Unlock instant, AI-driven research and patent intelligence for your innovation.

Deep learning oriented reservoir prediction sample expanding method

A technology for reservoir prediction and sample expansion, which is applied in the field of reservoir prediction sample expansion for deep learning, and can solve the problems of insufficient reservoir samples and inability to use deep learning.

Inactive Publication Date: 2019-02-15
CHINA PETROLEUM & CHEM CORP +1
View PDF4 Cites 1 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0003] The purpose of the present invention is to provide a deep learning-oriented reservoir prediction sample expansion method to solve the problem that deep learning and other big data analysis technologies cannot be used due to insufficient reservoir samples.

Method used

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View more

Image

Smart Image Click on the blue labels to locate them in the text.
Viewing Examples
Smart Image
  • Deep learning oriented reservoir prediction sample expanding method
  • Deep learning oriented reservoir prediction sample expanding method
  • Deep learning oriented reservoir prediction sample expanding method

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0033] Embodiment 1, a deep learning-oriented reservoir prediction sample expansion method, first find the seismic traces by the well-seismic combination, extract the seismic data next to the well, and convert the seismic data into a character string form that can reflect the distribution of the reservoir, The well-seismic reservoir correspondence relationship is established, and then according to the seismic string matching mode established by the reservoir distribution on the well, this mode is matched with the seismic traces within a certain range around the well, and the time point of matching seismic traces is recorded as the match of the reservoir. Finally, these marker points are recorded to obtain sample labels that can reflect different reservoir types and conform to geological characteristics.

[0034] The aforementioned method specifically includes the following steps:

[0035] a. Through well coordinate conversion, the well coordinates are converted into seismic li...

Embodiment 2

[0041] Embodiment 2, a deep learning-oriented reservoir prediction sample expansion method, figure 1 For this method main steps and flow process, its specific implementation is as follows:

[0042] (1) Match the seismic line number according to the well coordinates and extract the corresponding seismic line data.

[0043] (2) Find the maximum and minimum values ​​of the amplitude in the seismic data volume, and normalize the seismic data to the range of [-1,1] according to the maximum and minimum values; select the number of character strings according to the reservoir distribution characteristics on the well, And each character string corresponds to a value range, such as c=[-1,0], b=[0,0.2], a=[0.2,1]; after assigning each character a specific color, the Seismic traces are displayed, figure 2 Assign a to red to represent the peak, b to blue to represent near 0, c to yellow to represent the trough, and the seismic waveform features are converted into strings.

[0044] (3)...

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

PUM

No PUM Login to View More

Abstract

The invention provides a deep learning oriented reservoir prediction sample expanding method, which includes the following steps: near-well seismic data is extracted through well-to-seismic integration, and the seismic data is converted into a character string form capable of reflecting reservoir distribution; a corresponding relationship between well-seismic reservoirs is established by time-to-depth conversion, and a seismic character string matching pattern is established according to the reservoir distribution on wells; character string matching is carried out on the character string pattern and the seismic traces in a certain range around wells, the time points corresponding to the matched seismic traces are recorded, and the seismic sampling points corresponding to the points are marked as samples conforming to corresponding reservoirs; and the marked points are recorded in turn to obtain sample labels capable of reflecting different reservoir types and conforming to geological characteristics. A large number of effective training samples can be provided for big data analysis of reservoir prediction.

Description

technical field [0001] The invention belongs to the field of geophysical exploration seismic data interpretation and comprehensive research, and in particular relates to a deep learning-oriented reservoir prediction sample expansion method. Background technique [0002] Big data analysis technology has played a huge role in the fields of banking, communications, and the Internet. Multinational giants such as IBM, ORACLE, Microsoft, Google, Amazon, and Facebook, as well as domestic companies such as Ali, Tencent, and Baidu, have also become more and more powerful due to the development of big data technology. competitive. In the field of petroleum exploration, reservoir prediction is an important research content to find favorable exploration targets. The data involved include a large amount of data such as seismic, well logging, mud logging, and oil testing. These data are all indirect reflections of underground rocks. There is correlation between them, which conforms to th...

Claims

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

Application Information

Patent Timeline
no application Login to View More
IPC IPC(8): G01V1/50G01V1/30
CPCG01V1/306G01V1/50G01V2210/624
Inventor 朱剑兵韩宏伟王兴谋冯德永李长红罗荣涛赵庆国毕丽飞
Owner CHINA PETROLEUM & CHEM CORP