Mine ventilation safety intelligent monitoring method

Abstract

The present application relates to the field of mine, especially to a kind of mine ventilation safety intelligent monitoring method. First, based on the mine information obtained, construct graph structure, and the spatial distribution of monitoring point in graph structure is optimized by monitoring point coverage efficiency function, output optimal monitoring point set;Based on the optimal monitoring point set, sensor is laid out;Then, based on the ventilation parameter collected by sensor, generate state feature matrix;Based on state feature matrix, construct adjacency matrix;Based on state feature matrix and adjacency matrix, calculate the anomaly probability of monitoring point;Based on anomaly probability, get early warning level;Based on anomaly probability, the three-dimensional coordinates of hidden danger position are calculated by gradient guiding air flow propagation positioning algorithm;Finally, based on early warning level and the three-dimensional coordinates of hidden danger position, execute hierarchical automation processing and response mechanism. The technical problems of monitoring coverage not comprehensive, poor real-time data acquisition, insufficient anomaly recognition capability and delayed early warning response are solved.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of mines, in particular to a mine ventilation safety intelligent monitoring method. BACKGROUND

[0002] The mine ventilation system is one of the core guarantees for coal mine safety production. Its main function is to provide sufficient fresh air for underground work by reasonable airflow organization and control, and to remove harmful gases, dust and heat to ensure the safety of miners and the normal operation of production equipment. With the increase of coal mining depth and the expansion of production scale, the environment faced by the mine ventilation system is more complex, and safety hazards are increasingly prominent. Traditional mine ventilation monitoring mainly relies on manual inspection, fixed sensors and simple automation equipment, which has been difficult to meet the needs of modern mine development in efficiency, safety and intelligence. The rise of mine ventilation safety intelligent monitoring method is the result of the deep integration of modern information technology and coal mine safety production needs. Through the application of technologies such as Internet of Things, big data and artificial intelligence, intelligent monitoring methods can significantly improve the safety and efficiency of the ventilation system, providing strong technical support for coal mine safety production. With the continuous progress of technology, future mine ventilation safety monitoring will be more accurate and efficient, laying a solid foundation for realizing the goal of "unmanned mine" and "intelligent mine".

[0003] However, the existing mine ventilation safety intelligent monitoring method still has problems such as incomplete monitoring coverage, poor real-time data collection, insufficient abnormal identification capability, and untimely warning response. SUMMARY

[0004] The present application provides a mine ventilation safety intelligent monitoring method to solve the problems of lack of scientific basis in traditional monitoring point layout, difficulty in capturing ventilation parameter changes comprehensively, reliance on fixed threshold or simple statistics, difficulty in identifying complex abnormal patterns, long traditional warning response time, lack of grading mechanism and precise positioning.

[0005] A mine ventilation safety intelligent monitoring method according to the present application comprises the following steps:

[0006] S1. Based on the obtained mine information, a graph structure is constructed, and the spatial distribution of the monitoring points in the graph structure is optimized through a monitoring point coverage efficiency function to output an optimal monitoring point set. Based on the optimal monitoring point set, sensors are laid out;

[0007] S2. The sensors laid out based on the optimal monitoring point set collect ventilation parameters to generate a state feature matrix. Based on the state feature matrix, an adjacency matrix is constructed. Based on the state feature matrix and the adjacency matrix, the abnormal probability of the monitoring points is calculated. Based on the abnormal probability, the warning level is obtained;

[0008] S3. Based on the abnormal probability, the three-dimensional coordinates of the hidden danger position are calculated by a gradient-guided wind flow propagation positioning algorithm; based on the early warning level and the three-dimensional coordinates of the hidden danger position, a hierarchical automatic processing and response mechanism is executed.

[0009] Preferably, the S1 specifically comprises:

[0010] The three-dimensional Euclidean distance between the monitoring point and the neighbor node is calculated to obtain a distance attenuation factor.

[0011] Preferably, the S1 specifically comprises:

[0012] The risk weight and the ventilation importance of the neighbor node are introduced, and the coverage efficiency of the monitoring point is calculated in combination with the distance attenuation factor.

[0013] Preferably, the S1 specifically comprises:

[0014] The monitoring point that maximizes the coverage efficiency sum and satisfies the monitoring point number constraint and the minimum distance constraint is selected as the optimal monitoring point set.

[0015] Preferably, the S2 specifically comprises:

[0016] Based on the state feature matrix and the adjacency matrix, the hidden state of the monitoring point is generated; based on the hidden state of the monitoring point, the abnormal probability of the monitoring point is calculated.

[0017] Preferably, the S2 specifically comprises:

[0018] The abnormal probability of the monitoring point is compared with an abnormal threshold value; when the abnormal probability of the monitoring point exceeds a first abnormal threshold value, that is, the monitoring is abnormal, the early warning level is output.

[0019] Preferably, the S3 specifically comprises:

[0020] Based on the abnormal probability of the neighbor node of the monitoring point, the abnormal propagation gradient vector of the monitoring point is calculated.

[0021] Preferably, the S3 specifically comprises:

[0022] The modulus value of the abnormal propagation gradient vector is calculated, and the monitoring point with the largest modulus value is selected as the starting point of the abnormal propagation.

[0023] Preferably, the S3 specifically comprises:

[0024] Based on the starting point of the abnormal propagation and the abnormal propagation gradient vector, the three-dimensional coordinates of the hidden danger position are calculated.

[0025] The technical scheme of the present application has the following advantages:

[0026] 1. The spatial distribution of monitoring points has been optimized, solving the problem of incomplete coverage of traditional monitoring points and forming a three-dimensional intelligent sensing network. This ensures comprehensive collection of ventilation parameters and effectively reduces the probability of missing detection of hidden dangers such as the accumulation of harmful gases and abnormal ventilation resistance by covering key areas and potential risk points.

[0027] 2. By integrating spatial correlation and time series information between monitoring points using graph neural networks, the accuracy of abnormal state identification has been significantly improved, the problem of insufficient identification capability has been solved, valuable time has been gained for handling hidden dangers, and the ability to prevent accidents has been enhanced.

[0028] 3. The location of potential hazards can be directly located by using the gradient vector of anomaly propagation, which has high positioning accuracy and provides precise spatial guidance for subsequent targeted treatment.

[0029] 4. Based on the three-dimensional coordinates of the early warning level and the location of potential hazards, a graded automatic processing and response mechanism is implemented to reduce the risk of ventilation accidents and significantly improve the level of mine safety. Attached Figure Description

[0030] Figure 1 This is a flowchart of a mine ventilation safety intelligent monitoring method according to the present invention. Detailed Implementation

[0031] To further illustrate the technical means and effects adopted by the present invention to achieve its intended purpose, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0032] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains.

[0033] The following description, in conjunction with the accompanying drawings, details the specific scheme of the intelligent monitoring method for mine ventilation safety provided by the present invention.

[0034] See attached document Figure 1 The diagram illustrates a flowchart of an intelligent monitoring method for mine ventilation safety provided by an embodiment of the present invention. The method includes the following steps:

[0035] S1. Based on the acquired mine information, construct a graph structure and optimize the spatial distribution of monitoring points in the graph structure through the monitoring point coverage efficiency function to output the optimal monitoring point set; based on the optimal monitoring point set, deploy sensors.

[0036] Mine information is obtained by using mine data. The mine data includes mine geological exploration data, roadway design drawings and mine ventilation system operation records; the mine information includes mine physical structure and ventilation path information such as roadway coordinates, length, cross-sectional area and fan position; based on the obtained mine information, a graph structure is constructed wherein, is a node, that is, a monitoring point, such as a roadway intersection, is an edge, that is, a roadway.

[0037] In order to optimize the spatial distribution of the monitoring points, a monitoring point coverage efficiency function is defined to calculate the monitoring coverage efficiency of the monitoring points. The monitoring point coverage efficiency function introduces a distance attenuation factor, which is determined by calculating the three-dimensional Euclidean distance between the monitoring point and the neighbor nodes, is used to simulate the physical characteristics that the sensor monitoring range decreases with the increase of distance, adopts an exponential attenuation form to ensure that the area close to the monitoring point has a higher coverage weight; the monitoring coverage efficiency not only considers the Euclidean distance, but also combines the risk weight and ventilation importance of the neighbor nodes, the more frequent the area with more historical accidents, the higher the risk weight of the node, and the higher the ventilation importance of the node in the main air flow channel, which will contribute more to the monitoring coverage efficiency, so as to be covered by the monitoring point in priority; the main air flow channel is a channel with strategic significance in the mine ventilation system, which usually connects the main air inlet well, the main air outlet well and the main mining area, and is responsible for the transportation of the main air flow. The coverage efficiency of the monitoring point is obtained by weighted sum of the risk weight and ventilation importance of the neighbor nodes, which reflects the monitoring ability of the monitoring point, and the higher the coverage efficiency, the more effectively the monitoring point can cover the high-risk and high-importance areas.

[0038] The formula expression of the monitoring point coverage efficiency function is as follows:

[0039]

[0040] wherein, represents the coverage efficiency of the i-th monitoring point; represents the neighbor node set of the i-th monitoring point, the neighbor node represents the node directly connected with the i-th monitoring point through the edge; represents the risk weight of the j-th neighbor node, that is, the risk weight of the i-th monitoring point, which is used to quantify the safety importance of the i-th monitoring point, the higher the risk weight, the higher the monitoring priority, based on the historical accident data obtained from the mine ventilation system operation records, such as gas explosion records, the risk weight is determined by statistical analysis, such as by frequency analysis method, ;​​​​​​ denotes the distance attenuation factor, which is determined based on the three-dimensional Euclidean distance between the first monitoring point and the second monitoring point determined by an exponential decay function, reflecting the spatial decay of sensor monitoring capability; denotes the three-dimensional Euclidean distance between the first monitoring point and the second monitoring point denotes the distance attenuation coefficient, used to control the decay rate of the sensor monitoring range, determined based on experimental measurement method, which is to deploy sensors with known performance, such as wind speed and gas concentration sensors, in the actual mine environment, measure the signal strength or monitoring effect of the sensors at different distances, record the monitoring data of the sensors at different distances, and fit the exponential decay function to determine the value of denotes the ventilation importance of the first monitoring point, obtained by data mining techniques such as principal component analysis or clustering analysis, analyzing historical air flow data obtained from mine ventilation system operation records.

[0041] The optimization goal is to select a set of monitoring points as the optimal monitoring point set, so that the total coverage efficiency of the entire mine ventilation system is maximized, while satisfying the following constraints:

[0042] First, the monitoring point number constraint ensures that the number of monitoring points does not exceed the maximum number of monitoring points based on the existing multi-objective optimization algorithm, to control the cost;

[0043] Second, the minimum distance constraint ensures that the three-dimensional Euclidean distance between any two monitoring points is not less than the distance threshold set based on spatial statistical analysis method, to avoid excessive concentration of monitoring points and ensure uniform distribution of monitoring points, thereby improving the accuracy of subsequent hidden danger positioning.

[0044] A heuristic optimization method such as genetic algorithm or simulated annealing algorithm is used to solve the optimal monitoring point set wherein, denotes the three-dimensional coordinates of the first monitoring point, denotes the number of monitoring points, .

[0045] ​​​​​​​By optimizing the spatial distribution of monitoring points and rationally deploying different types of sensors, including those for wind speed, wind pressure, harmful gas concentration, temperature, and humidity, the monitoring coverage and accuracy of the mine ventilation system have been significantly improved, the sensor deployment cost has been reduced, and key monitoring of high-risk areas has been ensured, thereby enhancing the ability to prevent ventilation accidents.

[0046] S2. Collect ventilation parameters using sensors deployed based on the optimal set of monitoring points, and generate a state feature matrix; construct an adjacency matrix based on the state feature matrix; calculate the anomaly probability of the monitoring points based on the state feature matrix and the adjacency matrix; and obtain the warning level based on the anomaly probability.

[0047] Ventilation parameters, including wind speed, wind pressure, and concentration of harmful gases, are collected in real time by sensors deployed based on an optimal set of monitoring points, forming a state feature matrix. ,in, For the first Each monitoring point in time The parameter vector for each monitoring point contains real-time measurements of ventilation parameters in different dimensions, such as wind speed, wind pressure, and concentration of harmful gases. , Indicates the first The monitoring point number The actual measured values ​​of the ventilation parameters at each monitoring point reflect the ventilation status at that specific time point and are the basis for anomaly identification. This indicates the number of ventilation parameters, i.e., the number of dimensions; This indicates transpose.

[0048] Construct an adjacency matrix The adjacency matrix is ​​used to represent the relationship between monitoring points. Each element of the adjacency matrix comprehensively considers the spatial distance and parameter similarity between the monitoring points. The spatial distance also uses a distance attenuation factor to simulate the physical characteristic that the sensor monitoring range weakens with increasing distance. The parameter similarity is calculated by comparing the parameter vectors of two monitoring points to calculate the degree of similarity on different ventilation parameters, reflecting the dynamic correlation of airflow. The method of comprehensively considering spatial distance and parameter similarity enables the adjacency matrix to capture the physical and functional correlation between monitoring points during ventilation. For example, monitoring points that are close to each other or have similar airflow characteristics will have higher connection weights in the adjacency matrix.

[0049] The formula for calculating each element of the adjacency matrix is:

[0050]

[0051] in, Represents the elements of the adjacency matrix, used to reflect the first... The monitoring point and the first Spatial distance and parameter similarity correlation strength between monitoring points; Indicates the first The monitoring point and the first The spatial distance between monitoring points, i.e., the distance attenuation factor; To represent parameter similarity, the first parameter is calculated. The monitoring point and the first Each monitoring point in time The square norm of the parameter vector difference is used, and an exponential decay function is introduced to quantify the similarity of ventilation parameters between monitoring points, reflecting the correlation between airflow and abnormal propagation. The calculation formula is as follows: , Indicates the first The monitoring point and the first Each monitoring point in time The square norm of the difference between the parameter vectors; This represents a similarity adjustment parameter used to quantify the impact of differences in ventilation parameters on parameter similarity. It is obtained based on statistical analysis, which calculates the square norm of the parameter vector difference by analyzing the distribution of historical ventilation parameter data obtained from the mine ventilation system's operation records. Based on the statistical properties, such as mean, variance, or median, the similarity adjustment parameter is set to a multiple of the mean or variance of the squared norm of the parameter vector difference. .

[0052] The state feature matrix and adjacency matrix The data is input into a graph neural network, which iteratively calculates and aggregates the parameter vectors of each monitoring point with those of neighboring monitoring points. It also integrates data from historical time steps to form the hidden state of each monitoring point. The hidden state not only includes the ventilation parameters of the current monitoring point, but also captures the dynamic influence of neighboring monitoring points through an adjacency matrix, reflecting the spatial propagation and temporal evolution of anomalies.

[0053] The formula for generating the hidden state of monitoring points is:

[0054]

[0055] in, Indicates the first Each monitoring point in time The hidden state; Represents graph neural network operations; Indicates the first Each monitoring point in time The hidden state; The weight matrix representing the graph neural network is obtained by training the graph neural network with historical ventilation parameter data and abnormal event data obtained from the mine ventilation system operation record, and the reference value range is .

[0056] The hidden state of the monitoring point is generated by a fully connected layer and an activation function, such as Sigmoid, to generate the abnormal probability of each monitoring point, and the calculation formula is as follows:

[0057]

[0058] wherein, represents the abnormal probability of the th monitoring point at time , which is used to represent the possibility of the th monitoring point being abnormal; represents an activation function that maps linear output to to generate an abnormal probability value; represents the weight matrix of the output layer of the graph neural network, which maps the hidden state of the th monitoring point to the abnormal probability, and is obtained by training the graph neural network; represents the bias of the output layer of the graph neural network, which is used to adjust the offset of the abnormal probability, and is obtained by training the graph neural network.

[0059] The spatial correlation and time series information between monitoring points are integrated by the graph neural network, overcoming the limitations of isolated analysis of monitoring points in traditional methods.

[0060] Based on the mine safety standard, an abnormal threshold is set, which is compared with the abnormal probability of the monitoring point. If the abnormal probability of the monitoring point exceeds the first abnormal threshold, i.e., the monitoring is abnormal, the warning level is output, and the formula is expressed as follows:

[0061] Warning level determination:

[0062]

[0063] wherein, represents the warning level at time ; , , respectively represent the first, second and third abnormal thresholds, which are determined based on the mine safety standard.

[0064] S3, based on the abnormal probability, the three-dimensional coordinates of the hidden danger position are calculated by the gradient-guided air flow propagation positioning algorithm; based on the warning level and the three-dimensional coordinates of the hidden danger position, the hierarchical automatic processing and response mechanism is executed.

[0065] If an anomaly is detected, the location of the hazard is determined using a gradient-guided airflow propagation localization algorithm.

[0066] The gradient-guided airflow propagation localization algorithm calculates the anomaly propagation gradient vector for each monitoring point. Specifically, it performs a weighted summation of the anomaly probabilities of neighboring nodes, with the weights being the product of the adjacency matrix elements and the cosine of the angle between the airflow direction and the direction between the monitoring point. The direction vector is the spatial vector of the neighboring node relative to the current monitoring point. The anomaly propagation gradient vector represents the spatial propagation trend of the anomaly at the monitoring point, pointing to the direction of the possible source or spread of the anomaly.

[0067] The formula for calculating the anomaly propagation gradient vector is:

[0068]

[0069] in, Indicates the first The gradient vector of anomaly propagation at each monitoring point is used to reflect the direction and intensity of anomaly propagation in space, to locate potential hazards, and to guide the search along the anomaly propagation path. Indicates the first Each monitoring point in time The probability of an anomaly; Indicates weight; Indicates the first The three-dimensional coordinates of each monitoring point; Indicates the first The three-dimensional coordinates of each monitoring point; Indicates the first The monitoring point is relative to the first The direction vector of the nth monitoring point, i.e., the nth The monitoring point is relative to the first The spatial vectors of the monitoring points are used to determine the spatial direction of anomaly propagation, with a magnitude of [missing value]. ; Indicates the first The monitoring point and the first The three-dimensional Euclidean distance between each monitoring point is used to normalize the direction vector; The cosine of the angle between the airflow direction and the direction between points is used to reflect the influence of airflow on the propagation of anomalies. The formula for calculating the angle between the airflow direction and the direction between points is: , Indicates the first The wind direction vectors at each monitoring point are used to provide the wind direction and calculate the direction of abnormal propagation. These vectors are obtained through wind speed sensors and have a modulus of [missing value]. .

[0070] The anomaly propagation gradient vector of all monitoring points is traversed, the modulus value of each anomaly propagation gradient vector is calculated, the greater the modulus value, the more significant the influence of anomaly propagation on the monitoring point, the monitoring point with the largest modulus value is selected as the starting point of anomaly propagation, and the formula is expressed as follows:

[0071]

[0072] wherein, represents the index of the monitoring point with the largest modulus value of the anomaly propagation gradient vector; represents a maximum value selection operation; represents the modulus value of the anomaly propagation gradient vector of the th monitoring point.

[0073] The distance is moved along the direction of the anomaly propagation gradient vector to determine the final hazard location; the moving distance is determined by the wind speed of the monitoring point and the anomaly propagation time estimate value, the wind speed reflects the propagation speed of the anomaly, such as the propagation speed of harmful gas, the wind speed is obtained through the wind speed sensor, and the anomaly propagation time estimate value is obtained based on historical data obtained from the mine ventilation system operation record.

[0074] The three-dimensional coordinates of the hazard location are:

[0075]

[0076] wherein, represents the three-dimensional coordinates of the hazard location; represents the three-dimensional coordinates of the monitoring point with the largest modulus value of the anomaly propagation gradient vector; represents an adjustment coefficient, used to control the step length in the direction of the anomaly propagation gradient vector and adjust the offset distance of the hazard location, determined by an experimental calibration method, the experimental calibration method determines the appropriate value of by performing field experiments in the mine environment, combining sensor data and known anomaly propagation events, such as gas leakage, dust diffusion, etc.; represents the unit vector of the anomaly propagation gradient vector of the th monitoring point; represents the anomaly propagation gradient vector of the th monitoring point; represents the modulus value of the anomaly propagation gradient vector of the th monitoring point; represents the wind speed of the th monitoring point, used to affect the offset distance of the hazard location and reflect the wind flow propagation speed; Anomaly propagation time estimation value, used to reflect the time of anomaly propagation with wind flow, determine the offset of hidden danger location, based on historical anomaly data obtained from mine ventilation system operation records, obtained through regression analysis, such as linear regression analysis; Indicates the moving distance.

[0077] Based on the early warning level and the three-dimensional coordinates of the hidden danger location, a hierarchical automated processing and response mechanism is executed to quickly control the abnormal state in the mine ventilation system and ensure mine safety.

[0078] The hierarchical automated processing and response mechanism is as follows:

[0079] Low-level response: when the early warning level is low early warning, push the early warning information containing the early warning level, the three-dimensional coordinates of the hidden danger location and the recommended measures to the ventilation dispatch center;

[0080] Medium-level response: when the early warning level is medium early warning, on the basis of pushing the early warning information containing the early warning level, the three-dimensional coordinates of the hidden danger location and the recommended measures to the ventilation dispatch center, further, send an emergency notice to the mine safety management layer, and issue a regional safety warning through the mine broadcast system, automatically identify the air door and fan closest to the hidden danger location, execute the pre-set emergency ventilation scheme, increase the local air volume through the mine control system, such as increasing the local air volume by 10-20%;

[0081] High-level response: when the early warning level is high early warning, trigger the highest level of emergency response, send an emergency alarm to all emergency departments through sound and light alarm, short message, APP push and other channels, automatically switch to emergency ventilation mode, increase the whole network air volume by more than 30%, prioritize the ventilation effect of the hidden danger location and the upstream area, close the air door downstream of the hidden danger location to prevent the anomaly from spreading to the main return airway, and start the standby fan system to enhance the ventilation capacity.

[0082] In summary, a mine ventilation safety intelligent monitoring method is completed.

[0083] The order of the embodiments of the application is only for description, and does not represent the advantages and disadvantages of the embodiments. The processes depicted in the drawings do not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are possible or can be advantageous.

[0084] Each embodiment in the specification is described in a progressive manner, and the same or similar parts between each embodiment can be referred to each other, and each embodiment mainly describes the difference from other embodiments.

[0085] The above examples are only used to illustrate the technical solutions of the present application, and are not intended to limit the present application; although the present application has been described in detail with reference to the foregoing examples, those skilled in the art should understand that the technical solutions recorded in the foregoing examples can be modified, or some technical features can be replaced by equivalent features; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present application, and should be included in the protection scope of the present application.

Claims

1. A mine ventilation safety intelligent monitoring method, characterized in that, The method comprises the following steps: S1. Based on the obtained mine information, a graph structure is constructed, and the spatial distribution of the monitoring points in the graph structure is optimized through a monitoring point coverage efficiency function. The three-dimensional Euclidean distance between the monitoring points and the neighbor nodes is calculated to obtain a distance attenuation factor. The risk weight and the ventilation importance of the neighbor nodes are introduced. The coverage efficiency of the monitoring points is calculated in combination with the distance attenuation factor. The monitoring points that maximize the total coverage efficiency and meet the number constraint and the minimum distance constraint of the monitoring points are selected as the optimal monitoring point set. Sensors are arranged based on the optimal monitoring point set; S2. The sensors arranged based on the optimal monitoring point set collect ventilation parameters to generate a state feature matrix; Based on the state feature matrix, an adjacency matrix is constructed; Based on the state feature matrix and the adjacency matrix, the hidden state of the monitoring points is generated; Based on the hidden state of the monitoring points, the abnormal probability of the monitoring points is calculated; Based on the abnormal probability of the monitoring points, the abnormal probability is compared with an abnormal threshold value. When the abnormal probability of the monitoring points exceeds the first abnormal threshold value, i.e., the monitoring is abnormal, the warning level is outputted; S3. In the gradient-guided air flow propagation positioning algorithm, based on the abnormal probability of the neighbor nodes of the monitoring points, the abnormal propagation gradient vector of the monitoring points is calculated. The modulus of the abnormal propagation gradient vector is further calculated. The monitoring point with the largest modulus is selected as the starting point of the abnormal propagation. Based on the starting point of the abnormal propagation and the abnormal propagation gradient vector, the three-dimensional coordinates of the hidden danger position are calculated. Based on the warning level and the three-dimensional coordinates of the hidden danger position, a hierarchical automatic processing and response mechanism is executed.

Images