CN-121982836-A - Intelligent early warning method based on underground coal mine gas analysis

CN121982836ACN 121982836 ACN121982836 ACN 121982836ACN-121982836-A

Abstract

The invention discloses an intelligent early warning method based on underground gas analysis of a coal mine, and relates to the technical field of intelligent control; acquiring video information in a mine through an infrared thermal imager, acquiring multi-segment video information, carrying out image recognition processing to acquire gas leakage characteristics, carrying out traversal comparison on the gas leakage characteristics and a preset early warning characteristic set, determining gas early warning state information, carrying out fitting positioning with a mine simulation module, determining gas early warning states at all positions of the mine, matching an early warning strategy, and carrying out early warning control. The invention solves the technical problems of low sensitivity and low accuracy of the intelligent analysis and early warning of the underground gas of the coal mine in the prior art, and achieves the technical effect of reducing the accident probability by improving the sensitivity and the accuracy of the intelligent analysis and early warning of the underground gas of the coal mine.

Inventors

ZHANG LULU
HU ZHENDONG
YU DONGBO
ZHONG DAWEI
ZHANG YOUNING
CAO WENXIN
LU CUNYU
QIN NA
GUO CHEN
LIU ZUNLI
HUANG HANZENG
CUI XIAOMING
HOU SHOUQING
WANG FENGJIANG
DONG TAO
LI SHUMING

Assignees

山东鼎安检测技术有限公司

Dates

Publication Date: 20260505
Application Date: 20251225

Claims (8)

1. An intelligent early warning method based on underground coal mine gas analysis is characterized by comprising the following steps: collecting underground space arrangement and structure information of a coal mine, and constructing a mine simulation module, wherein the mine simulation module is used for simulating the structure of the coal mine; the mine simulation module is connected with a mine monitoring system, the mine operation state information is synchronized, operation main body identification is carried out based on a mine monitoring system acquisition picture, and a simulation label is generated and added into the mine simulation module; acquiring video information in a mine through a thermal infrared imager to obtain multi-segment video information, wherein the multi-segment video information has position information and the position information has regional continuity; performing image recognition processing on the multi-segment video information through a picture processing module to obtain gas leakage characteristics; Performing traversal comparison by using the gas leakage characteristics and a preset early warning characteristic set to determine gas early warning state information; And carrying out fitting positioning according to the gas early-warning state information and the mine simulation module, determining the gas early-warning state of each position of the mine, matching the early-warning strategy, decomposing and distributing the matched early-warning strategy to the early-warning execution module for early-warning control.
2. The method as recited in claim 1, further comprising: Activating a sensor control module when the gas leak characteristic meets a sensor activation threshold; The gas leakage characteristic is matched with a sensor set activation threshold value, and a sensor executing strategy is determined, wherein the sensor executing strategy comprises a multi-stage sensor control instruction, and the multi-stage sensors have different detection sensitivities; Based on the executing sensor strategy, acquiring sensor detection data, and carrying out early warning evaluation on the sensor detection data according to a threshold evaluation algorithm to obtain sensor evaluation state information; and carrying out fusion analysis by utilizing the gas early warning state characteristics and the sensor evaluation state information, and matching the early warning strategy by utilizing fusion analysis information.
3. The method of claim 1, wherein the performing, by a picture processing module, image recognition processing on the multi-segment video information to obtain a gas leakage feature comprises: preprocessing the multi-segment video information through a preprocessing sub-module in a picture processing module, wherein preprocessing is to remove randomly distributed noise points in a background area through non-local mean denoising; The method comprises the steps of respectively inputting preprocessed multi-segment video information into a state analysis submodule and a key frame analysis submodule, extracting the video information according to a preset step length through the state analysis submodule, and carrying out leakage state identification analysis on the extracted video to obtain leakage state information; and integrating the leakage state information and the gas leakage picture characteristics by a picture processing module to obtain the gas leakage characteristics.
4. The method of claim 3, wherein the extracting key frames of the video information by the key frame analysis submodule and the identifying and analyzing leakage characteristics of the extracted key frames to obtain the gas leakage picture characteristics comprise: defining a video initial frame, taking the initial frame as a key frame, and setting a frame number relation between the total frame number of the video and the key frame; Comparing the key frame with the next frame, and calculating a frame difference value through a color histogram; Setting a threshold value, taking the current frame and the next frame as new key frames when the frame difference value exceeds the threshold value, and the like until all frames are completed; Determining a key frame by using the obtained relation between the key frame and the frame number; and carrying out edge segmentation recognition on the key frame, determining picture edge information, carrying out in-region gray scale calculation based on the picture edge information, and determining picture gray scale value change characteristics to obtain the gas leakage picture characteristics.
5. The method as recited in claim 2, further comprising: determining a gas leakage characteristic change trend according to the gas leakage characteristic, and determining a gas concentration characteristic based on the gas leakage characteristic change trend; adding gas concentration characteristics into the mine simulation module according to the position information of the multi-segment video information; Carrying out mine risk partitioning according to the depth of the structure in the mine and the operation main body through the mine simulation module; Positioning a risk area according to the mine risk partition and the gas concentration characteristic; Activating a gas flow rate detection sensor of the area association partition according to the risk area positioning information, and constructing a flow rate data chain; and carrying out state prediction of the positioning risk area based on the flow velocity data chain, and generating early warning information by utilizing a prediction state matching early warning strategy.
6. The method of claim 5, wherein locating risk zones in accordance with the mine risk zone, the gas concentration profile, comprises: Determining the concentration of the operation main body of each partition and the space fluidity of each partition according to the mine risk partition; Determining a gas variation flow direction according to the gas concentration characteristics; Determining the concentration of the operation main body of each partition, the space fluidity of each partition and the distribution weight of the gas concentration characteristic through a case library; determining a flow direction correction coefficient based on the gas variation flow direction and the gas concentration characteristic, wherein the flow direction correction coefficient corresponds to the gas variation flow direction; calculating risk values of all the subareas according to the distribution weight operation main body concentration, the space fluidity and the gas concentration characteristics; and correcting the risk value of each partition by using the flow direction correction coefficient, and positioning the risk area according to the correction result, wherein the positioning risk area is the partition with the higher correction risk value.
7. The method as recited in claim 2, further comprising: Collecting a multi-accident infrared image set; Classifying the multi-accident infrared image set to construct a multi-type accident cluster; respectively carrying out feature decomposition on the multi-type accident clusters to construct an accident decomposition time sequence feature set; Determining basic event features and abnormal node core features based on the accident decomposition time sequence feature set, and marking the accident decomposition time sequence feature set according to the basic event features and the abnormal node core features; tracking and traversing the gas leakage characteristics by using the marked accident decomposition time sequence characteristic set, and determining characteristic fitness and coincidence time sequence relation; And matching event characteristics according to the characteristic fitness and the matching time sequence relationship, and predicting accident occurrence probability and accident types.
8. An intelligent early warning system based on underground coal mine gas analysis, which is characterized by comprising: The system comprises a mine simulation module construction module, a control module and a control module, wherein the mine simulation module construction module is used for acquiring the space setting and structure information of a coal mine underground and constructing a mine simulation module, and the mine simulation module is used for simulating the structure of the coal mine underground; The simulation label generation module is used for connecting the mine simulation module with a mine monitoring system, synchronizing mine operation state information, carrying out operation main body identification based on a mine monitoring system acquisition picture, generating a simulation label and adding the simulation label into the mine simulation module; The system comprises a multi-segment video information acquisition module, a video processing module and a video processing module, wherein the multi-segment video information acquisition module is used for acquiring video information in a mine through a thermal infrared imager to obtain multi-segment video information, and the multi-segment video information has position information and has regional continuity; The gas leakage characteristic obtaining module is used for carrying out image identification processing on the multi-segment video information through the picture processing module to obtain gas leakage characteristics; the gas early warning state information determining module is used for determining gas early warning state information by performing traversal comparison on the gas leakage characteristics and a preset early warning characteristic set; And the early warning control module is used for carrying out fitting positioning with the mine simulation module according to the gas early warning state information, determining the gas early warning state of each position of the mine, matching the early warning strategy, and carrying out decomposition and distribution on the matched early warning strategy to the early warning execution module for early warning control.

Description

Intelligent early warning method based on underground coal mine gas analysis Technical Field The invention relates to the technical field of intelligent control, in particular to an intelligent early warning method based on underground coal mine gas analysis. Background Harmful gas rushing out of coal rock in the underground coal mine excavating process is one of important reasons for causing major accidents such as fire, explosion and the like, so that the underground gas detection and early warning are very important for preventing the occurrence of coal mine events. However, the existing mine gas monitoring method also has the problems that the gas sample analysis result is lagged, the real-time performance of gas analysis is poor, and the accuracy of gas analysis is affected due to the conditions of pipeline blockage, breakage or overlarge resistance and the like easily occurring due to overlong gas extraction pipelines, so that the sensitivity and the accuracy of safety early warning are affected. Disclosure of Invention The application provides an intelligent early warning method based on underground coal mine gas analysis, which is used for solving the technical problems of low sensitivity and low accuracy of the intelligent underground coal mine gas analysis early warning in the prior art. The application provides an intelligent early warning method based on underground coal mine gas analysis, which comprises the steps of collecting underground coal mine space setting and structure information, constructing a mine simulation module, connecting the mine simulation module with a mine monitoring system, synchronizing mine operation state information, carrying out operation main body recognition based on a mine monitoring system collecting picture, generating simulation labels, adding the simulation labels into the mine simulation module, collecting video information in a mine through an infrared thermal imager, obtaining multi-segment video information, wherein the multi-segment video information has position information and has regional continuity, carrying out image recognition processing on the multi-segment video information through a picture processing module, obtaining gas leakage characteristics, carrying out traversal comparison on the gas leakage characteristics and a preset early warning characteristic set, determining gas state information, carrying out fitting positioning according to the gas state information and the mine simulation module, determining the gas early warning state of each position of the mine, and distributing the matching early warning strategy to an early warning executing module for early warning control. The intelligent early warning system based on the underground coal mine gas analysis comprises a mine simulation module construction module, a simulation label generation module, a gas early warning state information determination module, a multi-segment video information acquisition module, an early warning state information comparison module and an early warning state information determination module, wherein the mine simulation module construction module is used for acquiring underground coal mine space setting and structure information and constructing a mine simulation module which is used for simulating a coal mine structure, the simulation label generation module is used for connecting with a mine monitoring system through the mine simulation module, synchronizing mine operation state information and carrying out operation main body recognition on the basis of a mine monitoring system acquisition picture, generating simulation labels and adding the simulation labels into the mine simulation module, the multi-segment video information acquisition module is used for acquiring the video information in a mine through an infrared thermal imager and acquiring the multi-segment video information, the multi-segment video information has position information and area continuity, the gas leakage characteristic acquisition module is used for carrying out image recognition processing on the multi-segment video information through a picture processing module and acquiring the gas leakage characteristic, the gas early warning state information determination module is used for carrying out the matching of the gas early warning state information with the pre-warning state information determination module, and the early warning state information is used for carrying out the early warning state analysis and the early warning state determination module is matched with the position-warning state control strategy. One or more technical schemes provided by the application have at least the following technical effects or advantages: The application provides an intelligent early warning method based on underground gas analysis of a coal mine, which relates to the technical field of intelligent control, and constructs a mine simulation module by collecting underground space setting and s