CN-122014347-A - Mine accident intelligent detection collaborative management early warning method

CN122014347ACN 122014347 ACN122014347 ACN 122014347ACN-122014347-A

Abstract

The invention discloses a mine accident intelligent detection collaborative management early warning method, which relates to the technical field of mine accident early warning methods and specifically comprises the following steps of planning and triggering a multi-source collaborative inspection task, receiving full-quantity data of inspection, automatically associating data sources in the same time period and the same area by a system, carrying out multi-source data fusion analysis and risk judgment, carrying out grading early warning release and collaborative management response, triggering a blocking network sensing layer, carrying out air accurate check and change detection, carrying out dynamic risk assessment and light early warning zone demarcation, carrying out collaborative emergency response and closed loop management, and carrying out intelligent mine accident detection collaborative management early warning method.

Inventors

ZHANG JIANQIANG
NING SHUZHENG
ZHANG DEGAO
XUE HUANHUAN
NIU FANGPING
HOU YUEHUA
ZHOU YAO

Assignees

中国煤炭地质总局勘查研究总院

Dates

Publication Date: 20260512
Application Date: 20251230

Claims (10)

1. The intelligent detection collaborative management early warning method for the mine accidents is characterized by comprising the following steps of: step one, multi-source collaborative inspection task planning and triggering; step two, intelligent perception of the unmanned aerial vehicle and synchronous acquisition of multi-mode data; Step three, receiving the total data of the inspection, and simultaneously, automatically associating data sources in the same time period and the same area with the system, and carrying out multi-source data fusion analysis and risk judgment; Step four, grading early warning release and collaborative management response; Triggering a blocking net sensing layer; Step six, in-air accurate checking and change detection; Step seven, defining a dynamic risk assessment and lamplight early warning area; and step eight, cooperative emergency response and closed-loop management.
2. The mine accident intelligent detection collaborative management early warning method is characterized in that in the first step, collaborative inspection task planning generates daily and periodic inspection plans according to safety management regulations, a ground microseismic sensor monitors vibration abnormality, a slope radar finds tiny displacement super threshold value, gas concentration monitoring point data abnormality and receives manual report hidden danger, and a data center control system automatically acquires weather information to adjust special inspection according to the weather information.
3. The mine accident intelligent detection collaborative management early warning method is characterized in that in the second step, an unmanned aerial vehicle automatically flies along a preset route, night obstacle avoidance is achieved by using an airborne millimeter wave radar and a vision system, and in the second step, multi-mode data synchronous acquisition comprises high-definition night vision video streaming, infrared thermal image data and laser radar point cloud.
4. The mine accident intelligent detection collaborative management early warning method is characterized in that in the third step, the information receiving process is that an unmanned plane infrared finding area is abnormal in temperature, then fixed temperature sensor history data near the area is called, a temperature changing trend is checked, an unmanned plane image finding slope has new cracks, the new cracks are compared with displacement increment calculated by laser radar point cloud and slope monitoring radar data, severity and activity of the cracks are verified, visible light and infrared images are analyzed deeply, defects of protection facilities are identified through an identification model, point clouds which are scanned this time and last time are automatically compared through a point cloud change detection model, a high-precision differential model is generated, settlement, landslide and material accumulation change overrun areas are identified, various identified hidden danger features are combined with environment data and production activity data through a comprehensive risk assessment model, a risk assessment model is input, and risk grade and probability are output.
5. The mine accident intelligent detection collaborative management early warning method is characterized in that after the risk judgment is completed in the third step, an analysis report is generated, an early warning rule engine is started according to preset rules, a large screen of a command center, related area broadcasting and a responsible person mobile phone APP trigger strong warning simultaneously when high early warning occurs in the fourth step, early warning information is automatically associated with hidden danger details, field video streams and treatment plans when low early warning occurs, a maintenance and review work order is automatically generated, and the maintenance and review work order is distributed to mobile terminals of responsible department personnel.
6. The mine accident intelligent detection collaborative management early warning method is characterized in that an intelligent interception net is arranged below a slope potential instability area, multimode sensor nodes are arranged on the interception net according to a matrix, sensors integrate stress, vibration, inclination angle sensing and wireless transmission modules, the sensor network completes networking, a signal is accessed into an early warning center, and unique IDs and accurate coordinates of each sensor are marked in a digital twin model.
7. The mine accident intelligent detection collaborative management early warning method is characterized in that step five is characterized in that normal monitoring is carried out during operation, state data are periodically reported by sensors to form a background baseline, when an interception net is impacted by landslide and falling rocks, a group of sensors trigger super-threshold alarms at the same time and still keep communication, an early warning center automatically circles a suspected impact source area on a map by utilizing a triangular positioning or shock wave propagation algorithm according to the time-space relation of a sensor signal sequence, when part of sensor signals are suddenly and completely interrupted, the early warning center directly judges the physical area where the signal loss sensor is located as a direct dangerous area, and in both cases, the system automatically generates unmanned aerial vehicle emergency checking tasks with highest priority.
8. The mine accident intelligent detection collaborative management early warning method according to claim 1 is characterized in that in the sixth step, the unmanned aerial vehicle receives critical alarms from an intelligent interception net, the unmanned aerial vehicle does not execute conventional airlines, but flies directly into an alarm area, the height is reduced to perform approaching reconnaissance, multi-angle high-definition photographing, thermal imaging scanning and laser radar rapid scanning are performed on a suspected impact source area and a direct dangerous area, a picture is returned to a command center in real time, and the scale, the boundary, the stability and the threat degree of a landslide are analyzed in an auxiliary mode.
9. The intelligent detection collaborative management early warning method for mine accidents according to claim 1 is characterized in that in the step seven, an early warning center synthesizes interception net data and unmanned aerial vehicle check data, three layers of areas, namely a red high-risk area, a yellow warning area and a green safety gathering area, are dynamically generated on a three-dimensional map, a system automatically dispatches a special warning unmanned aerial vehicle to fly to a site airspace, and the unmanned aerial vehicle projects clear red and yellow dangerous area facula boundaries to the ground by using a strong light projection lamp in the air according to the geographic coordinates of the received areas, and projects a green set mark in the green area.
10. The intelligent detection collaborative management early warning method for mine accidents according to claim 1 is characterized in that a system in the eighth step automatically locks personnel positioning cards in red and yellow areas, sends forced evacuation alarms to handheld terminals of the personnel positioning cards, can automatically control power-off of operation equipment in the areas, and a command center commands emergency personnel to go to a green safety gathering area and plan a safe investigation or disposal path on a three-dimensional situation map, an unmanned plane continuously monitors the scene, tracks personnel evacuation conditions and slope stable states, each step of the disposal process is recorded in the system, and after the event is finished, all data are packed into cases for optimizing alarm thresholds of an interception network, an AI identification model and an emergency response plan.

Description

Mine accident intelligent detection collaborative management early warning method Technical Field The invention relates to the technical field of mine accident early warning methods, in particular to an intelligent detection collaborative management early warning method for mine accidents. Background The existing mine safety early warning technology focuses on single monitoring means or local risk identification, and multi-source information collaborative fusion and dynamic closed-loop response are difficult to achieve. In the prior art, chinese patent with the authority bulletin number of CN117036082A discloses an intelligent mine management system and method, and the system performs feature extraction based on a D-S evidence theory by collecting mine environment information and personnel physiological information, and cooperatively calculates accident probability by utilizing a double-neural network model so as to realize hierarchical early warning. Although the technology introduces a data fusion and model cooperation mechanism, the early warning process is still mainly dependent on a ground sensor network which is fixedly deployed, lacks the capability of real-time three-dimensional perception and quick response to sudden and local risks, and particularly has obvious defects in the aspects of monitoring and emergency treatment of geological disasters such as side slope instability, falling rocks and the like. In the actual working environment of mines, accidents such as slope instability, landslide, falling rocks and the like have the characteristics of strong burst property, wide influence range, high risk and the like, and the omnibearing and high-precision risk identification and positioning are difficult to realize only by a ground sensor. In addition, the existing system lacks an effective field visual guiding and collaborative emergency mechanism after early warning release, and particularly has low efficiency of personnel evacuation and field management and control at night or under low visibility conditions, so that the full-flow closed-loop management from early warning to treatment is difficult to realize. Therefore, a mine accident intelligent detection collaborative management early warning method which can integrate air and ground monitoring resources, realize multi-mode data fusion analysis and has intelligent triggering and dynamic response capability is needed, so that the instantaneity, the accuracy and the systematicness of mine safety management are improved. Disclosure of Invention The invention aims to provide an intelligent detection collaborative management early warning method for mine accidents, which aims to solve the problems in the background technology. In order to achieve the purpose, the invention provides the following technical scheme that the mine accident intelligent detection collaborative management early warning method specifically comprises the following steps: step one, multi-source collaborative inspection task planning and triggering; step two, intelligent perception of the unmanned aerial vehicle and synchronous acquisition of multi-mode data; Step three, receiving the total data of the inspection, and simultaneously, automatically associating data sources in the same time period and the same area with the system, and carrying out multi-source data fusion analysis and risk judgment; Step four, grading early warning release and collaborative management response; Triggering a blocking net sensing layer; Step six, in-air accurate checking and change detection; Step seven, defining a dynamic risk assessment and lamplight early warning area; and step eight, cooperative emergency response and closed-loop management. Preferably, in the first step, a daily and periodic inspection plan is generated according to a safety management rule by cooperating with the inspection task plan, the ground microseismic sensor monitors that the vibration is abnormal, the slope radar finds out that the tiny displacement exceeds a threshold value, the gas concentration monitoring point data is abnormal and receives the hidden trouble of manual report, and the data center control system automatically acquires meteorological information to adjust special inspection according to the meteorological information. Preferably, the unmanned aerial vehicle automatically flies along a preset route in the second step, night obstacle avoidance is realized by using an airborne millimeter wave radar and a vision system, and the multi-mode data synchronous acquisition in the second step comprises high-definition night vision video streaming, infrared thermal image data and laser radar point clouds. Preferably, in the third step, the information receiving process is that the temperature of an unmanned aerial vehicle infrared finding area is abnormal, then the history data of a fixed temperature sensor near the area is called, the temperature change trend is checked, the unmanned aerial vehicle image find