Rule engine-based mine emergency rescue command training guiding method and system
By constructing a mine emergency rescue training system based on a rule engine, the problem of insufficient procedure parsing in traditional systems is solved, intelligent training process guidance is achieved, and the operational capabilities and training effectiveness of rescue personnel are improved.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- CHINA COAL TECH & ENG GRP SHENYANG ENG CO
- Filing Date
- 2025-12-11
- Publication Date
- 2026-05-08
AI Technical Summary
Traditional mine emergency drill systems lack the ability to automatically analyze and execute regulations and clauses. Rescue procedures do not conform to regulations and requirements, and historical rescue cases are not deeply integrated with the physics engine. This results in insufficient realism and credibility of training and drills, low reusability, and affects the standardized operation capabilities of rescue personnel.
The rule engine-based approach constructs a rule base by acquiring historical cases and procedural clauses of mine rescue, generates dynamic response process templates, optimizes rescue rules using knowledge graphs and physical simulation rules, establishes a dynamic response process guidance mechanism, and realizes intelligent training process guidance.
This improved the authenticity and effectiveness of mine emergency rescue training, enhanced the standardized operational capabilities of rescue personnel, ensured the alignment of training procedures with actual rescue protocols, and increased the credibility and reusability of the training.
Smart Images
Figure CN121997534A_ABST
Abstract
Description
Technical Field
[0001] This application belongs to the field of emergency rescue training technology, specifically, it relates to a rule engine-based method and system for mine emergency rescue command training guidance. Background Technology
[0002] Mine emergency rescue is a high-risk operation that requires rescuers to possess strong comprehensive abilities in physical fitness, psychological resilience, knowledge, and experience. This is crucial for ensuring safe and efficient rescue operations and minimizing accident losses. Conducting rescue drills and training is a fundamental way to improve the capabilities of rescue personnel. The main types of accidents requiring mine emergency rescue include natural environmental disasters such as underground water inrush, fire, gas explosions, coal dust explosions, and roof collapses.
[0003] However, traditional mine emergency drill systems suffer from a disconnect between regulations and drill procedures, lacking the ability to automatically analyze and execute regulations, resulting in discrepancies between rescue procedures and regulations, and making it difficult for rescuers to quickly master standardized operations. Furthermore, historical rescue cases are mostly stored as static scripts, not deeply integrated with the physics engine and regulations, making it impossible to provide reference for current training and drills through intelligent matching, leading to a low reuse rate of historical cases. Additionally, the physical simulation lacks realism; traditional training systems oversimplify the physical simulation of the mine environment, deviating significantly from the actual evolution of disasters, thus affecting the credibility of training and drills. Summary of the Invention
[0004] To address the shortcomings of existing technologies, this invention provides a rule-based method for training and guiding emergency rescue command in mines, comprising the following steps:
[0005] Obtain historical cases of mine rescue, calculate the similarity of these cases, and generate a case database.
[0006] Obtain the terms and conditions of mine rescue regulations, and build a rule base based on the case database and the terms and conditions of mine rescue regulations, including a rescue rule sub-base, a regulation rule font library and a physical simulation rule sub-base;
[0007] Based on the rule base, prefabricated typical case process templates for different mine disasters are generated;
[0008] Establish a dynamic response process guidance mechanism, and generate mine emergency rescue and response processes based on the dynamic response process guidance mechanism.
[0009] Furthermore, historical cases of mine rescue are obtained, and the similarity of these cases is calculated to generate a case database. The specific method is as follows:
[0010] Obtain historical cases of mine rescue and extract disaster characteristic parameters and environmental characteristic parameters;
[0011] Disaster characteristic parameters include disaster type, location, scope of impact, and initial parameters. Initial parameters include: initial concentration of harmful gases, initial impact force of explosion, initial water inflow, initial water level rise rate, initial ignition temperature, or initial fire coverage area.
[0012] Environmental characteristic parameters refer to the mine's own and surrounding environmental indicators that affect the development of the disaster, the selection of rescue routes, the availability of rescue resources and waste, and the results of physical simulations when a mine disaster occurs. These include: external geographical environment parameters and internal facility environment parameters of the mine; external geographical environment parameters of the mine include: lithology of the surrounding rock of the roadway, roadway burial depth, ore layer thickness, geological structure, surface topography type, and surface water distribution; internal facility environment parameters of the mine include: ventilation system type, fan air volume, air pressure, roadway ventilation cross-sectional area, number and location of air doors or windows, mine width, height, location, number, and capacity of disaster refuge chambers;
[0013] Similarity calculations are performed on historical mine rescue cases based on disaster characteristic parameters and environmental characteristic parameters, and the information on historical mine rescue cases is tagged according to the calculation results.
[0014] Using knowledge graph technology, we associate tagged historical mine rescue case information to construct knowledge graphs of different types of historical rescue cases. Then, we weight and merge the knowledge graphs of various types of historical rescue cases to generate a case database.
[0015] Furthermore, the method for constructing the physical simulation rule sub-library is as follows:
[0016] Based on the environmental parameters of the mine's internal facilities, a three-dimensional mine roadway model is constructed. Combining this with physical theories related to mine hazards, physical simulation rules are generated, including:
[0017] One-dimensional incompressible fluid equations are used to simulate wind speed and direction in tunnels, generating airflow rules.
[0018] The concentration distribution of different gases is dynamically calculated based on the Gaussian plume model, and gas diffusion rules are generated.
[0019] Based on the heat conduction equation and fuel distribution, the fire spread path is predicted, and fire spread rules are generated.
[0020] Add the physical simulation rules to the physical simulation rule sub-library.
[0021] Furthermore, the method for constructing the rescue rule sub-library is as follows:
[0022] Historical mine rescue cases are categorized and extracted from the case database. Record data from each category of historical mine rescue cases is obtained, and the operational procedures during the rescue process are simulated to generate rescue rules based on the operational procedures.
[0023] The recorded data includes operational procedures, disaster characteristic parameters, and environmental characteristic parameters from historical rescue cases;
[0024] A three-dimensional dynamic simulation system is constructed based on physical simulation rules, a three-dimensional mine roadway model, disaster characteristic parameters, and environmental characteristic parameters. The generated rescue rules are imported into the three-dimensional dynamic simulation system. The rescue rules are reproduced in their entirety by generating virtual accident scenarios. A confusion matrix is used to analyze the matching degree between the execution results of the rescue rules and the actual rescue results. The deviation rate is calculated. When the deviation rate exceeds a preset threshold, the rescue rule is determined to be a failed rule. The failed rule is then optimized in terms of topology using a rule correction algorithm to obtain the optimized rescue rule.
[0025] The topology of invalid rules is optimized using a rule correction algorithm. The specific method is as follows:
[0026] Receive failure rule identifiers and deviation data, send failure rule identifiers to the expert knowledge base, input failure rule parameters and deviation data into the Bayesian network model, use the expert knowledge base to return compliance constraints based on failure rule identifiers and generate correction directions; use the Bayesian network model to output key influencing factors and quantitative correction suggestions based on parameter data.
[0027] By combining compliance constraints, correction directions, and quantitative correction suggestions, a correction and optimization scheme that meets compliance requirements and has the highest prediction success rate is generated.
[0028] The failure rules are optimized using a correction and optimization scheme that meets compliance requirements and has the highest prediction success rate. The optimized rescue rules are then added to the rescue rule sub-library along with rescue rules whose deviation rate does not exceed a preset threshold.
[0029] Furthermore, the method for constructing the procedure rule sub-library is as follows:
[0030] Each clause in the mine rescue regulations is broken down into executable logic with clear logical relationships. The executable logic is then converted into atomic logic units. The RETE algorithm is used to match and resolve conflicts between the atomic logic units to establish the regulations rules.
[0031] A mine type labeling system is constructed. Based on the atomic logical units of the mine rescue regulations, the cosine similarity algorithm is used to calculate the matching degree between the atomic logical units and different mine types, generating a visual compatibility heatmap. The visual compatibility heatmap is used to filter matching based on a preset threshold. For regulations rules with a matching degree lower than the preset threshold, a warning label is added and they are dynamically blocked. The combination of regulations rules with a matching degree higher than the preset threshold from the same type of mine rescue cases is selected.
[0032] Add combinations of procedures and rules that exceed the preset threshold in similar mine rescue cases to the procedure and rule sub-library.
[0033] Furthermore, the handling process template includes handling steps, required resources, corresponding rules, and physical parameters.
[0034] Furthermore, establish a dynamic handling process guidance mechanism, including:
[0035] Establish a database of rescue resources, including: the qualifications of rescue teams, equipment performance parameters, and material reserves;
[0036] When a trained user selects a pre-made typical case process template, the system compares the required resources of the selected template with the actual available resources. If the matching degree is lower than a preset threshold, an alternative template is generated based on the rescue rules in the rule base, and the corresponding rules and physical parameters are loaded to generate the corresponding mine emergency rescue and disposal process. If the matching degree is not lower than the preset threshold, the corresponding rules and physical parameters are loaded based on the pre-made typical case process template selected by the user to generate the corresponding mine emergency rescue and disposal process.
[0037] The dynamic response process guidance mechanism also includes a closed-loop optimization module. After training, this module compares the user's mine emergency rescue and response process with historical best-in-class rescue cases in the mine case database to establish a difference analysis matrix. It then combines relevant knowledge from the rule base and case knowledge graph to generate targeted improvement suggestions. The specific method is as follows:
[0038] Key operational data of historical best cases are extracted from the case knowledge graph, including: time series data, resource scheduling data, and decision parameter data. The time series data includes: the execution time of each disposal step and the time consumed between steps. The resource scheduling data includes: the order of calling various resources and the amount of resources used. The decision parameter data includes: the decision thresholds of key nodes and the operational logic. The key operational data are normalized and quantified to generate a matrix of historical best cases.
[0039] Furthermore, a mine emergency rescue and response process is generated based on a dynamic response process guidance mechanism. The specific method is as follows:
[0040] Based on the dynamic handling process guidance mechanism, a historical best case matrix and a user operation matrix are generated. The method for generating the user operation matrix is the same as that for generating the historical best case matrix. The user operation matrix is generated through the user's mine emergency rescue and handling process. The deviation between the user operation matrix and the historical best case matrix is calculated. The difference data with the deviation calculation result greater than the threshold is obtained. The corresponding search is performed in the case knowledge graph based on the difference data to obtain targeted improvement suggestions.
[0041] Based on the improvement suggestions, a mine emergency rescue and response process was generated for training and guidance.
[0042] On the other hand, the present invention also provides a mine emergency rescue command training and guidance system based on a rule engine, including a data storage module, a rule generation module, a template creation module, a process generation module and a closed-loop optimization module;
[0043] The data storage module is used to store historical mine rescue cases, a knowledge graph of historical rescue cases, disaster characteristic parameters, and environmental characteristic parameters; and to generate a case database based on disaster and environmental characteristics.
[0044] The rule generation module is used to obtain the clauses of the mine rescue regulations and build a rule base based on the case database and the clauses of the mine rescue regulations, including a rescue rule sub-library, a regulation rule font library and a physical simulation rule sub-library;
[0045] The template creation module is used to generate prefabricated typical case process templates for different mine disasters;
[0046] The process generation module is used to load the corresponding rules and physical parameters based on the pre-made typical case process template selected by the user, and generate a mine emergency rescue and disposal process.
[0047] The closed-loop optimization module is used to compare the differences between user operations and historical best cases after training, establish a difference analysis matrix, and combine the difference data with relevant knowledge in the rule base and case knowledge graph to generate targeted improvement suggestions.
[0048] The beneficial effects of adopting the above technical solution are as follows: The present invention provides a mine emergency rescue command training guidance method and system based on a rule engine. By acquiring historical cases of mine rescue, a case database is generated. The historical cases in the case database are structured, and the mine rescue procedure clauses are standardized into rules to build a rule base. Based on a dynamic handling process guidance mechanism, pre-made typical case process templates for different mine disasters are generated, allowing users to select pre-made typical case process templates. The system automatically loads the corresponding rules and physical parameters to generate the corresponding mine emergency rescue handling process, realizing intelligent guidance of mine emergency rescue command training and handling process, improving the authenticity and effectiveness of training exercises, and enhancing the standardized operation capabilities of rescue personnel. Attached Figure Description
[0049] Figure 1 A schematic diagram of the steps of the mine emergency rescue command training and guidance method based on a rule engine provided in Embodiment 1 of the present invention;
[0050] Figure 2 A schematic diagram of the structure of the mine emergency rescue command training and guidance system based on a rule engine provided in Embodiment 2 of the present invention. Detailed Implementation
[0051] The specific implementation methods of this application will be further described in detail below with reference to the accompanying drawings and embodiments.
[0052] Example 1
[0053] A rule-based approach to training and guiding emergency rescue command in mines, such as... Figure 1 As shown, it includes the following steps:
[0054] Step 1: Obtain historical cases of mine rescue, calculate the similarity of historical cases of mine rescue, and generate a case database;
[0055] Obtain historical cases of mine rescue and extract disaster characteristic parameters and environmental characteristic parameters;
[0056] Disaster characteristic parameters include disaster type, location, scope of impact, and initial parameters. Initial parameters include: initial concentration of harmful gases, initial impact force of explosion, initial water inflow, initial water level rise rate, initial ignition temperature, or initial fire coverage area.
[0057] Environmental characteristic parameters refer to the mine's own and surrounding environmental indicators that affect the development of the disaster, the selection of rescue routes, the availability of rescue resources and waste, and the results of physical simulations when a mine disaster occurs. These include: external geographical environment parameters and internal facility environment parameters of the mine; external geographical environment parameters of the mine include: lithology of the surrounding rock of the roadway, roadway burial depth, ore layer thickness, geological structure, surface topography type, and surface water distribution; internal facility environment parameters of the mine include: ventilation system type, fan air volume, air pressure, roadway ventilation cross-sectional area, number and location of air doors or windows, mine width, height, location, number, and capacity of disaster refuge chambers;
[0058] Similarity calculations are performed on historical mine rescue cases based on disaster characteristic parameters and environmental characteristic parameters, and the information on historical mine rescue cases is tagged according to the calculation results.
[0059] Using knowledge graph technology, we associate tagged historical mine rescue case information to construct knowledge graphs of different types of historical rescue cases. Then, we weight and merge the knowledge graphs of various types of historical rescue cases to generate a case database.
[0060] Step 2: Obtain the terms and conditions of the mine rescue regulations, and build a rule base based on the case database and the terms and conditions of the mine rescue regulations, including a rescue rule sub-library, a regulation rule font library, and a physical simulation rule sub-library;
[0061] The method for constructing the physical simulation rule sub-library is as follows:
[0062] Based on the environmental parameters of the mine's internal facilities, a three-dimensional mine roadway model is constructed. Combining this with physical theories related to mine hazards, physical simulation rules are generated, including:
[0063] One-dimensional incompressible fluid equations are used to simulate wind speed and direction in tunnels, generating airflow rules.
[0064] The concentration distribution of different gases is dynamically calculated based on the Gaussian plume model, and gas diffusion rules are generated.
[0065] Based on the heat conduction equation and fuel distribution, the fire spread path is predicted, and fire spread rules are generated.
[0066] Add the physical simulation rules to the physical simulation rule sub-library;
[0067] The method for constructing the rescue rule sub-library is as follows:
[0068] Historical mine rescue cases are categorized and extracted from the case database. Record data from each category of historical mine rescue cases is obtained, and the operational procedures during the rescue process are simulated to generate rescue rules based on the operational procedures.
[0069] The recorded data includes operational procedures, disaster characteristic parameters, and environmental characteristic parameters from historical rescue cases;
[0070] A three-dimensional dynamic simulation system is constructed based on physical simulation rules, a three-dimensional mine roadway model, disaster characteristic parameters, and environmental characteristic parameters. The generated rescue rules are imported into the three-dimensional dynamic simulation system. The rescue rules are reproduced in their entirety by generating virtual accident scenarios. A confusion matrix is used to analyze the matching degree between the execution results of the rescue rules and the actual rescue results. The deviation rate is calculated. When the deviation rate exceeds a preset threshold, the rescue rule is determined to be a failed rule. The failed rule is then optimized in terms of topology using a rule correction algorithm to obtain the optimized rescue rule.
[0071] The topology of invalid rules is optimized using a rule correction algorithm. The specific method is as follows:
[0072] Receive failure rule identifiers and deviation data, send failure rule identifiers to the expert knowledge base, input failure rule parameters and deviation data into the Bayesian network model, use the expert knowledge base to return compliance constraints based on failure rule identifiers and generate correction directions; use the Bayesian network model to output key influencing factors and quantitative correction suggestions based on parameter data.
[0073] By combining compliance constraints, correction directions, and quantitative correction suggestions, a correction and optimization scheme that meets compliance requirements and has the highest prediction success rate is generated.
[0074] The failure rules are optimized using a correction and optimization scheme that meets compliance requirements and has the highest prediction success rate. The optimized rescue rules are then added to the rescue rule sub-library along with rescue rules whose deviation rate does not exceed a preset threshold.
[0075] The method for constructing the procedure rule sub-library is as follows:
[0076] Each clause in the mine rescue regulations is broken down into executable logic with clear logical relationships. The executable logic is then converted into atomic logic units. The RETE algorithm is used to match and resolve conflicts between the atomic logic units to establish the regulations rules.
[0077] A mine type labeling system is constructed. Based on the atomic logical units of the mine rescue regulations, the cosine similarity algorithm is used to calculate the matching degree between the atomic logical units and different mine types, generating a visual compatibility heatmap. The visual compatibility heatmap is used to filter matching based on a preset threshold. For regulations rules with a matching degree lower than the preset threshold, a warning label is added and they are dynamically blocked. The combination of regulations rules with a matching degree higher than the preset threshold from the same type of mine rescue cases is selected.
[0078] Add combinations of procedures and rules that exceed the preset threshold in similar mine rescue cases to the procedure and rule sub-library;
[0079] Step 3: Generate pre-made typical case process templates for different mine disasters based on the rule base;
[0080] A standardized handling process template is generated based on the rules in the rule base. The handling process template includes handling steps, required resources, corresponding rules and physical parameters.
[0081] Step 4: Establish a dynamic response process guidance mechanism, and generate a mine emergency rescue and response process based on the dynamic response process guidance mechanism;
[0082] Establish a dynamic handling process guidance mechanism, including:
[0083] Establish a database of rescue resources, including: the qualifications of rescue teams, equipment performance parameters, and material reserves;
[0084] When a trained user selects a pre-made typical case process template, the system compares the required resources of the selected template with the actual available resources. If the matching degree is lower than a preset threshold, an alternative template is generated based on the rescue rules in the rule base, and the corresponding rules and physical parameters are loaded to generate the corresponding mine emergency rescue and disposal process. If the matching degree is not lower than the preset threshold, the corresponding rules and physical parameters are loaded based on the pre-made typical case process template selected by the user to generate the corresponding mine emergency rescue and disposal process.
[0085] The dynamic response process guidance mechanism also includes a closed-loop optimization module. After training, this module compares the user's mine emergency rescue and response process with historical best-in-class rescue cases in the mine case database to establish a difference analysis matrix. It then combines relevant knowledge from the rule base and case knowledge graph to generate targeted improvement suggestions. The specific method is as follows:
[0086] Key operational data of historical best cases are extracted from the case knowledge graph, including: time series data, resource scheduling data, and decision parameter data. The time series data includes: the execution time of each disposal step and the time consumed between steps. The resource scheduling data includes: the order of calling various resources and the amount of resources used. The decision parameter data includes: the decision thresholds of key nodes and the operational logic. The key operational data are normalized and quantified to generate a matrix of historical best cases.
[0087] Based on the dynamic handling process guidance mechanism, a historical best case matrix and a user operation matrix are generated. The method for generating the user operation matrix is the same as that for generating the historical best case matrix. The user operation matrix is generated through the user's mine emergency rescue and handling process. The deviation between the user operation matrix and the historical best case matrix is calculated. The difference data with the deviation calculation result greater than the threshold is obtained. The corresponding search is performed in the case knowledge graph based on the difference data to obtain targeted improvement suggestions.
[0088] Based on the improvement suggestions, a mine emergency rescue and response process was generated for training and guidance.
[0089] Example 2:
[0090] A rule-based mine emergency rescue command and training guidance system, such as Figure 2 As shown, it includes a data storage module, a rule generation module, a template creation module, a process generation module, and a closed-loop optimization module;
[0091] The data storage module is used to store historical mine rescue cases, a knowledge graph of historical rescue cases, disaster characteristic parameters, and environmental characteristic parameters; and to generate a case database based on disaster and environmental characteristics.
[0092] The rule generation module is used to obtain the clauses of the mine rescue regulations and build a rule base based on the case database and the clauses of the mine rescue regulations, including a rescue rule sub-library, a regulation rule font library and a physical simulation rule sub-library;
[0093] The template creation module is used to generate prefabricated typical case process templates for different mine disasters;
[0094] The process generation module is used to load the corresponding rules and physical parameters based on the pre-made typical case process template selected by the user, and generate a mine emergency rescue and disposal process.
[0095] The closed-loop optimization module is used to compare the differences between user operations and historical best cases after training, establish a difference analysis matrix, and combine the difference data with relevant knowledge in the rule base and case knowledge graph to generate targeted improvement suggestions.
[0096] The various embodiments in this application are described in a progressive manner. The same or similar parts between the various embodiments can be referred to each other. Each embodiment focuses on describing the differences from other embodiments.
[0097] The scope of protection of this application is not limited to the embodiments described above. Obviously, those skilled in the art can make various modifications and variations to this disclosure without departing from the scope and spirit of this disclosure. If such modifications and variations fall within the scope of this disclosure and its equivalents, then the intent of this disclosure also includes these modifications and variations.
Claims
1. A training and guidance method for mine emergency rescue command based on a rule engine, characterized in that, Includes the following steps: Obtain historical cases of mine rescue, calculate the similarity of these cases, and generate a case database. Obtain the terms and conditions of mine rescue regulations, and build a rule base based on the case database and the terms and conditions of mine rescue regulations, including a rescue rule sub-base, a regulation rule font library and a physical simulation rule sub-base; Based on the rule base, prefabricated typical case process templates for different mine disasters are generated; Establish a dynamic response process guidance mechanism, and generate mine emergency rescue and response processes based on the dynamic response process guidance mechanism.
2. The mine emergency rescue command training and guidance method based on a rule engine according to claim 1, characterized in that, The process involves acquiring historical cases of mine rescue, calculating the similarity between these cases, and generating a case database. The specific method is as follows: Obtain historical cases of mine rescue and extract disaster characteristic parameters and environmental characteristic parameters; Disaster characteristic parameters include disaster type, location, scope of impact, and initial parameters. Initial parameters include: initial concentration of harmful gases, initial impact force of explosion, initial water inflow, initial water level rise rate, initial ignition temperature, or initial fire coverage area. Environmental characteristic parameters refer to the mine's own and surrounding environmental indicators that affect the development of the disaster, the selection of rescue routes, the availability of rescue resources and waste, and the results of physical simulations when a mine disaster occurs. These include: external geographical environment parameters and internal facility environment parameters of the mine; external geographical environment parameters of the mine include: lithology of the surrounding rock of the roadway, roadway burial depth, ore layer thickness, geological structure, surface topography type, and surface water distribution; internal facility environment parameters of the mine include: ventilation system type, fan air volume, air pressure, roadway ventilation cross-sectional area, number and location of air doors or windows, mine width, height, location, number, and capacity of disaster refuge chambers; Similarity calculations are performed on historical mine rescue cases based on disaster characteristic parameters and environmental characteristic parameters, and the information on historical mine rescue cases is tagged according to the calculation results. Using knowledge graph technology, we associate tagged historical mine rescue case information to construct knowledge graphs of different types of historical rescue cases. Then, we weight and merge the knowledge graphs of various types of historical rescue cases to generate a case database.
3. The mine emergency rescue command training and guidance method based on a rule engine according to claim 1, characterized in that, The method for constructing the physical simulation rule sub-library is as follows: Based on the environmental parameters of the mine's internal facilities, a three-dimensional mine roadway model is constructed. Combining this with physical theories related to mine hazards, physical simulation rules are generated, including: One-dimensional incompressible fluid equations are used to simulate wind speed and direction in tunnels, generating airflow rules. The concentration distribution of different gases is dynamically calculated based on the Gaussian plume model, and gas diffusion rules are generated. Based on the heat conduction equation and fuel distribution, the fire spread path is predicted, and fire spread rules are generated. Add the physical simulation rules to the physical simulation rule sub-library.
4. The mine emergency rescue command training and guidance method based on a rule engine according to claim 1, characterized in that, The method for constructing the rescue rule sub-library is as follows: Historical mine rescue cases are categorized and extracted from the case database. Record data from each category of historical mine rescue cases is obtained, and the operational procedures during the rescue process are simulated to generate rescue rules based on the operational procedures. The recorded data includes operational procedures, disaster characteristic parameters, and environmental characteristic parameters from historical rescue cases; A three-dimensional dynamic simulation system is constructed based on physical simulation rules, a three-dimensional mine roadway model, disaster characteristic parameters, and environmental characteristic parameters. The generated rescue rules are imported into the three-dimensional dynamic simulation system. The rescue rules are reproduced in their entirety by generating virtual accident scenarios. A confusion matrix is used to analyze the matching degree between the execution results of the rescue rules and the actual rescue results. The deviation rate is calculated. When the deviation rate exceeds a preset threshold, the rescue rule is determined to be a failed rule. The failed rule is then optimized in terms of topology using a rule correction algorithm to obtain the optimized rescue rule. The topology of invalid rules is optimized using a rule correction algorithm. The specific method is as follows: Receive failure rule identifiers and deviation data, send failure rule identifiers to the expert knowledge base, input failure rule parameters and deviation data into the Bayesian network model, and use the expert knowledge base to return compliance constraints based on the failure rule identifiers and generate correction directions; Using a Bayesian network model, key influencing factors and quantitative correction suggestions are output based on parameter data; By combining compliance constraints, correction directions, and quantitative correction suggestions, a correction and optimization scheme that meets compliance requirements and has the highest prediction success rate is generated. The failure rules are optimized using a correction and optimization scheme that meets compliance requirements and has the highest prediction success rate. The optimized rescue rules are then added to the rescue rule sub-library along with rescue rules whose deviation rate does not exceed a preset threshold.
5. The mine emergency rescue command training and guidance method based on a rule engine according to claim 1, characterized in that, The method for constructing the procedure rule sub-library is as follows: Each clause in the mine rescue regulations is broken down into executable logic with clear logical relationships. The executable logic is then converted into atomic logic units. The RETE algorithm is used to match and resolve conflicts between the atomic logic units to establish the regulations rules. A mine type labeling system is constructed. Based on the atomic logical units of the mine rescue regulations, the cosine similarity algorithm is used to calculate the matching degree between the atomic logical units and different mine types, generating a visual compatibility heatmap. The visual compatibility heatmap is used to filter matching based on a preset threshold. For regulations rules with a matching degree lower than the preset threshold, a warning label is added and they are dynamically blocked. The combination of regulations rules with a matching degree higher than the preset threshold from the same type of mine rescue cases is selected. Add combinations of procedures and rules that exceed the preset threshold in similar mine rescue cases to the procedure and rule sub-library.
6. The mine emergency rescue command training and guidance method based on a rule engine according to claim 1, characterized in that, The handling process template includes handling steps, required resources, corresponding rules, and physical parameters.
7. The mine emergency rescue command training and guidance method based on a rule engine according to claim 1, characterized in that, Establish a dynamic handling process guidance mechanism, including: Establish a database of rescue resources, including: the qualifications of rescue teams, equipment performance parameters, and material reserves; When a trained user selects a pre-made typical case process template, the system compares the required resources of the selected template with the actual available resources. If the matching degree is lower than a preset threshold, an alternative template is generated based on the rescue rules in the rule base, and the corresponding rules and physical parameters are loaded to generate the corresponding mine emergency rescue and disposal process. If the matching degree is not lower than the preset threshold, the corresponding rules and physical parameters are loaded based on the pre-made typical case process template selected by the user to generate the corresponding mine emergency rescue and disposal process. The dynamic response process guidance mechanism also includes a closed-loop optimization module. After training, this module compares the user's mine emergency rescue and response process with historical best-in-class rescue cases in the mine case database to establish a difference analysis matrix. It then combines relevant knowledge from the rule base and case knowledge graph to generate targeted improvement suggestions. The specific method is as follows: Key operational data of historical best cases are extracted from the case knowledge graph, including: time series data, resource scheduling data, and decision parameter data. The time series data includes: the execution time of each disposal step and the time consumed between steps. The resource scheduling data includes: the order of calling various resources and the amount of resources used. The decision parameter data includes: the decision thresholds of key nodes and the operational logic. The key operational data are normalized and quantified to generate a matrix of historical best cases.
8. The mine emergency rescue command training and guidance method based on a rule engine according to claim 7, characterized in that, The mine emergency rescue and response process is generated based on a dynamic response process guidance mechanism. The specific method is as follows: Based on the dynamic handling process guidance mechanism, a historical best case matrix and a user operation matrix are generated. The method for generating the user operation matrix is the same as that for generating the historical best case matrix. The user operation matrix is generated through the user's mine emergency rescue and handling process. The deviation between the user operation matrix and the historical best case matrix is calculated. The difference data with the deviation calculation result greater than the threshold is obtained. The corresponding search is performed in the case knowledge graph based on the difference data to obtain targeted improvement suggestions. Based on the improvement suggestions, a mine emergency rescue and response process was generated for training and guidance.
9. A rule-based mine emergency rescue command training and guidance system, which conducts mine emergency rescue command training and guidance based on the method described in claim 1, characterized in that, It includes a data storage module, a rule generation module, a template creation module, a process generation module, and a closed-loop optimization module; The data storage module is used to store historical mine rescue cases, a knowledge graph of historical rescue cases, disaster characteristic parameters, and environmental characteristic parameters; and to generate a case database based on disaster and environmental characteristics. The rule generation module is used to obtain the clauses of the mine rescue regulations and build a rule base based on the case database and the clauses of the mine rescue regulations, including a rescue rule sub-library, a regulation rule font library and a physical simulation rule sub-library; The template creation module is used to generate prefabricated typical case process templates for different mine disasters; The process generation module is used to load the corresponding rules and physical parameters based on the pre-made typical case process template selected by the user, and generate a mine emergency rescue and disposal process. The closed-loop optimization module is used to compare the differences between user operations and historical best cases after training, establish a difference analysis matrix, and combine the difference data with relevant knowledge in the rule base and case knowledge graph to generate targeted improvement suggestions.