US 11,630,227 B1Granted

Method and system for recognizing mine microseismic event

AssigneeLiaoning University

InventorsYishan Pan, Linlin Ding, Lujie Cao, Hanlin Zhang, Hao Luo

Life of this patent

Estimating the term…

Loading patent…

Abstract

Embodiments of the present disclosure provide a method and system for recognizing a mine microseismic event, and belong to the field of mine data processing. The method includes: converting historical microseismic data monitored by a mine microseismic monitoring system into a microseismic waveform image, and then, converting the microseismic waveform image into a four-neighborhood microseismic waveform graph structure; performing area defining on the microseismic waveform graph structure, and extracting a similar feature layer of any node in the microseismic waveform graph structure based on the defined area; and taking the microseismic waveform image as an input layer of an improved convolutional neural network model, and sequentially connecting the input layer with the similar feature layer as well as a convolutional layer, a pooling layer, a fully connected layer and an output layer which are pre-configured for the improved convolutional neural network model to form a recognition model for recognizing the mine microseismic event. By using the recognition model designed in the present disclosure, the similar feature layer can be extracted, so that the mine microseismic event is effectively recognized.

Claims20

Dependent claims

A method for recognizing a mine microseismic event, wherein the method comprises:

monitoring historical microseismic data by a mine microseismic monitoring device;
and
performing the following steps by a calculation device:
converting the historical microseismic data monitored by the mine microseismic monitoring device into a microseismic waveform image, and then, converting the microseismic waveform image into a four-neighborhood microseismic waveform graph structure;
performing area defining on the microseismic waveform graph structure, and extracting a similar feature layer of any node in the microseismic waveform graph structure based on the defined area, wherein a feature of each node is represented by Red Green Blue (RGB) information of k similar nodes of the node, and the similar feature layer refers to k layers of similar nodes formed by arranging the k similar nodes in ascending order of similarity, wherein k is an integer greater than 1;
and
taking the microseismic waveform image as an input layer of an improved convolutional neural network model, and connecting the input layer with the similar feature layer, a convolutional layer, a pooling layer, a fully connected layer and an output layer to form a recognition model for recognizing the mine microseismic event,
wherein the convolutional layer, the pooling layer, the fully connected layer, and the output layer are preconfigured for the improved convolutional neural network model sequentially.

5 more elements

A system for recognizing a mine microseismic event, wherein the system comprises:

a mine microseismic monitoring device for monitoring historical microseismic data,
a processor, a memory and a program stored on the memory and capable of running on the processor, and the program is executed by the processor to implement the following steps:
converting the historical microseismic data into a microseismic waveform image, and then, converting the microseismic waveform image into a four-neighborhood microseismic waveform graph structure;
performing area defining on the microseismic waveform graph structure, and extracting a similar feature layer of any node in the microseismic waveform graph structure based on the defined area, wherein a feature of each node is represented by Red Green Blue (RGB) information of k similar nodes of the node, and the similar feature layer refers to k layers of similar nodes formed by arranging the k similar nodes in ascending order of similarity, wherein k is an integer greater than 1;
and
taking the microseismic waveform image as an input layer of an improved convolutional neural network model, and connecting the input layer with the similar feature layer, a convolutional layer, a pooling layer, a fully connected layer and an output layer which are to form a recognition model for recognizing the mine microseismic event,
wherein the convolutional layer, the pooling layer, the fully connected layer, and the output layer are preconfigured for the improved convolutional neural network model sequentially.

4 more elements

Dependencies are inferred from the claim text.

Facts

Application: 17/860,297Filed 8 Jul 2022
Priority: CN 202210281630.521 Mar 2022
Granted: 18 Apr 2023
Term ends: Not estimatedThe source records don’t support an estimate.

Classification8 codes

G01V 1/345
Displaying seismic recordings or visualisation of seismic data or attributes: visualisation of seismic data or attributes, e.g. in 3D cubes
G01V 1/282
Processing seismic data, e.g. for interpretation or for event detection: application of seismic models, synthetic seismograms
G01V 1/288
Processing seismic data, e.g. for interpretation or for event detection: event detection in seismic signals, e.g. microseismics
G01V 2210/123
Signal generation: passive source, e.g. microseismics