CN-121985004-A - Mine emergency rescue safety dynamic monitoring management system

CN121985004ACN 121985004 ACN121985004 ACN 121985004ACN-121985004-A

Abstract

The invention relates to the technical field of mine safety monitoring and discloses a mine emergency rescue safety dynamic monitoring management system which comprises a framework module, an anomaly detection module, an information value evaluation module, a resource scheduling module and a network coding and decoding module, wherein the framework module is used for constructing a three-layer fusion system framework, the three-layer fusion system framework comprises an intelligent perception layer, an information evaluation layer and a self-adaptive communication layer, the anomaly detection module is used for realizing multi-mode anomaly detection and threat level evaluation, the information value evaluation module is used for realizing priority evaluation based on information value, the resource scheduling module is used for realizing resource dynamic allocation based on reinforcement learning, and the network coding and decoding module is used for realizing self-adaptive network coding and decoding.

Inventors

QIAO NING
GUO TING
TANG ZILONG
WANG DAIHONG

Assignees

中煤华晋集团有限公司

Dates

Publication Date: 20260505
Application Date: 20260207

Claims (10)

1. The utility model provides a mine emergency rescue safety dynamic monitoring management system which characterized in that includes: the architecture module is used for constructing a three-layer fusion system architecture, wherein the three-layer fusion system architecture comprises an intelligent perception layer, an information evaluation layer and a self-adaptive communication layer; The abnormality detection module is used for realizing multi-mode abnormality detection and threat level assessment through the intelligent perception layer; The information value evaluation module is used for realizing priority evaluation based on the information value through the information evaluation layer; the resource scheduling module realizes the dynamic allocation of the resources based on reinforcement learning through the self-adaptive communication layer; And the network coding and decoding module is used for realizing self-adaptive network coding and decoding based on the resource allocation result of the self-adaptive communication layer.
2. The mine emergency rescue safety dynamic monitoring and management system according to claim 1, wherein the step of constructing a three-layer fusion system architecture comprises: Constructing an intelligent sensing layer based on multi-mode sensing fusion, wherein the intelligent sensing layer is responsible for collecting multi-source heterogeneous data from a mine environment, detecting potential abnormal events in real time through a deep learning model and evaluating threat levels; An information evaluation layer based on an information value theory is constructed, the layer is responsible for evaluating the value of monitoring information on the current decision, and the information transmission priority is calculated; and constructing an adaptive communication layer based on dynamic resource allocation, wherein the layer guarantees key information transmission by adopting a multipath cooperative transmission and coding adaptive technology according to the information priority and the communication resource condition.
3. The mine emergency rescue safety dynamic monitoring and management system according to claim 2, wherein the constructing the intelligent perception layer based on multi-mode sensing fusion comprises: setting a sensor network deployment and data acquisition module, and deploying multiple types of sensors in a mine key area; setting a data preprocessing module, and carrying out normalization, denoising and outlier processing on the acquired original sensing data; setting a multi-modal feature extraction module, and extracting features from different modal data by using a convolutional neural network and a long-term and short-term memory network; an anomaly detection and threat assessment module is arranged, and based on the extracted multi-mode characteristics, the type of the anomaly event and the threat level score are output.
4. The mine emergency rescue safety dynamic monitoring and management system according to claim 2, wherein the construction of the information evaluation layer based on the information value theory comprises the following steps: Setting an information cost function construction module and defining an information cost function Wherein In order to provide the information content with a content, In order to be in the state of the system, Time is; Setting a risk function model and constructing a risk function Evaluating the information given System status under conditions Risk level of (2); setting a timeliness evaluation model, and constructing a timeliness function Evaluating a loss of value of the information over time; Setting an information priority calculating module, and calculating information transmission priority by combining threat level, risk function and timeliness function to generate a transmission queue.
5. The mine emergency rescue safety dynamic monitoring and management system according to claim 2, wherein the constructing the adaptive communication layer based on dynamic allocation of resources comprises: setting a communication resource sensing module, and monitoring the communication network conditions in real time, wherein the communication network conditions comprise a bandwidth range, channel quality and link state; Setting a resource allocation strategy module, and constructing a resource dynamic allocation strategy based on a reinforcement learning algorithm; setting a multi-path routing module and constructing a network topology aware multi-path routing algorithm; and setting an adaptive coding module, and adaptively adjusting network coding parameters according to the communication quality and the information importance.
6. The mine emergency rescue safety dynamic monitoring and management system according to claim 1, wherein the steps of implementing multi-mode anomaly detection and threat level assessment through the intelligent perception layer comprise: preprocessing the collected multi-source heterogeneous sensing data; Extracting features from the preprocessed multi-modal data by using a deep learning model and fusing the features; Constructing an anomaly detection model based on deep learning, and training through historical data; Based on the detected anomalies, a multi-dimensional assessment of threat levels is achieved.
7. The mine emergency rescue safety dynamic monitoring and management system of claim 6, wherein the preprocessing of the collected multi-source heterogeneous sensing data comprises: Performing time stamp alignment on the distributed sensor data; Denoising the original data; converting the sensing data with different types and different dimensions into a unified interval through a normalization method; The problem of missing sensing data is processed.
8. The mine emergency rescue safety dynamic monitoring and management system according to claim 6, wherein the extracting features from the preprocessed multi-modal data by using the deep learning model and fusing the features comprises: extracting time dimension characteristics from time sequence data of various sensors; extracting the correlation characteristics of the space dimension from the sensor network data with the space distribution characteristic by using a graph rolling network; and adopting an attention mechanism to adaptively fuse different modal characteristics, and dynamically adjusting the weight of each modal characteristic.
9. The mine emergency rescue safety dynamic monitoring and management system according to claim 1, wherein the step of implementing priority assessment based on information value by the information assessment layer comprises: implementing information cost function The information cost function quantifies the decision value of the evaluation information in the current system state; calculating the value of each information and determining the priority based on the realized cost function; and an adaptive learning mechanism of the value evaluation model is realized, and the information value evaluation accuracy is continuously optimized.
10. A computer storage medium comprising a memory and one or more processors, the memory having executable code stored therein, the one or more processors, when executing the executable code, operative to implement the mine emergency rescue safety dynamic monitoring management system of any one of claims 1-9.

Description

Mine emergency rescue safety dynamic monitoring management system Technical Field The invention relates to the technical field of mine safety monitoring, in particular to a mine emergency rescue safety dynamic monitoring management system. Background Along with the increasing of the mining depth and scale, the mine safety risk is also increasingly prominent, and mine disasters, particularly serious disasters such as coal mine gas explosion, rock burst, flood and the like, occur. In order to reduce disaster risk and improve mine safety production level and emergency rescue capability, the mine safety dynamic monitoring and early warning system is widely applied. The traditional mine safety dynamic monitoring system mainly comprises three parts of sensor acquisition, data transmission and monitoring center processing so as to realize monitoring and abnormal alarm of mine environment parameters, however, the systems have obvious technical limitations when facing complex and changeable environments and disasters of mines, on one hand, the systems usually adopt a single type of sensor for monitoring, and are difficult to comprehensively capture the complex changes of the mine environments, and on the other hand, the data transmission and the abnormal detection are regarded as mutually independent links and lack organic cooperation. In mine disaster scenes, communication conditions are often severely restricted, a large amount of monitoring data are difficult to transmit in time, so that key early warning information is delayed or lost, a transmission mechanism with fixed priority is generally adopted in the prior art, and a resource allocation strategy cannot be dynamically adjusted according to real-time disaster conditions, so that real key early warning information cannot be transmitted to a decision maker in time in emergency. The existing researches begin to pay attention to the problem of data fusion in a mine safety monitoring system, for example, researches are carried out to provide a mine environment monitoring method based on multi-source data fusion, abnormal detection precision is improved by integrating different types of sensor data, and technical schemes are also carried out to try to improve the reliability of a mine emergency communication system, such as adopting strategies of mesh network topology, redundant transmission and the like to improve the communication reliability, most of the researches still consider data acquisition and transmission as independent technical problems, and the research lacks an optimization thought from the whole system, particularly under the condition of limited communication resources, how to ensure the preferential transmission of the most critical information under the condition of disaster environment, and still has the urgent problem to be solved. In summary, the existing mine safety dynamic monitoring system has the technical problems of information acquisition and transmission fracture, fixed communication resource allocation, weak capability of coping with extreme communication environment, missing information value evaluation mechanism and the like, and is difficult to meet urgent requirements for timely and reliable information transmission in mine emergency rescue. Therefore, an innovative technical scheme is needed, which can break through the technical limit of sensing and communication, construct a closed-loop system from anomaly detection to information transmission, and realize the full-flow intelligent management of 'discovery-understanding-transmission-response' of disaster information. Disclosure of Invention The invention provides a mine emergency rescue safety dynamic monitoring management system, which solves the technical problems of information acquisition and transmission fracture, fixed communication resource allocation, weak capability of coping with extreme communication environment and missing of an information value evaluation mechanism in the related technology. The invention provides a mine emergency rescue safety dynamic monitoring management system, which comprises: the architecture module is used for constructing a three-layer fusion system architecture, wherein the three-layer fusion system architecture comprises an intelligent perception layer, an information evaluation layer and a self-adaptive communication layer; The abnormality detection module is used for realizing multi-mode abnormality detection and threat level assessment through the intelligent perception layer; The information value evaluation module is used for realizing priority evaluation based on the information value through the information evaluation layer; the resource scheduling module realizes the dynamic allocation of the resources based on reinforcement learning through the self-adaptive communication layer; And the network coding and decoding module is used for realizing self-adaptive network coding and decoding based on the resource allocation result of