Underground emergency command system based on improved expert system

Abstract

The invention belongs to the field of mine emergency command, and particularly relates to an under-mine emergency command system based on an improved expert system. Comprising an accident analysis module, an accident type analysis module, an under-mine personnel command module and a rescue personnel command module. The accident analysis module performs accident analysis and risk assessment by using a sensor and historical data, and provides a report; the accident type analysis module analyzes the accident type by using a classification algorithm and generates an emergency response scheme; the under-mine personnel command module plans an optimal evacuation route by adopting improved A * and Dijkstra algorithms, and performs scheduling and resource optimization by utilizing genetic and particle swarm optimization algorithms; the rescue worker command module plans a rescue route through a minimum risk path algorithm and Monte Carlo simulation, and selects a path with the minimum risk. The system can improve mine emergency command efficiency and safety, and guarantee miner life safety and mine production continuity.

Description

Technical Field

[0001] The invention belongs to the field of mine emergency command, and in particular relates to an underground mine emergency command system based on an improved expert system. Background Art

[0002] As mine operations become more in-depth and complex, mine accidents occur frequently, including gas explosions, fires, landslides and other safety accidents. These accidents not only seriously threaten the lives of miners, but also cause huge economic losses and social impacts. The traditional underground mine emergency command system mainly relies on manual decision-making, which has problems such as slow response speed, incomplete information acquisition, and low decision-making accuracy. It cannot meet the needs of fast, safe and efficient emergency response.

[0003] Existing emergency command systems often lack intelligent accident analysis and prediction functions, and are unable to accurately analyze the causes and types of accidents in real time, resulting in delayed and inefficient emergency decision-making. At the same time, the mine environment is complex and changeable, and the safe evacuation of underground personnel and the dispatch of rescue personnel are also a huge challenge. Traditional path planning methods are difficult to find the optimal evacuation path and the rescue route with the least risk in a complex mine environment.

[0004] Therefore, there is an urgent need for an underground mine emergency command system based on an improved expert system, which can realize real-time analysis of accidents and optimization of emergency response by integrating optimization algorithms and risk assessment algorithms, improve the efficiency and safety of mine emergency command, and ensure the safety of miners' lives and the continuity of mine production. Summary of the invention

[0005] In view of the technical problems existing in the above-mentioned traditional underground mine emergency command system, the present invention proposes an underground mine emergency command system based on an improved expert system, which integrates accident analysis, accident type analysis, underground personnel command and rescue personnel command functions, utilizes optimization algorithms and risk assessment algorithms, realizes optimal route planning and minimum risk path selection, and improves the efficiency and safety of emergency command.

[0006] In order to achieve the above-mentioned purpose, the technical solution adopted by the present invention is to include an accident analysis module, an accident type analysis module, an underground personnel command module and a rescue personnel module;

[0007] The accident analysis module uses sensor data and historical accident data to perform accident cause analysis and risk assessment through an improved expert system, and provides a real-time accident analysis report;

[0008] The accident type analysis module uses a classification algorithm to analyze accident types based on accident characteristics and historical data, and provides emergency response plans for different accident types;

[0009] The underground personnel command module uses the improved A* algorithm and Dijkstra algorithm to plan the optimal evacuation route, ensuring that the underground personnel can reach the safe area in the shortest time. At the same time, the genetic algorithm and particle swarm optimization algorithm are used for dynamic scheduling and resource optimization;

[0010] The rescuer command module plans the rescue route through the minimum risk path algorithm and Monte Carlo simulation, selects the path with the minimum accident risk, and ensures that the rescuers can reach the accident site safely and quickly.

[0011] Preferably, the accident analysis module is implemented by:

[0012] Step 1: Data collection: sensors are used to collect various environmental data in the mine in real time, including gas concentration, temperature, humidity, vibration, noise, etc. At the same time, the system also integrates historical accident data to form a comprehensive database;

[0013] Step 2: Data preprocessing: preprocess the collected data, including denoising, normalization, feature extraction and other steps to ensure the accuracy and consistency of the data;

[0014] Step 3: Use the improved expert system to analyze the accident. The expert system combines rule reasoning and machine learning methods, including rule reasoning and machine learning; rule reasoning is based on predefined accident analysis rules, and the system conducts preliminary analysis and judgment on the data to generate preliminary accident cause and impact assessment; machine learning uses the trained random forest machine learning model to further conduct in-depth analysis and prediction of the data, optimize and verify the results of rule reasoning;

[0015] Step 4: Risk assessment: Based on the accident analysis results, the system conducts risk assessment, including the severity of the accident, possible impact range, subsequent potential risks, etc., and generates a detailed accident analysis report;

[0016] Step 5: Report generation: The system generates a visual accident analysis report based on the analysis and evaluation results and provides it to the emergency command center to guide subsequent emergency response decisions.

[0017] Preferably, the accident type analysis module is implemented by:

[0018] Step 1: Feature extraction: extract key features from accident data, such as accident location, time, environmental conditions, and equipment status;

[0019] Step 2: Use the random forest classification algorithm to classify the accident type. The specific steps are to use historical accident data to train the classification model, optimize the model parameters, and improve the classification accuracy. After the accident occurs, the system inputs the accident data into the classification model in real time to quickly determine the accident type;

[0020] Step 3: Generate emergency response plan. Based on the classification results, the system automatically generates emergency response plans for different accident types, including specific measures such as personnel evacuation, equipment shutdown, and rescue resource dispatch.

[0021] Step 4: Feedback and optimization. The system records the actual results of each accident analysis and response, and continuously optimizes the classification model and emergency response plan through the feedback mechanism to improve the accuracy and response efficiency of the system.

[0022] Preferably, the specific implementation steps of the improved A* algorithm of the underground personnel command module are as follows:

[0023] Step 1: Modeling of terrain information, slope calculation, measuring the height difference between adjacent nodes to evaluate the slope, assuming that the height of node x and its neighbor node y are H x and H y , the horizontal distance between them is d xy . Then the slope of node x is For multiple neighbor nodes, the average slope can be taken Where N is the number of nodes; surface roughness is the change of the surface around the calculated node to evaluate the surface roughness in, is the average height of neighbor nodes;

[0024] Step 2: Constructing the obstacle model. The first step is to evaluate the impact of obstacles. The impact of obstacles on the path is manifested as an increase in the path cost. The modeling method is to evaluate the obstacle density per unit area around the node. Suppose there are O obstacles within a radius r around the node x, and the area is A. The calculation formula for the obstacle density is: In addition, for obstacle proximity calculation, the distance from node x to the nearest obstacle is calculated. Assuming the nearest obstacle is o, the obstacle proximity calculation formula is:

[0025] Step 3: Modeling of environmental damage degree. The environmental damage degree function D(x) is a comprehensive evaluation of the damage situation at node x, mainly including damage intensity and damage diffusion. The damage intensity is a direct measurement of the damage situation at node x. In order to evaluate the difference before and after damage, the baseline state of node x before damage is set as Baseline(x), and its feature after damage is set as Feature(x). Then the damage intensity formula is: Among them, Feature(x) represents the i-th damage feature at node x, Baseline(x) represents the baseline state of node x before damage, and w irepresents the weight coefficient of the i-th feature, which is used to reflect the influence of the feature on the damage intensity, and n is the total number of damage features; the damage diffusion is an assessment of the damage impact range, which represents the influence of the damage on the area around the node x. Let the damage diffusion within the radius r around the node x be Among them, Neighbors r (x) represents the set of neighbor nodes within the radius r around node x, DamageIntensity(x) represents the damage degree of neighbor node y, N r (x) is the number of neighboring nodes; the comprehensive damage degree function D(x) combines the damage intensity and damage spread to provide a comprehensive assessment of the damage situation at node x: D(x) = αDamageIntensity(x) + βDamageSpread(x) where α and β are weight coefficients used to adjust the impact of damage intensity and damage spread on the comprehensive damage degree;

[0026] Step 4: Comprehensive improvement of heuristic function h(x) = γ1ASlop(x) + γ2R(x) + γ3P(x) + γ4D(x) where γ1γ2γ3γ4 are weight parameters of each factor, which are adjusted according to the specific application scenario;

[0027] Step 5: Use the clustering method to classify the calculated comprehensive damage degree. First, calculate the distance matrix between nodes based on the comprehensive damage degree D(x) Then, a bottom-up hierarchical clustering method is used to treat each node as a separate cluster, and the clusters with the highest similarity are gradually merged until the preset clustering conditions are met. Finally, the clustering effect is evaluated through indicators such as silhouette coefficient and intra-cluster distance to determine the optimal number of clusters.

[0028] Step 6: Based on the clustering results, identify the area with the lowest degree of damage and the largest space as the safe evacuation area; first, screen the safe area, select several clusters with the lowest comprehensive degree of damage from the clustering results, and further evaluate the spatial size and accessibility of these clusters; then, evaluate the spatial size of each cluster based on the distribution and distance of the nodes to ensure that the selected safe area can accommodate enough evacuees; finally, consider factors such as terrain complexity, obstacle distribution, and damage diffusion, evaluate the accessibility of the selected area, and ensure the safety and reliability of the evacuation path;

[0029] Step 7: Use the improved A* algorithm to plan the evacuation path to ensure that the personnel in the mine can safely reach the predetermined evacuation area; first, use the current node as the starting point and the target evacuation area as the end point, and use the improved A* algorithm to search for the optimal path. Among them, Distance(x1,x2) represents the Euclidean distance from node x1 to the target node x2, and α is the balance coefficient, which is used to adjust the weight of distance and degree of damage. Finally, according to real-time monitoring data and environmental changes, the evacuation path is dynamically adjusted to ensure that personnel can reach the safe area smoothly.

[0030] Preferably, the specific implementation steps of the improved Dijkstra algorithm of the underground personnel command module are as follows:

[0031] Step 1: Initialize the distances of all nodes, the starting distance is 0, and the distances of other nodes are infinite;

[0032] Step 2: Select the unprocessed node with the smallest distance and update the distance of its neighboring nodes;

[0033] Step 3: Repeat the above steps until the shortest path is found, and reconstruct the optimal evacuation route through the recorded path;

[0034] Step 4: Use genetic algorithm and particle swarm optimization algorithm to perform dynamic scheduling and resource optimization to ensure efficient evacuation of underground personnel.

[0035] Preferably, the rescuer command module plans the rescue route through the minimum risk path algorithm and Monte Carlo simulation, and selects the path with the minimum accident risk. The specific implementation steps are as follows:

[0036] Step 1: Minimum risk path algorithm, build a risk graph, assign a risk value to each path; initialize the node risk value, select the node with the minimum risk, and update the risk value of the neighboring node; repeat the above steps until the minimum risk path is found, and reconstruct the minimum risk path by backtracking;

[0037] Step 2: Monte Carlo simulation, generate a large number of possible path combinations, simulate each path, calculate the accident probability; statistically analyze the simulation results and select the path with the lowest accident probability.

[0038] Compared with the prior art, the advantages and positive effects of the present invention are:

[0039] 1. Through the improved expert system, combined with sensor data and historical accident data, accident cause analysis and risk assessment are realized, real-time accident analysis reports are provided, and corresponding emergency response plans are generated according to the accident type.

[0040] 2. Use the improved A* algorithm and Dijkstra algorithm, taking into account factors such as terrain, obstacles, and degree of environmental damage, to ensure that underground personnel can reach a safe area in the shortest time.

[0041] 3. The rescuer command module plans the rescue route through the minimum risk path algorithm and Monte Carlo simulation, selects the path with the lowest accident risk, and ensures that rescuers can reach the accident scene safely and quickly. DETAILED DESCRIPTION

[0042] In order to more clearly understand the above-mentioned purpose, features and advantages of the present invention, the present invention is further described below in conjunction with embodiments. It should be noted that the embodiments of the present application and the features in the embodiments can be combined with each other without conflict.

[0043] In the following description, many specific details are set forth to facilitate a full understanding of the present invention. However, the present invention may also be implemented in other ways than those described herein. Therefore, the present invention is not limited to the specific embodiments of the following disclosure.

[0044] Embodiment, taking into account that with the deepening and increasing complexity of mine operations, mine accidents occur frequently, including various safety accidents such as gas explosions, fires, and landslides. The present invention realizes real-time analysis of accidents and optimization of emergency response by integrating optimization algorithms and risk assessment algorithms, thereby improving the efficiency and safety of mine emergency command. First, the system uses sensor data and historical accident data to perform accident cause analysis and risk assessment through an improved expert system, and provides real-time accident analysis reports.

[0045] The accident analysis module ensures the accuracy and consistency of data through data collection and preprocessing. In addition, an improved expert system is used for accident analysis. The expert system combines rule reasoning and machine learning methods, specifically including rule reasoning and machine learning; rule reasoning is based on predefined accident analysis rules, and the system performs preliminary analysis and judgment on the data to generate preliminary accident causes and impact assessments; machine learning uses the trained random forest machine learning model to further perform in-depth analysis and prediction of the data, optimize and verify the results of rule reasoning. Finally, based on the accident analysis results, the system conducts risk assessment, generates reports, and guides subsequent emergency response decisions.

[0046] Considering that the judgment of accident type is the key to the entire emergency command system, the accident type analysis module extracts features and then uses the random forest classification algorithm to classify the accident type. Then the emergency response plan is generated. According to the classification results, the system automatically generates emergency response plans for different accident types. Finally, the system continuously optimizes the classification model and emergency response plan through the feedback mechanism to improve the accuracy and response efficiency of the system.

[0047] In order to ensure that the underground personnel can reach the safe area in the shortest time and minimize casualties, the underground personnel command module improves the A* and Dijkstra algorithms. The steps of improving A* are first to model the terrain information, calculate the slope, and measure the height difference between adjacent nodes to evaluate the slope; the construction of the obstacle model is first to evaluate the impact of obstacles. The impact of obstacles on the path is manifested as an increase in the path cost. The modeling method is to evaluate the obstacle density per unit area around the node. Assume that there are O obstacles within a radius r around the node x, and the area is A. Then the calculation formula of the obstacle density is: In addition, for obstacle proximity calculation, the distance from node x to the nearest obstacle is calculated. Assuming the nearest obstacle is o, the obstacle proximity calculation formula is: The next step is to model the degree of environmental damage, which mainly includes two aspects: damage intensity and damage diffusion; the damage intensity formula is Among them, Feature(x) represents the i-th damage feature at node x, Baseline(x) represents the baseline state of node x before damage, and w i represents the weight coefficient of the i-th feature, which is used to reflect the influence of the feature on the damage intensity, n is the total number of damage features; the damage spread is Among them, Neighbors r (x) represents the set of neighbor nodes within the radius r around node x, DamageIntensity(x) represents the damage degree of neighbor node y, N r (x) is the number of neighbor nodes; comprehensive damage degree function D(x) = αDamageIntensity(x) + βDamageSpread(x) where α and β are weight coefficients used to adjust the impact of damage intensity and damage spread on the comprehensive damage degree; comprehensive improvement of heuristic function h(x) = γ1ASlop(x) + γ2R(x) + γ3P(x) + γ4D(x) where γ1γ2γ3γ4 are weight parameters of each factor, which are adjusted according to the specific application scenario; clustering method is used to classify the calculated comprehensive damage degree. First, the distance matrix between nodes is calculated based on the comprehensive damage degree D(x). Then, a bottom-up hierarchical clustering method is used to treat each node as a separate cluster, and the clusters with the highest similarity are gradually merged until the preset clustering conditions are met; finally, the clustering effect is evaluated by indicators such as silhouette coefficient and intra-cluster distance to determine the optimal number of clusters; then, based on the clustering results, the area with the lowest degree of damage and larger space is identified as the safe evacuation area; first, the safe area is screened, and several clusters with the lowest comprehensive degree of damage are selected from the clustering results, and the spatial size and accessibility of these clusters are further evaluated; then, based on the distribution and distance of the nodes, the spatial size of each cluster is evaluated to ensure that the selected safe area can accommodate enough evacuees; finally, factors such as terrain complexity, obstacle distribution and damage diffusion are considered to evaluate the accessibility of the selected area to ensure the safety and reliability of the evacuation path; finally, the improved A* algorithm is used to plan the evacuation path to ensure that the personnel underground can safely reach the predetermined evacuation area; first, the current node is used as the starting point and the target evacuation area is used as the end point, and the improved A* algorithm is used to search for the optimal path Among them, Distance(x1,x2) represents the Euclidean distance from node x1 to target node x2, and α is the balance coefficient, which is used to adjust the weight of distance and degree of damage; finally, according to real-time monitoring data and environmental changes, the evacuation path is dynamically adjusted to ensure that personnel can reach the safe area smoothly. The steps of improving the Dijkstra algorithm are to initialize the distance of all nodes, with the starting distance being 0 and the distance of other nodes being infinite; select the unprocessed node with the smallest distance and update the distance of its neighboring nodes; repeat the above steps until the shortest path is found, and reconstruct the optimal evacuation route through the recorded path; use genetic algorithm and particle swarm optimization algorithm for dynamic scheduling and resource optimization to ensure that personnel in the mine can evacuate efficiently.

[0048] Finally, in order to ensure that the selected rescue route is the path with the lowest accident risk, the present invention uses the minimum risk path algorithm and Monte Carlo simulation in the rescue personnel command module to plan the rescue route and select the path with the lowest accident risk. First, the minimum risk path algorithm constructs a risk graph and assigns a risk value to each path; initializes the node risk value, selects the node with the lowest risk, and updates the risk value of the neighboring node; repeats the above steps until the minimum risk path is found; reconstructs the minimum risk path by backtracking; Monte Carlo simulation generates a large number of possible path combinations, simulates each path, and calculates the accident probability; statistically analyzes the simulation results and selects the path with the lowest accident probability.

[0049] The above description is only a preferred embodiment of the present invention and does not limit the present invention in other forms. Any technician familiar with the profession may use the technical content disclosed above to change or modify it into an equivalent embodiment with equivalent changes and apply it to other fields. However, any simple modification, equivalent change and modification made to the above embodiment based on the technical essence of the present invention without departing from the content of the technical solution of the present invention still falls within the protection scope of the technical solution of the present invention.

Claims

1. An underground mine emergency command system based on an improved expert system, characterized in that: It includes accident analysis module, accident type analysis module, underground personnel command module and rescue personnel module; The accident analysis module uses sensor data and historical accident data to perform accident cause analysis and risk assessment through an improved expert system, and provides a real-time accident analysis report; The accident type analysis module uses a classification algorithm to analyze accident types based on accident characteristics and historical data, and provides emergency response plans for different accident types; The underground personnel command module uses the improved A* algorithm and Dijkstra algorithm to plan the optimal evacuation route, ensuring that the underground personnel can reach the safe area in the shortest time. At the same time, the genetic algorithm and particle swarm optimization algorithm are used for dynamic scheduling and resource optimization; The rescuer command module plans the rescue route through the minimum risk path algorithm and Monte Carlo simulation, selects the path with the minimum accident risk, and ensures that the rescuers can reach the accident site safely and quickly.

2. The underground mine emergency command system based on the improved expert system according to claim 1 is characterized in that: The implementation steps of the accident analysis module are: Step 1: Data collection: sensors are used to collect various environmental data in the mine in real time, including gas concentration, temperature, humidity, vibration, noise, etc. At the same time, the system also integrates historical accident data to form a comprehensive database; Step 2: Data preprocessing: preprocess the collected data, including denoising, normalization, feature extraction and other steps to ensure the accuracy and consistency of the data; Step 3: Use the improved expert system to analyze the accident. The expert system combines rule reasoning and machine learning methods, specifically including rule reasoning and machine learning; rule reasoning is based on predefined accident analysis rules, and the system conducts preliminary analysis and judgment on the data to generate a preliminary accident cause and impact assessment; Machine learning uses the trained random forest machine learning model to further analyze and predict data, optimize and verify the results of rule reasoning; Step 4: Risk assessment: Based on the accident analysis results, the system conducts risk assessment, including the severity of the accident, possible impact range, subsequent potential risks, etc., and generates a detailed accident analysis report; Step 5: Report generation: The system generates a visual accident analysis report based on the analysis and evaluation results and provides it to the emergency command center to guide subsequent emergency response decisions.

3. The underground mine emergency command system based on the improved expert system according to claim 1 is characterized in that: The implementation steps of the accident type analysis module are: Step 1: Feature extraction: extract key features from accident data, such as accident location, time, environmental conditions, and equipment status; Step 2: Use the random forest classification algorithm to classify the accident type. The specific steps are to use historical accident data to train the classification model, optimize the model parameters, and improve the classification accuracy. After the accident occurs, the system inputs the accident data into the classification model in real time to quickly determine the accident type; Step 3: Generate emergency response plan. Based on the classification results, the system automatically generates emergency response plans for different accident types, including specific measures such as personnel evacuation, equipment shutdown, and rescue resource dispatch. Step 4: Feedback and optimization. The system records the actual results of each accident analysis and response, and continuously optimizes the classification model and emergency response plan through the feedback mechanism to improve the accuracy and response efficiency of the system.

4. The underground mine emergency command system based on the improved expert system according to claim 1 is characterized in that: The specific implementation steps of the improved A* algorithm of the underground personnel command module are as follows: Step 1: Modeling of terrain information, slope calculation, measuring the height difference between adjacent nodes to evaluate the slope, assuming that the height of node x and its neighbor node y are H x and H y , the horizontal distance between them is d xy . Then the slope of node x is For multiple neighbor nodes, the average slope can be taken Where N is the number of nodes; surface roughness is the change of the surface around the calculated node to evaluate the surface roughness in, is the average height of neighbor nodes; Step 2: Constructing the obstacle model. The first step is to evaluate the impact of obstacles. The impact of obstacles on the path is manifested as an increase in the path cost. The modeling method is to evaluate the obstacle density per unit area around the node. Suppose there are O obstacles within a radius r around the node x, and the area is A. The calculation formula for the obstacle density is: In addition, for obstacle proximity calculation, the distance from node x to the nearest obstacle is calculated. Assuming the nearest obstacle is o, the obstacle proximity calculation formula is: Step 3: Modeling of environmental damage degree. The environmental damage degree function D(x) is a comprehensive evaluation of the damage situation at node x, mainly including damage intensity and damage diffusion. The damage intensity is a direct measurement of the damage situation at node x. In order to evaluate the difference before and after damage, the baseline state of node x before damage is set as Baseline(x), and its feature after damage is set as Feature(x). Then the damage intensity formula is: Among them, Feature(x) represents the i-th damage feature at node x, Baseline(x) represents the baseline state of node x before damage, and w i represents the weight coefficient of the i-th feature, which is used to reflect the influence of the feature on the damage intensity, and n is the total number of damage features; the damage diffusion is an assessment of the damage impact range, which represents the influence of the damage on the area around the node x. Let the damage diffusion within the radius r around the node x be Among them, Neighbors r (x) represents the set of neighbor nodes within the radius r around node x, DamageIntensity(x) represents the damage degree of neighbor node y, N r (x) is the number of neighboring nodes; the comprehensive damage degree function D(x) combines the damage intensity and damage spread to provide a comprehensive assessment of the damage situation at node x: D(x) = αDamageIntensity(x) + βDamageSpread(x) where α and β are weight coefficients used to adjust the impact of damage intensity and damage spread on the comprehensive damage degree; Step 4: Comprehensive improvement of heuristic function h(x) = γ1ASlop(x) + γ2R(x) + γ3P(x) + γ4D(x) where γ1γ2γ3γ4 are weight parameters of each factor, which are adjusted according to the specific application scenario; Step 5: Use the clustering method to classify the calculated comprehensive damage degree. First, calculate the distance matrix between nodes based on the comprehensive damage degree D(x) Then, a bottom-up hierarchical clustering method is used to treat each node as a separate cluster, and the clusters with the highest similarity are gradually merged until the preset clustering conditions are met. Finally, the clustering effect is evaluated through indicators such as silhouette coefficient and intra-cluster distance to determine the optimal number of clusters. Step 6: Based on the clustering results, identify the area with the lowest degree of damage and the largest space as the safe evacuation area; first, screen the safe area, select several clusters with the lowest comprehensive degree of damage from the clustering results, and further evaluate the spatial size and accessibility of these clusters; then, evaluate the spatial size of each cluster based on the distribution and distance of the nodes to ensure that the selected safe area can accommodate enough evacuees; finally, consider factors such as terrain complexity, obstacle distribution, and damage diffusion, evaluate the accessibility of the selected area, and ensure the safety and reliability of the evacuation path; Step 7: Use the improved A* algorithm to plan the evacuation path to ensure that the personnel in the mine can safely reach the predetermined evacuation area; first, use the current node as the starting point and the target evacuation area as the end point, and use the improved A* algorithm to search for the optimal path. Among them, Distance(x1,x2) represents the Euclidean distance from node x1 to the target node x2, and α is the balance coefficient, which is used to adjust the weight of distance and degree of damage. Finally, according to real-time monitoring data and environmental changes, the evacuation path is dynamically adjusted to ensure that personnel can reach the safe area smoothly.

5. The underground mine emergency command system based on the improved expert system according to claim 1 is characterized in that: The specific implementation steps of the improved Dijkstra algorithm of the underground personnel command module are as follows: Step 1: Initialize the distances of all nodes, the starting distance is 0, and the distances of other nodes are infinite; Step 2: Select the unprocessed node with the smallest distance and update the distance of its neighboring nodes; Step 3: Repeat the above steps until the shortest path is found, and reconstruct the optimal evacuation route through the recorded path; Step 4: Use genetic algorithm and particle swarm optimization algorithm to perform dynamic scheduling and resource optimization to ensure efficient evacuation of underground personnel.

6. The underground mine emergency command system based on the improved expert system according to claim 1 is characterized in that: The rescuer command module plans the rescue route through the minimum risk path algorithm and Monte Carlo simulation, and selects the path with the minimum accident risk. The specific implementation steps are as follows: Step 1: Minimum risk path algorithm, build a risk graph, assign a risk value to each path; initialize the node risk value, select the node with the minimum risk, and update the risk value of the neighboring nodes; Repeat the above steps until the minimum risk path is found, and then reconstruct the minimum risk path by backtracking; Step 2: Monte Carlo simulation, generate a large number of possible path combinations, simulate each path, and calculate the accident probability; Statistically analyze the simulation results and select the path with the lowest accident probability.

Images