Login| Sign Up| Help| Contact|

Patent Searching and Data


Title:
AUTOMATIC SEISMIC FACIES IDENTIFICATION METHOD BASED ON COMBINATION OF SELF-ATTENTION MECHANISM AND U-SHAPED STRUCTURE
Document Type and Number:
WIPO Patent Application WO/2024/000709
Kind Code:
A1
Abstract:
An automatic seismic facies identification method based on a combination of a self-attention mechanism and a U-shaped structure, comprising: obtaining a post-stack seismic data volume, and preprocessing the post-stack seismic data volume to construct a sample training and verification data set; constructing an encoder by using an overlapping block merging module having a down-sampling function and an efficient self-attention converter module capable of global modeling representation; constructing a decoder by using a block expansion module having a linear up-sampling function, the efficient self-attention converter module, and a skip connection module capable of fusing high and lower layer features; constructing a seismic facies identification model by using the encoder, the decoder and a hypercolumn module, the seismic facies identification model comprising a Hypercolumns-U-Segformer; constructing a hybrid loss function, and iteratively training the seismic facies identification model by using training and verification sets in the sample training and verification data set; and inputting test data to obtain seismic facies of test seismic data. By means of public data and actual data experiments, compared with a seismic facies identification model of a conventional convolutional neural network, the present application has fewer noise points, more obvious boundary details and stronger lateral continuity, and can explain different underground deposition environments more accurately.

Inventors:
WANG ZHIGUO (CN)
CHEN YUMIN (CN)
YANG YANG (CN)
GAO ZHAOQI (CN)
LI ZHEN (CN)
WANG QIANNAN (CN)
GAO JINGHUAI (CN)
Application Number:
PCT/CN2022/108319
Publication Date:
January 04, 2024
Filing Date:
July 27, 2022
Export Citation:
Click for automatic bibliography generation   Help
Assignee:
UNIV XI AN JIAOTONG (CN)
International Classes:
G06Q10/04; G06N3/04; G06N3/08; G06Q50/26; G06V10/22; G06V10/32; G06V10/764; G06V10/774; G06V10/82
Foreign References:
CN114660656A2022-06-24
CN112731522A2021-04-30
CN112711072A2021-04-27
CN111626355A2020-09-04
CN113296152A2021-08-24
CN110516740A2019-11-29
US20210063594A12021-03-04
US20210247531A12021-08-12
Attorney, Agent or Firm:
WAYNE INTELLECTUAL PROPERTY AGENCY CO., LTD. (CN)
Download PDF: