A mine accident emergency rescue simulation system and method

Abstract

The present invention relates to the field of accident emergency rescue technology, and specifically to a mine accident emergency rescue simulation system and method. The present invention first establishes a digital twin model for mine accident emergency rescue simulation and connects to a multi-rescue role collaborative rescue system. Then, by analyzing the entry urgency and basic response efficiency of each type of rescue role, the entry priority of each type of rescue role is obtained; then, Monte Carlo simulation is used to simulate the response benefits of each type of rescue role under different random variables; finally, a response surface is established based on "disaster evolution-entry timing-response benefit". By maximizing the response benefit and constraining the entry priority, the optimal entry timing of each type of rescue role is obtained. Through the present invention, the handover and entry timing of multiple rescue roles can be dynamically scheduled according to disaster evolution, random variables, etc. during real-time emergency rescue simulation, so that rescue tasks can be carried out in a timely and orderly manner, and the efficiency of multi-role collaborative rescue can be improved.

Description

Technical Field

[0001] The present invention relates to the technical field of surveying and mapping data analysis and entry, and in particular to a mine accident emergency rescue simulation system and method. Background Art

[0002] Mining accidents often occur underground, in complex environments prone to secondary disasters such as gas explosions, roof falls, and water seepage. These accidents often result in multiple casualties, making rescue difficult and time-sensitive, making advance emergency response drills crucial. Modern mine emergency response drill systems rely on high-precision sensor networks and IoT technologies to achieve comprehensive awareness. For example, an open-pit mine in the Inner Mongolia Autonomous Region deployed a "Mine Safety Black Box System" that integrates 4,128 smart sensors. Leveraging technologies such as digital twins, edge computing, and knowledge graphs, it constructed an accident evolution model containing 270,000 association rules, enabling the generation of dynamic evacuation routes within 120 seconds of an accident.

[0003] Improving the rescue system through simulation can improve accident rescue efficiency. However, existing emergency rescue simulation systems lack a dynamic response mechanism for the coordinated scheduling of multiple rescue roles. When three or more rescue teams are involved simultaneously, the execution efficiency of the plan drops by approximately 41% compared to the ideal value. This is because the static planning and scheduling algorithm is decoupled from the dynamic risk field, resulting in a disconnect between the simulation process and the real-time changes in the actual rescue. This is especially true when multiple rescue roles are involved in the on-site rescue mission, which involves a large number of random variables. Failure to accurately predict the entry timing of different rescue roles can cause chaos at the rescue site, directly affecting the overall rescue progress and efficiency. Summary of the Invention

[0004] In order to solve the above technical problems, the purpose of the present invention is to provide a mine accident emergency rescue simulation system and method.

[0005] According to a first aspect of an embodiment of the present invention, a mine accident emergency rescue simulation method is provided, the method comprising:

[0006] Establish a digital twin model for mine accident emergency rescue simulation, introduce historical sample database, and obtain the disaster evolution process;

[0007] Integrate a multi-rescue role collaborative rescue system into the digital twin model, combine historical sample data, analyze the entry urgency and basic response efficiency of each type of rescue role, and obtain the entry priority of each type of rescue role;

[0008] Random variables are introduced into the digital twin model. Combined with historical sample data, the impact of each type of random variable on the rescue effect of each type of rescue role at different entry times is analyzed. The positive impact factor of the completion of the preceding task on the subsequent entry of the rescue role under each type of random variable is analyzed. The loss rate of the entry delay rate on the completion of the core responsibilities of the rescue role under each type of random variable is analyzed. Monte Carlo simulation is then used to evaluate the response benefits of each type of rescue role under the random variable.

[0009] A response surface combining “disaster evolution-entry timing-response benefit” is established. By maximizing the response benefit and constraining the entry priority, the optimal entry timing for each type of rescue role is obtained.

[0010] In some embodiments of the present invention, historical sample data is combined to analyze the urgency of each type of rescue role entering the scene, including:

[0011] The LSTM model is trained based on the task execution time and environmental parameters of each type of rescue role in the historical sample data to predict the completion time of each type of rescue role and obtain the predicted completion time of each type of rescue role.

[0012] Obtaining the expected completion time for each type of rescue role based on the real-time updated environmental parameters and input emergency events in the digital twin model;

[0013] Obtaining a predecessor task delay rate of each type of rescue role in time sequence according to the predicted completion time and the expected completion time;

[0014] Set the entry weight for each type of rescue role to be handed over;

[0015] The entry urgency of each type of rescue role is obtained by combining the predecessor task delay rate and the entry weight.

[0016] In some embodiments of the present invention, the basic response efficiency of each type of rescue role is analyzed in combination with historical sample data, including:

[0017] Based on historical sample data, the average entry time interval of each type of rescue role is obtained;

[0018] Through the digital twin model, the entry preparation time and entry path time of each type of rescue role are obtained;

[0019] The basic response efficiency of each type of rescue role is obtained by combining the average entry time interval, the entry preparation time, and the entry path time.

[0020] In some embodiments of the present invention, random variables are introduced into the digital twin model, and combined with historical sample data, the impact of each type of random variable on the rescue effect at different entry times of each type of rescue role is analyzed, including:

[0021] Random variables are introduced into the digital twin model, and combined with historical sample data, the sensitivity of the rescue effect to each type of random variables and the volatility of each type of random variables are analyzed to obtain the degree of influence of each type of random variable on the rescue effect at different entry times of each type of rescue role.

[0022] In some embodiments of the present invention, random variables are introduced into the digital twin model, and combined with historical sample data, the positive impact factors of the completion of the preceding task on the subsequent on-site rescue role under each type of random variable are analyzed, including:

[0023] Random variables are introduced into the digital twin model. Based on historical sample data, the Pearson correlation coefficient of the predecessor task completion data set and the subsequent rescue role entry response speed data set for each type of random variable is calculated to obtain the positive impact factor of the predecessor task completion on the subsequent rescue role under each type of random variable.

[0024] In some embodiments of the present invention, random variables are introduced into the digital twin model, and combined with historical sample data, the loss rate of the entry delay rate on the completion of the core responsibilities of the rescue role under each type of random variable is analyzed, including:

[0025] Random variables are introduced into the digital twin model. Combined with historical sample data, the range of the core responsibility completion data set of each type of rescue role under each type of random variable is calculated, and the ratio of the range of the entry delay rate data set of each type of rescue role under each type of random variable is calculated, so as to obtain the loss rate of the entry delay rate to the core responsibility completion of the rescue role under each type of random variable.

[0026] In some embodiments of the present invention, a digital twin model for mine accident emergency rescue simulation is established, and a historical sample database is introduced to obtain the disaster evolution process, including:

[0027] Deploy multiple types of sensors in the mine to obtain real-time monitoring data;

[0028] Based on real-time monitoring data, digital twin technology is used to establish a digital twin model for mine accident emergency rescue simulation;

[0029] Introducing a historical sample database into the digital twin model, constructing a mathematical-mechanical coupling model of the disaster, and combining it with real-time monitoring data to invert the disaster process and obtain the disaster evolution process;

[0030] Edge computing is used to solve risk field parameters in real time, and dynamically coupled with the path planning algorithm to obtain a previewed escape route.

[0031] In some embodiments of the present invention, the multi-rescue role collaborative rescue system includes a command center, an emergency rescue team, a fire and disaster prevention team, a medical rescue team, a warning and evacuation team, a material supply team, an environmental monitoring team, and an information notification team.

[0032] According to a second aspect of an embodiment of the present invention, a mine accident emergency rescue simulation system is provided, the system comprising:

[0033] Model building module, used to build a digital twin model for mine accident emergency rescue simulation;

[0034] The optimal entry timing deduction management module is used to control and manage the optimal entry timing deduction process for each type of rescue role;

[0035] The user-side operation module is used to conduct virtual drills of the drill plan using human-computer interactive technology.

[0036] In some embodiments of the present invention, the optimal entry timing deduction management module includes:

[0037] An entry priority analysis unit is used to integrate a multi-rescue role collaborative rescue system into the digital twin model, analyze the entry urgency and basic response efficiency of each type of rescue role in combination with historical sample data, and obtain the entry priority of each type of rescue role;

[0038] A response benefit analysis unit is used to introduce random variables into the digital twin model, and combine historical sample data to analyze the impact of each type of random variable on the rescue effect of each type of rescue role at different entry times, analyze the positive impact factor of the completion of the preceding task on the subsequent entry of the rescue role under each type of random variable, and analyze the loss rate of the entry delay rate on the completion of the core responsibilities of the rescue role under each type of random variable, so as to obtain the response benefit of each type of rescue role under each type of random variable;

[0039] The optimal entry timing acquisition unit is used to establish a response surface based on the combination of "disaster evolution-entry timing-response benefit" to obtain the optimal entry timing for each type of rescue role by maximizing the response benefit and constraining the entry priority.

[0040] Compared with the existing technology, the mine accident emergency rescue simulation system and method provided by the present invention has the following beneficial effects:

[0041] The present invention first establishes a digital twin model for mine accident emergency rescue simulation, and connects to a multi-rescue role collaborative rescue system, combining historical sample data; then, by analyzing the entry urgency and basic response efficiency of each type of rescue role, the entry priority of each type of rescue role is obtained; then, by analyzing the impact of each type of random variable on the rescue effect of each type of rescue role at different entry times, analyzing the positive impact factor of the completion of the preceding task on the subsequent entry of the rescue role under each type of random variable, and analyzing the loss rate of the entry delay rate on the completion of the core responsibilities of the rescue role under each type of random variable, and then using Monte Carlo simulation to evaluate the response benefit of each type of rescue role under random variables; finally, a response surface based on "disaster evolution-entry timing-response benefit" is established, and by maximizing the response benefit and constraining the entry priority, the optimal entry time of each type of rescue role is obtained. Through the present invention, during real-time emergency rescue simulation, the entry timing of multiple rescue roles can be dynamically scheduled according to disaster evolution, random variables, etc., effectively avoiding the disconnection between the simulation process and the real-time changes of the actual rescue, enabling the rescue task to be carried out in a timely and orderly manner, and greatly improving the efficiency of multi-role collaborative rescue. BRIEF DESCRIPTION OF THE DRAWINGS

[0042] In order to more clearly illustrate the technical solutions and advantages of the embodiments of the present invention or the prior art, the following briefly introduces the drawings required for use in the embodiments or the prior art descriptions. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0043] Figure 1 A schematic diagram of the basic process of a mine accident emergency rescue simulation method provided by one embodiment of the present invention;

[0044] Figure 2 A schematic diagram of the operation flow of an accident response system provided by one embodiment of the present invention;

[0045] Figure 3 A schematic diagram of a response surface of "disaster evolution-entry timing-response benefits" provided by one embodiment of the present invention;

[0046] Figure 4 A schematic diagram of the basic composition of a mine accident emergency rescue simulation system provided by one embodiment of the present invention. DETAILED DESCRIPTION

[0047] To further illustrate the technical means and effectiveness of the present invention to achieve its intended purpose, the following, in conjunction with the accompanying drawings and preferred embodiments, describes in detail a mine accident emergency rescue simulation system and method according to the present invention, including its specific implementation, structure, features, and effectiveness. In the following description, different references to "one embodiment" or "another embodiment" do not necessarily refer to the same embodiment. Furthermore, specific features, structures, or characteristics of one or more embodiments may be combined in any suitable manner.

[0048] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by those skilled in the art to which this invention pertains. Terms such as "comprises," "comprising," or any other variations thereof are intended to encompass non-exclusive inclusion, such that a circuit structure, article, or device comprising a series of elements includes not only those elements but also other elements not explicitly listed, or elements inherent to such article or device. In the absence of further limitations, the phrase "comprising a ..." to define an element does not preclude the presence of other identical elements in the article or device comprising the element.

[0049] The specific scheme of a mine accident emergency rescue simulation system and method provided by the present invention is described in detail below with reference to the accompanying drawings.

[0050] See also Figure 1 , which shows the basic process of a mine accident emergency rescue simulation method provided by an embodiment of the present invention.

[0051] like Figure 1 As shown, an embodiment of the present invention provides a mine accident emergency rescue simulation method, which specifically includes:

[0052] S100: Establish a digital twin model for mine accident emergency rescue simulation, introduce a historical sample database, and obtain the disaster evolution process.

[0053] Establish a digital twin model for mine accident emergency rescue simulation, introduce historical sample database, and obtain the disaster evolution process. Further analysis includes:

[0054] First, multiple types of intelligent sensors are deployed in the mine, such as gas concentration, temperature, micro-seismicity, displacement, etc. The sampling frequency is increased to 200ms / level, and a full-factor perception network is built to obtain real-time monitoring data.

[0055] Then, based on real-time monitoring data, digital twin technology is used to construct a three-dimensional virtual mine accident site, integrating spatial (three-dimensional geological model) + time dimension data, dynamically updating the tunnel topology structure, obstacle distribution, gas diffusion path and temperature field changes, etc., to obtain a digital twin model for mine accident emergency rescue simulation.

[0056] In addition, a historical sample database is introduced into the digital twin model, such as a disaster model database driven by multi-source data. Based on more than 3,000 characteristic parameters of 128 accident cases, a mathematical-mechanical coupling model of disasters such as gas explosion, water seepage, and roof collapse is constructed. The disaster process is inverted by combining real-time monitoring data to obtain the disaster evolution process.

[0057] Finally, edge computing is used to solve risk field parameters in real time, such as gas concentration gradient, flue gas conditions, structural stress thresholds, etc., and dynamically coupled with the path planning algorithm to avoid the disconnection between static paths and risks and obtain a previewed escape route.

[0058] All rehearsed escape routes can be obtained through the digital twin model. There are generally fewer routes to choose from in a mine. Existing technology can already quickly generate dynamically adjustable escape routes based on the evolution of on-site disasters. When an accident occurs, the personnel in the mine can evacuate in an orderly manner according to the simulation drill content, which can greatly increase the probability of escape. Existing technology is not the focus of this application and will not be elaborated on.

[0059] During the simulation, the locations of miners, fires, and emergency destinations (refuge rooms and shafts) were randomized. It should be noted that in a real underground mine, the locations of workstations and safe havens are known, and all evacuation path lengths will adhere to the limits set by safety regulations.

[0060] After receiving the accident report, the mine dispatch room will immediately start the emergency plan and notify the command center members. Usually, the command center will be established within 15-30 minutes after the accident. The center will be responsible for overall command of rescue, coordination of resource dispatch, approval of rescue plan, reporting of rescue progress, etc. The operation process of the accident response system is as follows: Figure 2 shown.

[0061] S200: Integrate a multi-rescue role collaborative rescue system into the digital twin model, combine historical sample data, analyze the entry urgency and basic response efficiency of each type of rescue role, and obtain the entry priority of each type of rescue role.

[0062] The digital twin model for mine accident emergency rescue simulation established in step S100 can generate all evacuation paths, but it lacks a dynamic response mechanism for the coordinated scheduling of multiple rescue roles. When more than three rescue teams are present, the execution efficiency of the plan drops by approximately 41% compared to the ideal value. This is because the static planning and scheduling algorithm is decoupled from the dynamic risk field changes, resulting in a disconnect between the simulation process and the real-time changes in the actual rescue. This is especially true when multiple rescue roles are involved in the rescue mission, as a large number of random variables are present. Failure to properly predict the entry timing of different rescue roles can cause chaos at the rescue site, directly impacting the overall rescue progress and efficiency.

[0063] Therefore, in an embodiment of the present invention, by integrating a multi-rescue role collaborative rescue system into the digital twin model and combining historical sample data, the entry urgency and basic response efficiency of each type of rescue role are analyzed to obtain the entry priority of each type of rescue role.

[0064] The digital twin model incorporates a collaborative rescue system with multiple rescue roles. Specifically, a collaborative decision-making sandbox is added to the digital twin model to support the integration of a collaborative rescue system with multiple rescue roles. This system includes a command center, emergency rescue team, fire prevention team, medical rescue team, warning and evacuation team, material supply team, environmental monitoring team, and information reporting team. These rescue roles are introduced into the sandbox simulation system, and real-time data is shared through the digital twin platform. Based on the knowledge graph, historical plans, equipment status, and other information are linked.

[0065] Among them, whether the multi-rescue role collaborative rescue system needs to enter the mine, the operating depth limit and the execution content are shown in the following table.

[0066]

[0067] Dynamic scheduling of multiple rescue roles, specifically:

[0068] The exploration team of the emergency rescue team mainly goes to all emergency points and rescue points; emergency points are control nodes of dangerous sources such as gas accumulation areas, water seepage points, fire sources, etc. After the exploration team sends back the on-site data, the fire prevention and disaster prevention team enters the emergency point for control; the rescue point refers to the concentrated area of ​​​​people in distress who have not been evacuated, which is determined by the exploration team through the personnel positioning system, thermal imaging detection system, etc. If the exploration team finds several trapped people in the alley, it will immediately report to the command center, and then lead the trapped people to evacuate along the evacuation route. If there are injured patients, the medical rescue team will send a first aid team carrying hemodialysis equipment according to the vital signs data sent back from the rescue point. The environmental monitoring team dynamically updates the safety environment in the mine based on the existing sensor data in the mine or the environmental data sent back by the exploration team. For example, if the digital twin system detects that the CH4 concentration in a certain area exceeds the limit, it will automatically adjust the evacuation route of the emergency rescue team. The access rights to the accident site are clear. The warning and evacuation team is not allowed to enter the accident site before receiving the order to avoid chaos in the rescue operation. The material supply team dispatches specialists to cooperate with the exploration team to set up supply stations along the proven disaster evacuation routes, including oxygen masks, drinking water, hemostatic bandages, etc. The information notification group cooperates with the command center and is responsible for all communication information management.

[0069] Due to the sudden and emergency nature of mining accidents, continuous deterioration, and limited rescue space, the handover and coordination of rescue personnel are idealized during emergency rescue simulations, and the fault tolerance rate is too high. However, in actual rescue operations, a minor mistake, such as information transmission errors and ambiguous decisions, may cause chaos in the rescue unit. Therefore, the existing system still needs to be optimized in the dynamic scheduling of multiple rescue roles.

[0070] Different rescue roles need to coordinate with other rescue roles when operating. If one rescue role's task is a prerequisite for another, the other rescue roles can only begin their operations after the predecessor task is completed. Poor completion of the predecessor task can delay the arrival of the supporting rescue role or disrupt the arrival of the subsequent rescue role. Unfavorable timing for the arrival of different rescue roles can disrupt the overall rescue mission. For example, a tunnel collapse caused the exploration team to take 20 minutes longer than expected to transmit data. The fire prevention team's delayed arrival caused a fire to spread, potentially requiring an additional 30 minutes to contain the source. The medical rescue team was unable to reach the rescue point due to a collapse or uncontrolled fire, resulting in deteriorating vital signs and reduced survival rates for critically injured personnel. The environmental monitoring team, unable to receive sensor data due to the damaged environment within the mine, misjudged the safe zone, causing the supply team to mistakenly set up a supply station in a collapsed or high-temperature area, resulting in the destruction of supplies. The warning and evacuation team allowed external reinforcements to enter prematurely without receiving confirmation from the exploration team that the route was safe, causing rescue chaos.

[0071] Therefore, based on the digital twin model connected to the collaborative rescue system of multiple rescue roles and combined with historical sample data, the entry urgency and basic response efficiency of each type of rescue role are analyzed to obtain the entry priority of each type of rescue role.

[0072] Among them, combined with historical sample data, the urgency of each type of rescue role's entry is analyzed, further including:

[0073] First, the LSTM model is trained based on the task execution time and environmental parameters (such as tunnel complexity and gas concentration) of each type of rescue role in the historical sample data to predict the completion time of each type of rescue role and obtain the predicted completion time of each type of rescue role.

[0074] At the same time, based on the real-time updated environmental parameters and input emergencies in the digital twin model, the expected completion time of each type of rescue role is obtained.

[0075] Then, based on the predicted completion time and expected completion time, the delay rate of the predecessor task of each type of rescue role in the time sequence is obtained:

[0076]

[0077] Where, Indicates the The delay rate of the predecessor task of the rescue-like role in time sequence; Indicates the Predicted completion time for rescue-like roles; Indicates the The expected completion time for rescue-like roles.

[0078] Additionally, entry weights are assigned to each type of rescue role to be handed over. Specifically, these weights are set based on the mission's urgency (e.g., the vital signs of the person in distress) and environmental risks (e.g., the probability of a gas explosion). These weights range from 0 to 1 (they can be manually set after evaluation by the command center or based on historical impact assessments). These weights are then input into the digital twin model, adding a mechanism for dynamically adjusting the order of entry. For example, if the trapped person's vital signs are critical, the medical rescue team may need to enter earlier, increasing their entry weight. If the gas spread at a dangerous point is rapid, the fire and disaster prevention team must be dispatched immediately, increasing their entry weight. When an emergency deployment is required, the reconnaissance team of the previous emergency rescue team will need to temporarily adjust their mission to prioritize establishing entry conditions for the incoming rescue roles, such as clearing passages and placing markers along the route.

[0079] Then, combining the predecessor task delay rate and entry weight, we can determine the entry urgency for each rescue role. Specifically, the product of each rescue role's entry weight and its predecessor task delay rate is recorded as the entry urgency. A larger product indicates a higher entry urgency for that rescue role.

[0080] Combined with historical sample data, the basic response efficiency of each rescue role is analyzed, further including:

[0081] First, based on historical sample data, when the front-end rescue role meets the triggering conditions, the average entry time interval of each type of rescue role is obtained.

[0082] In addition, through the simulation results of the digital twin model, the entry preparation time of each type of rescue role (including command transmission time, material preparation time, etc.) and the entry path time required to reach the target location in the mine (rescue point or emergency point) from the current location are obtained.

[0083] Furthermore, the basic response efficiency of each rescue role is derived by combining the average entry time interval, entry preparation time, and entry path time. Specifically, the basic response efficiency of each rescue role is calculated as the ratio of the average entry time interval of the rescue role to the sum of the entry path time and entry preparation time of the rescue role simulated in the digital twin model.

[0084] Finally, combining the entry urgency and basic response efficiency of each type of rescue role, the entry priority of each type of rescue role is obtained as follows:

[0085]

[0086]

[0087] Where, Indicates the Entry priority for rescue-like characters; Indicates the The urgency of the rescue-like role's entry; Indicates the The average time interval between the entry of rescue-like characters; Indicates the The entry path time required for the rescue-like character to reach the target location in the mine from the current location; Indicates the Preparation time for the rescue-like characters; represents the linear normalization function; Indicates the Delay rate of predecessor tasks for rescue-like roles; Indicates the The entry weight of rescue-like roles.

[0088] Represents the basic response efficiency of each type of rescue role. The larger the ratio, the higher the basic response efficiency of this type of rescue role under the current simulation results, that is, it can enter the scene quickly after the command is issued, so the rescue role has a higher priority for entry; conversely, the smaller the ratio, the lower the basic response efficiency of this type of rescue role, so the rescue role has a lower priority for entry; the urgency of entry The larger the value, the higher the entry priority of the corresponding rescue character.

[0089] Entry priority for each type of rescue role The larger the value, the earlier the rescue role is needed to enter the scene, because the simulated accident scene has a higher demand for the rescue role to enter the scene, the shorter the reserved response time is, and priority entry when the rescue role has a high basic response can be more conducive to shortening the overall rescue time.

[0090] When the introduced simulation variables change, the entry priority of different rescue roles will also change accordingly. It should be noted that all simulation variables introduced in the digital twin model need to obey the Gaussian distribution to make the simulation results random.

[0091] S300: Introducing random variables into the digital twin model and combining historical sample data, we analyze the impact of each type of random variable on the rescue effect of each type of rescue role at different entry times, analyze the positive impact factor of the completion of the preceding task on the subsequent entry of the rescue role under each type of random variable, and analyze the loss rate of the entry delay rate on the completion of the core responsibilities of the rescue role under each type of random variable, and obtain the response benefit of each type of rescue role under each type of random variable.

[0092] By introducing random variables into the digital twin model and combining historical sample data, we analyze the impact of each type of random variable on the rescue effect of each type of rescue role at different entry times, analyze the positive impact factor of the completion of the previous task on the subsequent entry of the rescue role under each type of random variable, and analyze the loss rate of the entry delay rate on the completion of the core responsibilities of the rescue role under each type of random variable, and obtain the response benefit of each type of rescue role under each type of random variable.

[0093] Introducing random variables into the digital twin model, combined with historical sample data, analyzes the degree of influence of each type of random variable on the rescue effect at different entry times for each type of rescue role. Specifically, random variables are introduced into the digital twin model, and the probability of each type of random variable occurring during the simulation is set to follow a Gaussian distribution, such as information transmission delay, equipment failure, etc.; then, combined with historical sample data, the rescue effect of each type of random variable at different entry times for each type of rescue role is simulated; based on the simulation results, the sensitivity of the rescue effect to each type of random variable is analyzed, as well as the volatility of each type of random variable, to obtain the degree of influence of each type of random variable on the rescue effect at different entry times for each type of rescue role. The quantitative formula for the influence of each type of random variable on the rescue effect at different entry times for each type of rescue role is constructed as follows:

[0094]

[0095] Where, represents a random variable For the first The degree of influence of rescue effect under different entry timings of rescue-like characters, i.e., the dependent random variable The total change in rescue effectiveness caused by various factors (such as information delays and equipment failures); 、 They represent the start time and end time of the continuous simulation process respectively; Indicates the Rescue effect of rescue-like characters For random variables sensitivity (partial derivatives); represents a random variable The standard deviation of , reflecting its volatility; Indicates arrive Integration is performed within the time range.

[0096] Used to quantify the dynamic impact of the fluctuation of random variables on the rescue effect. The integral represents the cumulative effect over time, and the standard deviation Amplified sensitivity , and finally the influence of random variables on the rescue effect is obtained.

[0097] By introducing random variables into the digital twin model and combining them with historical sample data, we analyze the positive impact factors of the completion of predecessor tasks on the subsequent rescue roles under each type of random variable. Specifically, we introduce random variables into the digital twin model and calculate the Pearson correlation coefficient of the predecessor task completion (core responsibility completion) data set of each type of rescue role under each type of random variable and the subsequent rescue role entry response speed data set based on historical sample data. We obtain the positive impact factors of the predecessor task completion on the subsequent rescue roles under each type of random variable. For example, the marking accuracy of the exploration team increases by 10%, and the response speed of the medical rescue team increases by 15%.

[0098] Random variables are introduced into the digital twin model, and combined with historical sample data, the loss rate of the entry delay rate to the completion of the core responsibilities of the rescue role under each type of random variable is analyzed. Specifically, the range of the core responsibility completion data set of each type of rescue role under each type of random variable is calculated, and the ratio of the range of the entry delay rate data set of each type of rescue role under each type of random variable is calculated. The loss rate of the entry delay rate to the completion of the core responsibilities of the rescue role under each type of random variable is obtained. For example, if the fire prevention and disaster prevention team arrives 1 minute late, the probability of a gas explosion will increase by 8%.

[0099] Combining the random variable simulation results and historical sample data, Monte Carlo simulation is used to evaluate the response benefits of each rescue role under random variables. Random variables are obtained in the simulation results to determine the degree of influence of each type of random variable on the rescue contribution of each type of role; the positive impact factor of the completion of each type of rescue role's pre-task obtained from the historical sample data of each random variable on the rescue role; and the loss rate of the entry delay rate of each type of rescue role on the completion of the core responsibility of the rescue role obtained from the historical sample data of each random variable. Then, the response benefits of each type of rescue role under the random variables are:

[0100]

[0101] Where, Indicates the The response benefits of the rescue-like role under the influence of all random variables; represents the total number of random variables; represents a random variable The intervention weight (range 0-1, adjusted according to the actual accident scenario); Representative Rescue role, Responding to the rescue role performing its predecessor mission; represents a random variable The impact of the rescue contribution of the rescue role in performing the preceding task is that the random variable has an inverse proportional relationship to the rescue contribution; Indicates the degree of completion of the predecessor task Positive impact factors of rescue-like roles; Indicates that the random variable Next The delay rate of entry of a rescue role is the loss rate of the completion of the core responsibilities of the rescue role.

[0102] represents a random variable The impact on the previous task further weakens the impact of the previous task completion on the next task. The positive impact factor of the rescue role makes the The response benefits of the class role are reduced; Indicates the Class rescue role in random variables The response yields under The weighted average of the response benefits of the rescue role under all random variables is obtained. The final response payoff of the rescue-like character under all random variables.

[0103] It should be noted that when the random variables introduced in the digital twin model are different, the final response benefits of each type of rescue role under different random variables are different.

[0104] S400: Establish a response surface based on the combination of "disaster evolution-entry timing-response benefit" to obtain the optimal entry timing for each type of rescue role by maximizing the response benefit and constraining the entry priority.

[0105] Through steps S200 and S300, we obtain the entry priorities of different rescue roles and their response benefits under random variables. From this, we construct a response surface combining "disaster evolution, entry timing, and response benefit." By maximizing the response benefit and constraining the entry priority, we determine the optimal entry timing for each rescue role.

[0106] First, using the response surface methodology, a response surface was established based on the “disaster evolution-entry timing-response benefit” combination. Specifically, it includes:

[0107] The X-axis represents the disaster evolution rate. Based on the real-time changes in the core disaster parameters such as harmful gas diffusion, water level rise, and mine collapse speed introduced in the digital twin model, the overall evolution rate of the disaster is quantified. Any scoring method is acceptable.

[0108] The Y-axis represents the entry timing. The original entry time set in the plan is taken as the 0 value. The actual entry time is the amount of advance or delay compared to the 0 value, which is the entry timing.

[0109] The Z-axis represents the response benefit, and the size of the response benefit represents the probability that the rescue role is affected by the random variable in different accident scenarios.

[0110] Combining historical sample data with multiple simulation results of the digital twin model, the obtained X, Y, and Z variables are used to construct a nonlinear response surface using Gaussian process regression, such as Figure 3 shown.

[0111] Then, by maximizing the response benefit and constraining the entry priority, the optimal entry time for each type of rescue role is obtained. Specifically, it includes:

[0112] In the digital twin model, enter all constraints:

[0113] Role entry order dependency: For example, the emergency team must enter 10 minutes earlier than the firefighting team;

[0114] Safety threshold: The firefighting team is allowed to enter only when the gas concentration is less than 5%;

[0115] Resource limit: The number of characters entering the game at the same time is less than or equal to 3;

[0116] Route feasibility: Landslide and fire areas are prohibited from passing during simulation;

[0117] …

[0118] Then, a genetic algorithm is used to find the optimal entry time for each role in the response surface, that is, the highest response benefit value (Z) of the rescue role under a specific disaster evolution rate (X) and entry timing (Y) predicted by the Gaussian process regression model.

[0119] In this process, a penalty constraint is added to the Y-axis, that is, the entry timing sought by the rescue character. The further it deviates from the value of its entry priority function, the greater the penalty.

[0120] The penalty term is calculated as:

[0121]

[0122] Where, Indicates the The Y-axis penalty value for rescue-like characters; Indicates the The initial value of the Y axis of the rescue character class; Indicates the Entry priority for rescue-like characters; represents the linear normalization function.

[0123] For the The entry timing of the rescue-like role in the response surface Normalized deviation and its entry priority The absolute value of the difference represents the penalty term, that is, the rescue character's entry timing cannot deviate too much from its entry priority, otherwise the penalty will increase.

[0124] Therefore, the objective function is:

[0125]

[0126] Where, represents the objective function value; Indicates the The Z-axis value of the rescue character, that is, Response benefits corresponding to rescue-like roles; Indicates the The Y-axis penalty value for rescue-like characters; Represents the maximum function.

[0127] When the objective function The entry timing corresponding to the maximum value is the current optimal entry timing for the rescue role. The entry timing sought has the maximum response benefit and the minimum entry priority penalty under a specific disaster evolution rate.

[0128] Before entering the scene, each rescue role inputs its current disaster evolution rate, constraints, and all real-time simulation variables into the digital twin model to obtain the optimal entry time for the rescue role. This can dynamically adjust the collaborative smoothness of multiple rescue roles, ensure smooth handover, and reduce rescue risks.

[0129] Based on the same inventive concept as the above-mentioned system, this embodiment also provides a mine accident emergency rescue simulation system.

[0130] See also Figure 4 , which shows the basic composition of a mine accident emergency rescue simulation system provided by an embodiment of the present invention.

[0131] like Figure 4 As shown, a mine accident emergency rescue simulation system includes:

[0132] The model building module 10 is used to establish a digital twin model for mine accident emergency rescue simulation, introduce a historical sample database, and obtain the disaster evolution process;

[0133] The optimal entry timing deduction management module 20 is used to control and manage the optimal entry timing deduction process for each type of rescue role;

[0134] The user-side operation module 30 is used to conduct a virtual drill on the drill plan using human-computer interactive technology.

[0135] Furthermore, the optimal entry timing deduction management module 20 includes:

[0136] The entry priority analysis unit 21 is used to integrate the multi-rescue role collaborative rescue system into the digital twin model, analyze the entry urgency and basic response efficiency of each type of rescue role in combination with historical sample data, and obtain the entry priority of each type of rescue role;

[0137] The response benefit analysis unit 22 is used to introduce random variables into the digital twin model and, in combination with historical sample data, analyze the impact of each type of random variable on the rescue effect of each type of rescue role at different entry times, analyze the positive impact factor of the completion of the preceding task on the subsequent entry of the rescue role under each type of random variable, and analyze the loss rate of the entry delay rate on the completion of the core responsibilities of the rescue role under each type of random variable, so as to obtain the response benefit of each type of rescue role under each type of random variable;

[0138] The optimal entry timing acquisition unit 23 is used to establish a response surface based on the combination of "disaster evolution-entry timing-response benefit" to obtain the optimal entry timing for each type of rescue role by maximizing the response benefit and constraining the entry priority.

[0139] It should be noted that the order in which the embodiments of the present invention are described above is for illustrative purposes only and does not necessarily represent the superiority or inferiority of the embodiments. The processes depicted in the accompanying drawings do not necessarily require the specific order or sequential order shown to achieve the desired results. In certain embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0140] The various embodiments in this specification are described in a progressive manner, and the same or similar parts between the various embodiments can be referred to each other. Each embodiment focuses on the differences from other embodiments.

Claims

1. A mine accident emergency rescue simulation method, characterized in that: The method comprises: Establish a digital twin model for mine accident emergency rescue simulation, introduce historical sample database, and obtain the disaster evolution process; Integrate a multi-rescue role collaborative rescue system into the digital twin model, combine historical sample data, analyze the entry urgency and basic response efficiency of each type of rescue role, and obtain the entry priority of each type of rescue role; Random variables are introduced into the digital twin model. Combined with historical sample data, the impact of each type of random variable on the rescue effect of each type of rescue role at different entry times is analyzed. The positive impact factor of the completion of the preceding task on the subsequent entry of the rescue role under each type of random variable is analyzed. The loss rate of the entry delay rate on the completion of the core responsibilities of the rescue role under each type of random variable is analyzed. Monte Carlo simulation is then used to evaluate the response benefits of each type of rescue role under the random variable. Establish a response surface based on the "disaster evolution-entry timing-response benefit" combination. By maximizing the response benefit and constraining the entry priority, the optimal entry timing for each type of rescue role is obtained. Among them, random variables are introduced into the digital twin model, and combined with historical sample data, the sensitivity of the rescue effect to each type of random variable is analyzed, as well as the volatility of each type of random variable, to obtain the degree of influence of each type of random variable on the rescue effect under different entry times of each type of rescue role; Introducing random variables into the digital twin model, based on historical sample data, calculates the Pearson correlation coefficient between the pre-task completion dataset of each type of rescue role and the subsequent rescue role entry response speed dataset under each type of random variable, and obtains the positive impact factor of the pre-task completion on the subsequent rescue role under each type of random variable; Random variables are introduced into the digital twin model. Combined with historical sample data, the range of the core responsibility completion data set of each type of rescue role under each type of random variable is calculated, and the ratio of the range of the entry delay rate data set of each type of rescue role under each type of random variable is calculated, so as to obtain the loss rate of the entry delay rate to the core responsibility completion of the rescue role under each type of random variable.

2. The mine accident emergency rescue simulation method according to claim 1, characterized in that: Combined with historical sample data, the urgency of each type of rescue role entering the scene is analyzed, including: The LSTM model is trained based on the task execution time and environmental parameters of each type of rescue role in the historical sample data to predict the completion time of each type of rescue role and obtain the predicted completion time of each type of rescue role. Obtaining the expected completion time for each type of rescue role based on the real-time updated environmental parameters and input emergency events in the digital twin model; Obtaining a predecessor task delay rate of each type of rescue role in time sequence according to the predicted completion time and the expected completion time; Set the entry weight for each type of rescue role to be handed over; The entry urgency of each type of rescue role is obtained by combining the predecessor task delay rate and the entry weight.

3. The mine accident emergency rescue simulation method according to claim 2, characterized in that: Combined with historical sample data, the basic response efficiency of each rescue role is analyzed, including: Based on historical sample data, the average entry time interval of each type of rescue role is obtained; Through the digital twin model, the entry preparation time and entry path time of each type of rescue role are obtained; The basic response efficiency of each type of rescue role is obtained by combining the average entry time interval, the entry preparation time, and the entry path time.

4. The mine accident emergency rescue simulation method according to claim 1, characterized in that: A digital twin model for mine accident emergency rescue simulation was established, and a historical sample database was introduced to obtain the disaster evolution process, including: Deploy multiple types of sensors in the mine to obtain real-time monitoring data; Based on real-time monitoring data, digital twin technology is used to establish a digital twin model for mine accident emergency rescue simulation; Introducing a historical sample database into the digital twin model, constructing a mathematical-mechanical coupling model of the disaster, and combining it with real-time monitoring data to invert the disaster process and obtain the disaster evolution process; Edge computing is used to solve risk field parameters in real time, and dynamically coupled with the path planning algorithm to obtain a previewed escape route.

5. The mine accident emergency rescue simulation method according to claim 1, characterized in that: The multi-rescue role collaborative rescue system includes a command center, an emergency rescue team, a fire and disaster prevention team, a medical rescue team, a warning and evacuation team, a material supply team, an environmental monitoring team, and an information notification team.

6. A mine accident emergency rescue simulation system, characterized in that: The system is used to implement the steps of the mine accident emergency rescue deduction method according to any one of claims 1 to 5, and the system includes: Model building module, used to build a digital twin model for mine accident emergency rescue simulation; The optimal entry timing deduction management module is used to control and manage the optimal entry timing deduction process for each type of rescue role; The user-side operation module is used to conduct virtual drills of the drill plan using human-computer interactive technology.

7. The mine accident emergency rescue simulation system according to claim 6, characterized in that: The optimal entry timing deduction management module includes: An entry priority analysis unit is used to integrate a multi-rescue role collaborative rescue system into the digital twin model, analyze the entry urgency and basic response efficiency of each type of rescue role in combination with historical sample data, and obtain the entry priority of each type of rescue role; A response benefit analysis unit is used to introduce random variables into the digital twin model, and combine historical sample data to analyze the impact of each type of random variable on the rescue effect of each type of rescue role at different entry times, analyze the positive impact factor of the completion of the preceding task on the subsequent entry of the rescue role under each type of random variable, and analyze the loss rate of the entry delay rate on the completion of the core responsibilities of the rescue role under each type of random variable, so as to obtain the response benefit of each type of rescue role under each type of random variable; The optimal entry timing acquisition unit is used to establish a response surface based on the combination of "disaster evolution-entry timing-response benefit" and obtain the optimal entry timing for each type of rescue role by maximizing the response benefit and constraining the entry priority.

Images