the same plane where the four corner points Q 11 ,Q 21 ,Q 12 ,Q 22 are located in the y-axis direction of the three-dimensional space; And Respectively, the indexes of two position points R 1 and R 2 on the intermediate section C i in the z-axis direction of the three-dimensional space, and the three-dimensional space coordinate index of each pixel P (x, y, z) on the ith intermediate section C i to be generated.
4. The method for modeling a three-dimensional formation combining bilinear interpolation and multipoint statistical simulation according to claim 1, wherein step S30 is specifically: For a complete middle section C i , it is divided into two parts, wherein the part of the geological attribute value contained in the original set of geological attributes is regarded as a determined conditional data set Cond, the part of the geological attribute value not contained in the original set of geological attributes is regarded as a priori probability part, and the probability distribution of each point is obtained, and the areas are regarded as areas to be multi-point simulated.
5. The method for modeling a three-dimensional formation in combination with bilinear interpolation and multipoint statistical simulation according to claim 1, wherein in step S40, the step of performing the multipoint statistical simulation is specifically: Rasterizing the section A, the section B and the current middle section C i in a consistent size, and distributing a condition data part Cond of the current middle section C i to a grid SG to be simulated; Setting direct sampling multipoint statistic simulation parameters including the sizes (a, b, c) of the multipoint data templates T and a mode distance threshold T; defining a random simulation Path for a region to be simulated in the grid SG, and covering all nodes to be simulated on the current grid SG; The current simulation position P is obtained by traversing the simulation Path in sequence, and pattern matching is carried out on the training image section A and the training image section B by utilizing the data template T according to the existing conditional data on the grid SG until the simulation results r 1 and r 2 with the pattern distance smaller than the threshold value T are successfully matched.
6. The method for modeling a three-dimensional formation combining bilinear interpolation and multipoint statistical simulation according to claim 5, wherein the step of assigning probabilities to simulation results obtained from matching in training images comprises: According to the inverse proportion of the distance from the current middle section C i to the two known sections A and B, probability is distributed to simulation results r 1 and r 2 , and two multipoint simulation probability distributions P (r 1 )、P(r 2 ) are obtained; The prior probability P 0 (j, k) obtained by interpolation of the current point before is obtained, wherein j is the index of the current point, k represents possible attribute values, and P 0 represents the value probability of each value.
7. The method for modeling a three-dimensional formation in combination with bilinear interpolation and multipoint statistical modeling of claim 6, wherein the calculation formulas of P (r 1 ) and P (r 2 ) are: in the formula, And Representing the separation of the intermediate section from the known sections a and B; and the attenuation coefficient representing the spatial correlation is used for controlling the influence of the distance between the generated profile and the training profile on the value probability.
8. The method for modeling a three-dimensional formation combining bilinear interpolation and multipoint statistical simulation according to claim 6, wherein the prior probability P 0 (j, k) has a calculation formula as follows: wherein K is the number of all possible attribute categories on the geological section; The attribute value of the j-th pixel; Is a typical value for the kth category.
9. The method for modeling a three-dimensional stratum by combining bilinear interpolation and multipoint statistical simulation according to claim 1, wherein a calculation formula of the bayesian probability fusion is as follows: in the formula, The probability distribution after fusion; The simulation probability distribution is obtained by utilizing multipoint statistical simulation; and K is the number of all possible attribute categories on the geological section.
10. The application of the three-dimensional stratum modeling method according to any one of claims 1 to 9, which is applied to the fields of urban geological engineering, water resource development and protection, oil reservoirs and mine geology.

Description

Three-dimensional stratum modeling method combining bilinear interpolation and multipoint statistical simulation Technical Field The application belongs to the field of three-dimensional geological modeling, and particularly relates to a three-dimensional stratum modeling method combining bilinear interpolation and multipoint statistical simulation. Background The three-dimensional geological modeling aims at integrating discrete and multi-source geological data by utilizing a computer technology, constructing a digital model capable of quantitatively representing underground space structure and attribute distribution, and providing powerful support for deep understanding of geological rules and scientific analysis and decision making. The fourth stratum is closely related to human activities, and modeling work is one of the cores of the related engineering field. However, the fourth-series deposit has the remarkable characteristics of strong heterogeneity, complex cause and severe phase transition, so that the spatial distribution of the fourth-series deposit is extremely irregular, and modeling faces a plurality of challenges. On one hand, the sparsity of drilling and profile data makes the complex and changeable geological interface difficult to accurately capture, and on the other hand, the traditional spatial interpolation method based on two-point geostatistics is difficult to accurately describe complex geological phenomena such as faults, pinch-outs and the like. For this reason, the prior art has introduced a multi-point geostatistical method capable of learning complex spatial structures in "training images". While this approach has advantages in characterizing complex geologic structures, it is generally assumed that geologic structures are statistically stable, which is not applicable in areas with significant spatial trends or causal zones, such as sedimentary formations, faults, and the like. In addition, the pixel-based pattern searching and matching process is huge in calculation amount, and when a three-dimensional high-resolution model is processed, calculation efficiency and storage overhead become key bottlenecks. Therefore, how to combine the calculation efficiency with the real depicting ability of the complex geological structure in the geological modeling is a current urgent problem to be solved. Disclosure of Invention Aiming at the defects of the prior art, the application aims to provide a three-dimensional stratum modeling method combining bilinear interpolation and multipoint statistical simulation, which can give consideration to both calculation efficiency and real depicting ability of complex geological structures in geological modeling. To achieve the above object, in a first aspect, the present application provides a three-dimensional formation modeling method combining bilinear interpolation and multipoint statistical simulation, comprising the steps of: S10, acquiring two parallel geological section data discretized according to geological attributes, and respectively marking the two parallel geological section data as a section A and a section B; S20, sequentially generating a plurality of middle sections between the section A and the section B through bilinear interpolation based on the section A and the section B according to a plurality of preset different positions; s30, dividing each generated middle section into a conditional data part and a prior probability part, wherein the geological attribute of the conditional data part belongs to a known geological attribute set, and the geological attribute of the prior probability part does not belong to the known geological attribute set; S40, for each middle section, performing multi-point statistical simulation based on a conditional data part, an priori probability part and the sections A and B serving as training images, wherein the multi-point statistical simulation comprises the steps of distributing probability for simulation results obtained by matching from the training images according to the space distance from the current middle section to the sections A and B, and fusing the multi-point statistical simulation results distributed with the probability with the priori probability part through a Bayesian probability fusion mechanism to obtain the final geological attribute of each position in the middle section; S50, synthesizing all the intermediate sections processed in the S40 into a three-dimensional geological model. The three-dimensional stratum modeling method combining bilinear interpolation and multipoint statistical simulation has the following effects that by fusing bilinear interpolation and multipoint statistical simulation, the method bridges the gap between the calculation efficiency and the geological reality. Specifically, a basic model frame of the stratum is quickly constructed by using a bilinear interpolation technology with low calculation cost and high speed, and efficiency advantages are p