Seismic velocity spectrum automatic pickup method and device based on unsupervised learning

Abstract

The invention discloses a seismic velocity spectrum automatic pickup method and device based on unsupervised learning. The method comprises the steps: acquiring a seismic stacking velocity spectrum; scanning the maximum reference stacking speed and the minimum reference stacking speed of the corridor according to a preset speed, and cutting off the acquired seismic stacking speed spectrum; carrying out normalization and threshold filtering processing on the seismic stacking velocity spectrum after removal processing to obtain a seismic stacking velocity spectrum after scatter processing; in the time direction, at each time sampling points, executing the following steps to carry out stacking velocity automatic pickup on the scattered seismic stacking velocity spectrum: constructing two subspace-time windows with an overlapping relationship in a main space-time window of each time sampling point; respectively clustering the data in the two sub space-time windows by adopting a weighted K-mean algorithm; and determining the clustering center of the main space-time window according to the clustering centers of the two sub space-time windows in the main space-time window. The method anddevice can determine the center of the velocity spectrum energy group in a self-adaptive manner.

Description

technical field [0001] The invention relates to the technical field of oil and gas exploration, in particular to a method and device for automatically picking up seismic velocity spectrum based on unsupervised learning. Background technique [0002] This section is intended to provide a background or context to embodiments of the invention that are recited in the claims. The descriptions herein are not admitted to be prior art by inclusion in this section. [0003] As we all know, seismic velocity spectrum picking is a very important link in seismic data processing, and the picking results directly affect the quality of seismic imaging. The artificial picking method of seismic velocity spectrum, on the one hand, is affected by the level of processors or human subjectivity, and on the other hand, it is increasingly unable to meet the requirements of users for massive data and high-density velocity spectrum picking and processing cycles. In recent years, it has gradually been...

AI Technical Summary

Problems solved by technology

[0006] The embodiment of the present invention provides an automatic picking method of seismic velocity spectrum based on unsupervised learning, which is used to solve the problem that the existing automatic picking method of seismic velocity spectrum based on unsupervised learning cannot accurately set the clustering center that changes with space Quantitative technical problems, the method includes: collecting and obtaining the seismic stacking velocity spectrum; according to the maximum reference stacking velocity and the minimum reference stacking velocity of the preset speed scanning corridor, cutting the collected seismic stacking velocity spectrum; The seismic stacking velocity spectrum is normalized and threshold value filtered to obtain the scattered seismic stacking velocity spectrum; along the time direction, sample points one by one, and perform the following steps to automatically stack the velocity spectrum of the scattered seismic stacking velocity spectrum. Picking: Construct the first sub-time-space window and the second sub-time-space window with an overlapping relationship within the main time-space window of each time sampling point; use the weighted K-mean algorithm, respectively Cluster the data in the time-space window to obtain the cluster centers of the first sub-time-space window and the second sub-time-space window; determine the main time-space according to the cluster centers of the first sub-time-space window and the second sub-time-space window window cluster center

[0007] The embodiment of the present invention also provides an automatic seismic velocity spectrum picking device based on unsupervised learning, which is used to solve the problem that the existing automatic picking method of seismic velocity spectrum based on unsupervised learning cannot accurately set the clustering with spatial changes The technical problem of the number of centers, the device includes: a stacking velocity spectrum acquisition module, used to acquire the seismic stacking velocity spectrum; a data removal processing module, used to scan the corridor's maximum reference stacking velocity and minimum reference stacking velocity according to the preset speed, for The acquired seismic stacking velocity spectrum is cut and processed; the data scatter processing module is used to normalize and threshold filter the cut seismic stacked velocity spectrum to obtain the scattered seismic stacked velocity spectrum; Velocity spectrum automatic picking module, including: sub-space-time window construction module, sub-space-time window cluster center determination module and main space-time window cluster center determination module, used to implement sub-space-time window construction module, The sub-time-space window cluster center determination module and the function of the main space-time window cluster center determination module are used to automatically pick up the stacking velocity of the scattered seismic stacking velocity spectrum; among them, the sub-time-space window construction module is used for each The first sub-time-space window and the second sub-time-space window with overlapping relationship are constructed in the main time-space window of the time sampling point; the sub-time-space window clustering center determination module is used to adopt the K-mean algorithm with weighting, respectively for the first sub-time-space window The data in the time-space window and the second sub-time-space window are clustered to obtain the cluster centers of the first sub-time-space window and the second sub-time-space window; the main space-time window cluster center determination module is used to The cluster centers of the sub-space-time window and the second sub-space-time window determine the cluster center of the main space-time window

[0008] In the embodiment of the present invention, a computer device is also provided to solve the technical problem that the existing automatic picking method of seismic velocity spectrum based on unsupervised learning cannot accurately set the number of cluster centers that vary with space. The computer device It includes a memory, a processor, and a computer program stored on the memory and operable on the processor. When the processor executes the computer program, the above-mentioned method for automatically picking up seismic velocity spectrum based on unsupervised learning is realized.

Images