CN-121995448-A - Ultra-deep walk-slip break-control grid-shaped reservoir geophysical characterization method, device, electronic equipment and medium

CN121995448ACN 121995448 ACN121995448 ACN 121995448ACN-121995448-A

Abstract

The application discloses a geophysical characterization method, a device, electronic equipment and a medium for an ultra-deep walk sliding breaking control grid-shaped reservoir. The method comprises the steps of establishing a forward model based on the grid-shaped reservoir body, carrying out forward simulation analysis on the grid-shaped reservoir body aiming at the forward model, constructing a deep learning sample library of the grid-shaped reservoir body according to the result of the forward simulation analysis, establishing a U-net network model, training, and carrying out grid-shaped reservoir body representation through the trained network model. Based on a conventional seismic attribute initial sample library, the application constructs an adaptive deep learning sample library for artificial intervention of a reservoir body by manually explaining the result, innovatively researching and developing structural gradient attribute, instantaneous amplitude ratio and similarity transverse change rate attribute, and applies deep learning prediction and characterization of a fault control reservoir body.

Inventors

DING SHENG
YIN TIANHAO
GAO RUIYU
LI FEI

Assignees

中国石油化工股份有限公司
中石化石油物探技术研究院有限公司

Dates

Publication Date: 20260508
Application Date: 20241105

Claims (10)

1. An ultra-deep walk-slip break-control grid-like reservoir geophysical characterization method, comprising: establishing a forward model based on a grid-shaped accumulator; performing grid-shaped accumulator forward modeling analysis on the forward model; Constructing a deep learning sample library of the grid-shaped reservoir body according to the result of forward modeling analysis; and establishing a U-net network model, training, and carrying out grid-shaped reservoir body characterization through the trained network model.
2. The ultra-deep walk sliding off-control grid-like reservoir geophysical characterization method of claim 1 wherein building a forward model based on the grid-like reservoir comprises: designing a geological model according to an actual seismic section and combining reservoir geological awareness, and determining reservoir characteristics of a target well section from the seismic section and the real drilling blow-out loss; calculating actual speeds of various reservoirs according to logging data, and establishing the forward model according to aspect ratio of geologic bodies obtained through actual logging and logging.
3. The ultra-deep walk-slip-off-grid-like reservoir geophysical characterization method of claim 2 wherein the reservoir characteristics comprise reservoir number, beaded response characteristics, reflection characteristics.
4. The ultra-deep walk sliding off-gate type reservoir geophysical characterization method of claim 1 wherein, for the forward model, kirchihoff shift imaging is simulated with wave equation to obtain forward simulation analysis results.
5. The ultra-deep walk sliding off-grid-like reservoir geophysical characterization method of claim 4 wherein the forward modeling analysis results comprise: The amplitude characteristics of different types of reservoirs have differences, the speed of the reservoirs is low, the absolute value of the amplitude of the corresponding trough is large, namely, the probability that a string bead with a large amplitude value is drilled into a good reservoir is large, and the probability that the part with a large amplitude value in the string bead is subjected to emptying leakage is large, so that different string beads in the same area are classified and ordered through the transverse change of the amplitude, the part with a large amplitude energy is used as a target, and the part with a large amplitude in the string bead is determined to be used as a target of well track design.
6. The ultra-deep walk-slip-break-control grid-like reservoir geophysical characterization method of claim 1 wherein constructing a deep learning sample library of the grid-like reservoir from the results of forward modeling analysis comprises: When the sample library is constructed, manually interpreted data is supplemented as a label, and forward modeling analysis results are also supplemented into the sample library, so that attribute research and development aiming at medium-small scale fracture zones are carried out, the sample library established by conventional seismic attributes is expanded, and further, data enhancement processing of the sample library is carried out, so that the training quality of the deep learning network is ensured.
7. The ultra-deep walk sliding off-grid-like reservoir geophysical characterization method of claim 1 wherein building a U-net network model and training comprises: Dividing the data of the deep learning sample library into a training set and a verification set; and establishing the U-net network model, training the network model through the training set, and verifying through the verification set.
8. An ultra-deep walk-slip break-control grid-like reservoir geophysical characterization device comprising: the forward model building module is used for building a forward model based on the grid-shaped storage body; The forward modeling analysis module is used for performing grid-shaped accumulator forward modeling analysis on the forward model; the sample library construction module is used for constructing a deep learning sample library of the grid-shaped reservoir body according to the result of forward modeling analysis; and the characterization module is used for establishing a U-net network model, training and carrying out grid-shaped reservoir characterization through the trained network model.
9. An electronic device, the electronic device comprising: A memory storing executable instructions; A processor executing the executable instructions in the memory to implement the ultra-deep walk sliding off-grid reservoir geophysical characterization method of any one of claims 1-7.
10. A computer readable storage medium, characterized in that it stores a computer program which, when executed by a processor, implements the ultra-deep walk sliding break-control grid-like reservoir geophysical characterization method according to any one of claims 1-7.

Description

Ultra-deep walk-slip break-control grid-shaped reservoir geophysical characterization method, device, electronic equipment and medium Technical Field The invention relates to the field of seismic and geological engineering integration, in particular to a method, a device, electronic equipment and a medium for representing geophysical of an ultra-deep walk-slip break-control grid-shaped reservoir body. Background Ultra-deep walk-slip off-control reservoirs have great exploration potential. However, the 'break-store' relationship is complex, the characterization difficulty is high, the non-uniformity of the break-control reservoir body is high, the spatial configuration of the reservoir body is difficult to establish, and the relationship between the non-uniformity structure in the break-control reservoir body and the seismic reflection characteristics is unclear. The technical problem of ultra-deep sliding break-control grid-shaped reservoir surface is not widely studied at home and abroad. Many papers have focused on a certain technical link (e.g., research on methods such as imaging algorithms) and have not developed systematic studies based on geologic and seismic response characteristics to form an overall technique for geophysical characterization of ultra-deep carbonate walk-slip-off-control grid-like reservoirs. Therefore, there is a need to develop a method, apparatus, electronic device and medium for ultra-deep walk-slip-off-control grid-like reservoir geophysical characterization. The information disclosed in the background section of the invention is only for enhancement of understanding of the general background of the invention and should not be taken as an acknowledgement or any form of suggestion that this information forms the prior art already known to a person skilled in the art. Disclosure of Invention The invention provides a geophysical characterization method, a device, electronic equipment and a medium for ultra-deep walk-slip-break control grid-shaped reservoirs, which are used for constructing an adaptive deep learning sample library for manual intervention of the reservoirs by manually explaining achievements, innovatively researched and developed structural gradient attributes, instantaneous amplitude ratios and similarity transverse change rate attributes on the basis of a conventional seismic attribute initial sample library and applying deep learning prediction to characterize the break control reservoirs. In a first aspect, embodiments of the present disclosure provide a method of ultra-deep walk-slip off-control grid-like reservoir geophysical characterization comprising: establishing a forward model based on a grid-shaped accumulator; performing grid-shaped accumulator forward modeling analysis on the forward model; Constructing a deep learning sample library of the grid-shaped reservoir body according to the result of forward modeling analysis; and establishing a U-net network model, training, and carrying out grid-shaped reservoir body characterization through the trained network model. As a specific implementation manner of an embodiment of the present disclosure, establishing a forward model based on a grid-like repository includes: designing a geological model according to an actual seismic section and combining reservoir geological awareness, and determining reservoir characteristics of a target well section from the seismic section and the real drilling blow-out loss; calculating actual speeds of various reservoirs according to logging data, and establishing the forward model according to aspect ratio of geologic bodies obtained through actual logging and logging. As a specific implementation of an embodiment of the disclosure, the reservoir characteristics include a reservoir number, a bead response characteristic, and a reflection characteristic. As a specific implementation manner of the embodiment of the disclosure, for the forward model, a wave equation is used for simulating Kirchihoff offset imaging, and a forward simulation analysis result is obtained. As a specific implementation manner of the embodiment of the present disclosure, the forward modeling analysis result includes: The amplitude characteristics of different types of reservoirs have differences, the speed of the reservoirs is low, the absolute value of the amplitude of the corresponding trough is large, namely, the probability that a string bead with a large amplitude value is drilled into a good reservoir is large, and the probability that the part with a large amplitude value in the string bead is subjected to emptying leakage is large, so that different string beads in the same area are classified and ordered through the transverse change of the amplitude, the part with a large amplitude energy is used as a target, and the part with a large amplitude in the string bead is determined to be used as a target of well track design. As a specific implementation manner of an embodiment of the present di