CN-121982827-A - Mining broadcast linkage early warning method and system based on environment awareness

CN121982827ACN 121982827 ACN121982827 ACN 121982827ACN-121982827-A

Abstract

The invention belongs to the technical field of mining safety monitoring, and in particular relates to a mining broadcasting linkage early warning method and system based on environment awareness, wherein the method comprises the steps of collecting an infrared thermal image sequence and an audio frame sequence of a mine environment by using a mining broadcasting substation; the method comprises the steps of determining an audio abnormal sound index by analyzing the entropy deviation degree and the energy spectrum discrete condition of any frame of audio frequency spectrum, determining a thermal abnormal diffusion index by combining the divergence change of a heat source divergence set of any frame of infrared thermal image and the area change rate of a high Wen Liantong area, and generating a comprehensive risk linkage coefficient by introducing the association degree of two index sequences in a reference frame, thereby driving grading alarm. The invention can effectively distinguish normal high load and fault states of equipment, reduce false alarm rate and realize underground acoustic-thermal linkage early warning.

Inventors

WU GUANREN
GAO HONGPENG
CHEN XIAOPENG

Assignees

山西省元智控科技有限公司

Dates

Publication Date: 20260505
Application Date: 20260402

Claims (10)

1. The mining broadcasting linkage early warning method based on environment awareness is characterized by comprising the following steps of: collecting an infrared thermal image sequence and an audio frame sequence of a mine environment by utilizing a mine broadcasting substation; Carrying out frequency domain analysis on any frame of audio to obtain the energy spectrum of each frequency point and the frequency spectrum entropy of the frame of audio, and determining an audio abnormal sound index according to the deviation degree of the frequency spectrum entropy relative to the preset normal frequency spectrum entropy and the discrete condition of the energy spectrum of the frequency point; acquiring a heat source divergence set of any frame of infrared thermal image and the area of a high Wen Liantong area, and determining a thermal anomaly diffusion index according to the divergence change of the heat source divergence set and the area change rate of the high Wen Liantong area; Determining a comprehensive risk linkage coefficient according to the product of the audio abnormal sound index and the thermal abnormal diffusion index and the association degree between two index sequences in the reference frame; and according to the comprehensive risk linkage coefficient, matching a preset risk level threshold, and driving the broadcasting substation chip to execute corresponding voice alarm.
2. The mining broadcast linkage early warning method based on environment awareness according to claim 1, wherein the step of performing frequency domain analysis on any frame of audio to obtain an energy spectrum of each frequency point and a spectrum entropy of the frame of audio comprises the following steps: performing fast Fourier transform on any frame of audio to obtain the frequency domain amplitude of each frequency point, squaring the frequency domain amplitude of each frequency point to obtain an energy spectrum of each frequency point, performing sum normalization processing on the energy spectrum to obtain the spectrum probability of each frequency point, and substituting the spectrum probability into a shannon entropy formula to obtain the spectrum entropy of the frame of audio.
3. The mining broadcast linkage early warning method based on environment awareness according to claim 1, wherein the acquisition mode of the preset normal spectrum entropy is as follows: and extracting the average value of the spectrum entropy of all the frame audios in the mining safety test process as a preset normal spectrum entropy.
4. The mining broadcast linkage early warning method based on environment awareness according to claim 1, wherein the thermal anomaly diffusion index satisfies the expression: ; In the formula, Is the first An audio abnormal sound index of the frame; Is the first Spectral entropy of frame audio; presetting normal frequency spectrum entropy; Is the first Standard deviation and mean value of energy spectrum of frame audio frequency inner frequency point; Is a weight coefficient; taking absolute value symbols; As a natural exponential function.
5. The mining broadcast linkage early warning method based on environment awareness according to claim 1, wherein the acquiring the heat source divergence set of any frame of infrared thermal image comprises: And extracting the temperature value of any frame of infrared thermal image to form a temperature matrix, applying a negative Laplacian operator to the temperature matrix to carry out convolution operation to obtain a divergence matrix, and taking pixel points with the median value of the divergence matrix being greater than zero as a heat source divergence set.
6. The mining broadcast linkage early warning method based on environment awareness according to claim 1, wherein the method for acquiring the area of the high-temperature communication area is as follows: And (3) carrying out threshold segmentation on any frame of infrared thermal image by using an Ojin method, extracting connected domains from the region larger than the threshold, and taking the total number of pixels of the largest connected domain as the area of a region Wen Liantong high in the frame of infrared thermal image.
7. The mining broadcast linkage early warning method based on environment awareness according to claim 1, wherein the thermal anomaly diffusion index satisfies the expression: ; In the formula, Is the first A thermal anomaly diffusion index for the frame; Is the first The average value of all the divergences in the frame infrared thermal image heat source divergence set; The method comprises the steps of acquiring the mean value and standard deviation of all the divergences in a heat source divergence set based on a mining safety test process; Is the first Frame and th The area of the high Wen Liantong area in the frame infrared thermal image; the area of the infrared thermal image is the total number of pixel points; Is a natural constant; As a logarithmic function; as a function of the maximum value.
8. The mining broadcast linkage early warning method based on environment awareness according to claim 1, wherein the comprehensive risk linkage coefficient satisfies the expression: ; In the formula, Is the first A frame comprehensive risk linkage coefficient; Is the first An audio abnormal sound index and a thermal abnormal diffusion index of the frame; Is the first Pearson correlation coefficients of an audio abnormal sound index sequence and a thermal abnormal diffusion index sequence in a reference frame of the frame; Is gain strength, wherein the reference frame is a succession preceding the current frame The frame of the frame is a frame of a frame, Is a positive integer greater than 1.
9. The mining broadcast linkage early warning method based on environment awareness according to claim 1, wherein the driving broadcast substation chip executes corresponding voice warning, and the method comprises the following steps: The preset risk level threshold value is And If the comprehensive risk linkage coefficient is greater than And is smaller than Judging yellow early warning, driving the broadcasting substation chip to play a prestored prompt tone file, and if the comprehensive risk linkage coefficient is greater than or equal to that And judging that the alarm is red, and driving the chip to circularly play the full-volume emergency evacuation instruction.
10. The mining broadcasting linkage early warning system based on environment awareness is characterized by comprising a processor and a memory, wherein the memory stores computer program instructions, and the computer program instructions, when executed by the processor, realize the mining broadcasting linkage early warning method based on environment awareness according to any one of claims 1-9.

Description

Mining broadcast linkage early warning method and system based on environment awareness Technical Field The invention relates to the technical field of mining safety monitoring. More particularly, the invention relates to a mining broadcast linkage early warning method based on environmental awareness. Background In modern coal mining operation, underground environment is complex and changeable, various potential safety hazards such as gas protrusion, roof permeability, electromechanical equipment failure and the like exist, the mining intrinsic safety type broadcasting substation is used as key equipment for underground communication and emergency command, and bears the heavy duty of daily production scheduling instruction transmission and personnel evacuation guidance in emergency, and along with the deep advancement of intelligent construction of intelligent mines and coal mines, the urgent requirements of real-time perception, intelligent analysis and quick response on underground complex operation environments are difficult to be met by the traditional passive voice broadcasting. The existing mining broadcasting system focuses on stability and coverage of signal transmission, the early warning function of the mining broadcasting system is usually triggered by a switching value signal of an external sensor, and the mining broadcasting system lacks of deep autonomous analysis capability on environment audio and video data. In an underground actual scene, huge background noise and continuous heat radiation are generated when heavy machinery such as a coal cutter, a heading machine and the like run, the existing monitoring means often adopt audio frequency decibel detection or infrared highest temperature detection based on a single threshold value, the simple threshold value judgment logic is extremely easy to be interfered by the normal running state of equipment, for example, high noise and high temperature of the equipment during normal high-load running are often misjudged as faults or fire precursors, and aperiodic abnormal sound or local tiny temperature rise generated by the abrasion of bearings in the equipment is often ignored because the values are not out of limits, so that the broadcasting system has frequent false alarm or missing alarm conditions and the efficacy of linkage early warning cannot be exerted. Disclosure of Invention In order to solve the technical problems of high false alarm rate and delayed response of single feature detection under the complex working condition, the invention provides the following schemes. In a first aspect, the invention provides a mining broadcasting linkage early warning method based on environment awareness, which comprises the steps of collecting an infrared thermal image sequence and an audio frame sequence of a mining environment by a mining broadcasting substation, carrying out frequency domain analysis on any frame of audio, obtaining an energy spectrum of each frequency point and a frequency spectrum entropy of the frame of audio, determining an audio abnormal sound index according to the deviation degree of the frequency spectrum entropy relative to a preset normal frequency spectrum entropy and the discrete condition of the energy spectrum of the frequency point, obtaining the heat source divergence set of any frame of infrared thermal image and the area of a high Wen Liantong area, determining a thermal abnormal diffusion index according to the divergence change of the heat source divergence set and the area change rate of the high Wen Liantong area, determining a comprehensive risk linkage coefficient according to the product of the audio abnormal sound index and the thermal abnormal diffusion index and the association degree between the two types of index sequences in a reference frame, and driving a broadcasting substation chip to execute corresponding voice warning according to the comprehensive risk linkage coefficient matching preset risk level threshold. According to the method, the infrared thermal image and the audio frame sequence are collected, the frequency spectrum entropy deviation and the energy spectrum dispersion are fused to construct an audio abnormal sound index, the heat source dispersion and the regional area change are combined to construct a thermal abnormal diffusion index, the time sequence association degree of the two indexes is utilized to generate a comprehensive risk linkage coefficient, normal high-load operation and real fault states of mine equipment can be effectively distinguished, the problem of high false alarm rate of single threshold detection is solved, and accurate sensing and quick linkage early warning of underground complex environments are realized. Preferably, the frequency domain analysis is performed on any frame of audio to obtain an energy spectrum of each frequency point and a spectrum entropy of the frame of audio, and the method comprises the steps of performing fast Fourier transform on any