Looking for breakthrough ideas for innovation challenges? Try Patsnap Eureka!

Wide-azimuth pre-stack seismic reflection mode analysis method of tensor depth self-encoding network

A self-encoding network and pre-stack seismic technology, applied in the field of wide-azimuth pre-stack seismic reflection mode analysis, can solve problems that are not suitable for seismic data reflection mode analysis

Inactive Publication Date: 2020-10-30
UNIV OF ELECTRONICS SCI & TECH OF CHINA
View PDF6 Cites 5 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

When using wide-azimuth pre-stack seismic data for reflection pattern analysis, multiple wide-azimuth pre-stack data will form a four-dimensional data volume. Therefore, image-based deep learning models are no longer suitable for wide-azimuth pre-stack seismic data. Reflection Pattern Analysis

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
  • Wide-azimuth pre-stack seismic reflection mode analysis method of tensor depth self-encoding network
  • Wide-azimuth pre-stack seismic reflection mode analysis method of tensor depth self-encoding network
  • Wide-azimuth pre-stack seismic reflection mode analysis method of tensor depth self-encoding network

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0055] The technical principle of the present invention will be described below.

[0056] This invention will use the following notation: Tensors are represented using bold calligraphic letters, such as Matrices are represented by bold uppercase letters, such as T; vectors are represented by bold lowercase letters, such as t; scalars are represented by lowercase letters, such as t. Tensor space for 3D seismic data Use i, j, k to denote the xline, inline and time dimensions of 3D seismic data. Therefore, i∈[N x ], j ∈ [N y ], k∈[N t ], where [N] represents the set {1, 2, ..., N}. for tensor elements in available express. The i-th horizontal slice (horizontal slice), the j-th side slice (lateral slice) and the k-th frontal slice (frontal slice) are denoted as where the k-th frontal slice is available express.

[0057] Assume is a three-dimensional tensor of dimension l×m×n. Fixed third dimension, forward slicing It is a matrix with a dimension of l×m, where...

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 wide-azimuth pre-stack seismic reflection mode analysis method for a tensor depth self-encoding network, and the method comprises the following steps: S1, data preprocessing:extracting the wide-azimuth pre-stack seismic data of a target layer according to a layer target analyzed by a seismic reflection mode, and forming a three-dimensional tensor sample set; S2, traininga TDOEN network by using the sample set; S3, obtaining deep features of the sample set; and S4, clustering depth features obtained in the step S3 by combining an SOM clustering algorithm to generate aseismic facies division result. According to the method, a tensor model represented by high-dimensional data is adopted, through tensor product operation model analysis, a tensor-based weight networkhas fewer weight parameters, and feature expression can be performed more fully. Therefore, the tensor-based learning network is more suitable for processing wide azimuth data than a vector (matrix)-based deep learning model.

Description

technical field [0001] The invention belongs to the technical field of geophysical exploration and interpretation, and in particular relates to a wide-azimuth pre-stack seismic reflection mode analysis method of a tensor depth self-encoding network. Background technique [0002] Lithology identification and fracture identification methods based on wide-azimuth pre-stack seismic data have a very broad application prospect in seismic data interpretation and processing, especially with the gradual improvement of exploration technology, the analysis of reflection patterns based on wide-azimuth pre-stack seismic data Methods began to develop gradually. Due to the very high similarity between the data in the prestack gather, the noise in the data is random. Therefore, using adjacent multi-channel prestack seismic data for joint depth feature extraction can improve the robustness of depth feature extraction. The feature extraction of prestack angle gather data can be carried out ...

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): G01V1/30G06K9/62G06N3/04G06N3/08
CPCG01V1/306G06N3/084G01V2210/624G06N3/048G06N3/045G06F18/23
Inventor 钱峰冯令田廖松杰胡光岷
Owner UNIV OF ELECTRONICS SCI & TECH OF CHINA
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Patsnap Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Patsnap Eureka Blog
Learn More
PatSnap group products

Browse by: Latest US Patents, China's latest patents, Technical Efficacy Thesaurus, Application Domain, Technology Topic, Popular Technical Reports.

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap|About US| Contact US: help@patsnap.com