CN-121997054-A - Interpretation method for stratum horizon non-integration

CN121997054ACN 121997054 ACN121997054 ACN 121997054ACN-121997054-A

Abstract

The invention discloses an interpretation method of stratum layer unconformity, which relates to the technical field of geological detection, and comprises the steps of constructing a geological data set, drawing a geological profile, identifying a potential unconformity interface according to the geological profile, screening out a plurality of sampling points with similarity between upper and lower rock layers at the unconformity interface lower than expected, constructing distribution coefficients, positioning a risk area on the unconformity interface according to the distribution states of the distribution coefficients, training to obtain an area structure evolution model, generating a relation between geological structure activities and stratum unconformity by using the area structure evolution model as a geological mapping relation to output, constructing a stratum interpretation knowledge graph in advance, and matching interpretation texts from the stratum interpretation knowledge graph according to the correspondence between the geological mapping relation and interpretation contents of stratum unconformity. With the assistance of the knowledge graph, time consumption of comprehensive explanation can be saved, the efficiency of explanation is improved, and the explanation content is more reliable.

Inventors

YU HAITAO
WANG XINYUE
ZHANG ZHAOYU
LIU XINGYU
HE XULIN
WU HONGJUN

Assignees

东北石油大学

Dates

Publication Date: 20260508
Application Date: 20240122

Claims (10)

1. A method for explaining the non-integration of stratum layers is characterized by comprising the following steps, Selecting a plurality of observation points in a target area, collecting geological data at the observation points, constructing a geological data set, drawing a geological profile by the geological data set, identifying a potential non-integrated interface according to the geological profile, and verifying the potential non-integrated interface by constructing a variation coefficient Fpp; Screening a plurality of sampling points with similarity between upper and lower rock layers at the non-integrated interface lower than expected, constructing a distribution coefficient Cob by using the age detection data at the sampling points, and positioning a risk area on the non-integrated interface according to the distribution state of the plurality of distribution coefficients Cob; using geological data in the risk area as sample data, training to obtain an area structure evolution model, generating a relation between geological structure activities and stratum non-integration by using the area structure evolution model, and outputting the relation as a geological mapping relation; and (3) pre-constructing a stratum interpretation knowledge graph, and matching an interpretation text from the stratum interpretation knowledge graph according to the correspondence between the geological mapping relation of stratum non-integration and the interpretation content, wherein the interpretation text is used as the interpretation of stratum layer non-integration.
2. The method of claim 1, wherein the interpretation of formation horizon non-integration is: And determining a target area from the potential non-integrated area, constructing an electronic map covering the target area, setting a plurality of observation points on the electronic map, sampling geological data of the target area at the observation points, and summarizing geological data obtained by sampling to construct a geological data set.
3. A method of interpretation of stratigraphic horizon non-integration according to claim 2 wherein: drawing a geological profile from the geological data set, identifying a potential non-integrated interface through the geological profile, and collecting rock samples on the upper side and the lower side of the non-integrated interface; And constructing a variation coefficient Fpp from the variation data set, determining a potential non-integration area as a non-integration area if the variation coefficient Fpp exceeds a variation threshold, marking the non-integration area on the electronic map, and sending a detection instruction to the outside.
4. A method of interpretation of stratigraphic horizon non-integration according to claim 3 wherein: The variation coefficient Fpp is obtained by performing linear normalization processing on the isotope difference Hp and the texture similarity Cp, mapping corresponding data values into intervals [0,1], and then according to the following formula: wherein i is the serial number of sampling points, i=1, 2, & gt, n is a positive integer greater than 1, and is the number of the sampling points, the weight coefficient is that F 1 ≤1,0≤F 2 is more than or equal to 0 and less than or equal to 1, and F 2 +F 1 =1, and the weight coefficient is that F 1 ≤1,0≤F 2 is more than or equal to 1 and F 2 +F 1 =1 Is a qualified standard value of isotope difference, Hp avg is the mean value of isotope differences, and Cp avg is the mean value of texture similarity.
5. The method of claim 4, wherein the interpretation of formation horizon non-integration is: After receiving the detection instruction, selecting sampling points in the non-integration area, wherein the density and the variation coefficient Fpp of the sampling points accord with: wherein n is the number of sampling points in the non-integrated area, S ij is the distance from the data sampling point i to the data sampling point j, and r is the maximum diameter in the non-integrated area.
6. A method of interpretation of stratigraphic horizon non-integration according to claim 3 wherein: Collecting data at sampling points of the non-integrated area, constructing a data set of the non-integrated area after summarizing, calculating the similarity of lithology of the upper strata and the lower strata of the non-integrated interface, sequencing a plurality of acquired similarities, and acquiring a similarity sequence; And constructing a distribution coefficient Cob by using the age difference Xs, determining a corresponding sampling point as a target sampling point if the distribution coefficient Cob exceeds a distribution threshold value, connecting a plurality of target sampling points, and then taking an surrounding area as a risk area to send an analysis instruction to the outside.
7. The method of claim 6, wherein the interpretation of formation horizon non-integration is: The age difference Xs is subjected to linear normalization processing, corresponding data values are mapped into intervals [0,1], and a distribution coefficient Cob is obtained according to the following mode: Where i=1, 2,..n, n is the number of sampling points, xv avg is the mean of the time differences, and Xs i is the time difference between the upper and lower layers at the i-th sampling point.
8. The method of claim 6, wherein the interpretation of formation horizon non-integration is: After receiving analysis instructions, collecting geological data in a risk area, collecting the above data from observation, analysis and exploration, identifying characteristic data, constructing a geological characteristic data set, extracting part of sample data in the geological characteristic data set, training an initial model constructed by a Bp neural network by using the sample data, taking the trained initial model as an area structure evolution model, outputting the relation between geological structure activities and stratum incompetence by using the trained area structure evolution model as a geological mapping relation, and outputting a plurality of groups of geological mapping relations.
9. The method of claim 1, wherein the interpretation of formation horizon non-integration is: According to training, obtaining a network representation learning model, taking stratum unconformity and related words thereof as search words, adopting depth-first search to search and collect data from public channels, summarizing the data to generate a map data set, and selecting data from the map data set by using the trained semantic network model based on machine learning to perform fusion processing to complete construction of stratum interpretation knowledge maps.
10. The method of claim 9, wherein the interpretation of formation horizon non-integration is performed by: according to the correspondence between the geological mapping relation of the unconformity of the stratum and the explanatory content, a plurality of explanatory texts are matched from the stratum explanatory knowledge graph by using the trained matching model and are used as an explanatory method for the unconformity of the stratum layer in the unconformity area, and the number of the explanatory methods is positively correlated with the number of the unconformity areas.

Description

Interpretation method for stratum horizon non-integration Technical Field The invention relates to the technical field of geological detection, in particular to an interpretation method for stratum horizon non-integration. Background Formation horizon disaggregation is a concept in geology that refers to the occurrence of discontinuities or deletions between otherwise continuously deposited formations in a sequence of formations due to the occurrence of certain geological events historically. These geological events may include erosion, lifting, sinking, or other structural activities, etc. The lack of integration represents a time gap in the geological history, which may be short-term or long-term events lasting thousands of years. Formation unconformities are generally classified as unconformities, angular unconformities, parallel unconformities, or suture unconformities. Different types of unconformities reflect historically complex structural activities and environmental transitions of the earth, and are important clues for studying the earth's history and geologic structure. By analyzing the unconformity, geologist can reconstruct past geological events and environmental changes, read the evolution history of sedimentary basins, predict the distribution situation of mineral resources and the like. In the Chinese patent of the invention with the publication number of CN107269269B, a rapid interpretation method of an unconformity stratum layer is disclosed, and comprises the steps of 1, dividing a single well fine sand layer group of a drilled well, 2, carrying out fine synthetic record calibration, 3, obtaining a time gradient of standard layer control, 4, calculating fitting time difference of a time-varying window algorithm, and 5, carrying out automatic interpretation of the unconformity stratum. The rapid interpretation method for the unconformity stratum layer automatically tracks the unconformity stratum by utilizing the time-varying window calculation fit controlled by the mark layer, and the result of the unconformity stratum interpreted by the method is more objective, accurate and effective, thereby improving the working efficiency and reducing the interpretation error. However, the technical solutions described in the above applications have the following drawbacks: After the potential formation unconformity area has been confirmed, due to randomness of sampling point selection, if the unconformity of the sampling area actually exists and is unevenly distributed, but the sampled geological data comes from a slight area, when the geological data of the slight area is used as sample data to generate explanatory content, the correspondence between the explanatory content and the actually existing horizon unconformity is low, so that the reliability of interpretation of actual output is insufficient, and geological risks brought by the actual formation horizon unconformity may be reduced. To this end, the present invention provides an interpretation method of formation horizon non-integration. Disclosure of Invention (One) solving the technical problems Aiming at the defects of the prior art, the invention provides an interpretation method of stratum horizon non-integration, which comprises the steps of constructing a geological data set, drawing a geological profile, identifying a potential non-integration interface according to the geological profile, screening out a plurality of sampling points with similarity between upper and lower rock layers at the non-integration interface lower than expected, constructing distribution coefficients, positioning a risk area on the non-integration interface according to the distribution state of the plurality of distribution coefficients, training to obtain an area structure evolution model, generating a relation between geological structure activities and stratum non-integration by using the area structure evolution model as a geological mapping relation to output, constructing a stratum interpretation knowledge graph in advance, and matching an interpretation text from the stratum interpretation knowledge graph according to the correspondence between the geological mapping relation and interpretation content of stratum non-integration. With the assistance of the knowledge graph, the time consumption of comprehensive explanation can be saved, and the efficiency of explanation is improved, so that the problem in the background technology is solved. (II) technical scheme The interpretation method of stratum horizon non-integration comprises selecting a plurality of observation points in a target area, collecting geological data at the observation points, constructing a geological data set, drawing a geological profile by the geological data set, identifying a potential non-integration interface according to the geological profile, verifying the potential non-integration interface by constructing a variation coefficient Fpp, and sending out a detection instru