A method and system for optimizing subway station emergency response plans under fire conditions

Abstract

The present invention belongs to the field of fire emergency command technology and specifically discloses a method and system for optimizing subway station emergency response plans under fire conditions. The method includes: constructing a hierarchical timed colored Petri net model to simulate different scenarios of subway fire emergency response using CPN Tools; establishing a multi-objective optimization model for subway station fire emergency response with the goals of minimizing emergency response time and the number of firefighters; based on the simulated data set, using the Skyline Operator algorithm to solve the multi-objective optimization model to obtain the Pareto front, and using a fuzzy-based ideal point method to obtain the optimal subway station emergency response plan. The present invention can simultaneously optimize the time control and resource allocation of subway station fire emergency response, providing scientific guidance for real-time decision-making in subway fire emergency response.

Description

Technical Field

[0001] The present invention belongs to the field of fire emergency command technology, and more specifically, relates to a method and system for optimizing emergency response plans for subway stations under fire conditions. Background Art

[0002] Subway systems face significant challenges in safety management, particularly in emergency response to fire incidents. Delayed response time is one of the main reasons for increased fire losses. Therefore, a timely and effective emergency response system is crucial for successful rescue efforts and fire hazard management. With the increasing popularity of the Internet of Things (IoT), wireless sensor networks are increasingly being applied in disaster management. The IoT can provide emergency managers with essential, up-to-date information and enhance first responder capabilities. To optimize the emergency response process, it is necessary to integrate sensing technology, analysis, judgment, decision-making, and action functions into a hybrid approach. This approach enables effective emergency response based on real-time sensor feedback, thereby reducing potential losses caused by fire hazards and helping firefighters determine the most appropriate plan to complete fire rescue and control tasks. Petri nets, as an excellent dynamic discrete system modeling tool, have great potential in the modeling and analysis of fire emergency response. They can be used to define conditions, activate activities and actions, and facilitate coordination among various departments. When the system is large, the Petri net structure can become very complex. Using HTCPN can make Petri net models clearer and more logical. Response time is a critical factor in subway fire emergency response. Reducing fire emergency response time can effectively reduce casualties and property losses. Another core issue in subway emergency response is the rational allocation of emergency resources. Research on human resource allocation in emergency response is relatively underdeveloped. In fact, the level of coordination among firefighters during an emergency impacts the time required to control the incident. Therefore, the allocation of firefighters should be carefully considered by the emergency command department. Shortening emergency response time and rationally allocating firefighters often conflict. Balancing these two objectives is a key issue, which requires multi-objective optimization.

[0003] Based on the above defects and deficiencies, this field urgently needs to propose a method and system for optimizing subway station emergency response plans under fire conditions, filling the gap in the current research on optimizing subway station fire emergency response time control and resource allocation, and thus providing scientific guidance for real-time decision-making in subway fire emergency response. Summary of the Invention

[0004] To address the aforementioned shortcomings or improvements in existing technologies, the present invention provides a method and system for optimizing subway station emergency response plans under fire conditions. This method constructs a Hierarchical Timed Colored Petri Net (HTCPN) model using CPN Tools to simulate different subway fire emergency response scenarios. A multi-objective optimization model for subway station fire emergency response is established, taking into account minimizing emergency response time and the number of firefighters. Based on the simulated dataset, the Skyline Operator is used to solve the multi-objective optimization problem, obtaining a Pareto front, and employing a fuzzy ideal point method to determine the optimal compromise solution. This method can simultaneously optimize both time control and resource allocation for subway station fire emergency response, providing scientific guidance for real-time decision-making in subway fire emergency response.

[0005] To achieve the above objectives, according to one aspect of the present invention, a method for optimizing emergency response plans for subway stations under fire conditions is proposed, which is characterized by comprising:

[0006] S1 constructs a hierarchical timed colored Petri net model to simulate different scenarios of subway fire emergency response using CPN Tools;

[0007] S2 establishes a multi-objective optimization model for subway station fire emergency response with the goal of minimizing emergency response time and the number of firefighters;

[0008] Based on the simulated data set, S3 uses the Skyline Operator algorithm to solve the multi-objective optimization model to obtain the Pareto front, and adopts the fuzzy-based ideal point method to obtain the optimal subway station emergency response plan.

[0009] As a further preferred embodiment, in step S1, an HTCPN model and a subsystem of the HTCPN model are established according to the emergency response process preset in the subway station, and the information collected by the sensors in the subway fire emergency management system is placed in the Petri net through tokens and color states. When the token changes, it means that the state has changed and will enter the next state. The intermediate process is represented by a transition. The execution time of the transition activity is set according to the actual situation. The activation and triggering of the transition and the corresponding tag change are defined by the activation and triggering rules of the actual Petri net respectively. When the required number of tags and color information are present at the input of the transition, the transition will be activated. The activated transition can trigger an event, obtain a specified number of tags from the input, and store a specified number of tags at the output.

[0010] The subsystems of the HTCPN model include a fire extinguishing and rescue subsystem in the rescue section, which connects the location where the rescue team receives the signal and the location where the fire is successfully extinguished and the rescue is completed.

[0011] The subsystem of the HTCPN model also includes a staff fire extinguishing subsystem, which connects the warehouse representing the staff to assemble and the warehouse where the staff extinguish the fire;

[0012] The subsystems of the HTCPN model also include a staff-assisting passenger evacuation subsystem, which is respectively connected to the library representing the completion of broadcast information, the generation of evacuation plan, and the library representing the completion of passenger evacuation.

[0013] As a further preference, the information detected by the sensors is transmitted to the control center via a heterogeneous network. The control center processes and analyzes the collected data and performs operations through feedback components. Specifically, the fire emergency management system in the subway system relies on a sensor network to detect and locate dangers, locate personnel, including fire points, trapped personnel, and firefighters, and establish communication between the environment and the control center. In the HTCPN model, the library is used to represent the active state in the fire emergency management system, the transition represents the active process of the state, and the execution time of the transition is assigned based on experience. The colored marks combined with the Internet of Things represent specific types of information. The state of the HTCPN model is transmitted in the form of real-time data updates from the sensors.

[0014] As a further preference, with the assistance of Petri net modeling software CPN Tools, the established HTCPN model is simulated in different scenarios to obtain the emergency response time and the number of firefighters allocated. Specifically, tokens in the library are used to represent the resources and information required in the emergency response process, transitions are used to represent the action process, and time functions are added to the transitions to represent the time required for the action. In order to obtain simulation data on the emergency response time and the number of firefighters in each team, counting monitors are added to key libraries and transitions before the model is run.

[0015] As further preferred, in step S2, the multi-objective optimization model includes:

[0016]

[0017]

[0018] Among them, F1 is the number of firefighters, F2 is the total emergency response time, and are the number of fire brigade and passenger rescue team respectively, is the number of firefighters in the nth fire brigade, is the number of firefighters in the mth passenger rescue team, and the objective function F2 is obtained by simulating the HTCPN model in CPN Tools;

[0019] The constraints are:

[0020]

[0021]

[0022] in, are the maximum and minimum capacities of the fire-fighting teams, respectively. They are the maximum and minimum capacities of the fire rescue team respectively.

[0023] As a further preferred embodiment, in step S3, based on the data set obtained by CPN Tools, the SkylineOperator algorithm is used to find the Pareto front of the multi-objective optimization problem, including: given a set of n objectives , the Skyline Operator algorithm returns all objects ,make Not affected by any other object Domination, where i, j = 1,…, n.

[0024] As a further preferred method, the fuzzy ideal point method is used to obtain the best subway station emergency response plan, specifically including: calculating the membership function , the i-th objective function among the k Pareto solutions is expressed as:

[0025]

[0026] in, and are the maximum and minimum values of the i-th target respectively;

[0027] For non-dominated solutions, the normalized membership function The optimization is performed using the following weighted sum model:

[0028]

[0029] Where M is the number of non-dominated solutions; is the weight of the i-th objective function; N is the total number of optimization objectives; then, The maximum value It is the optimal solution.

[0030] According to another aspect of the present invention, a system for optimizing emergency response plans for subway stations under fire conditions is provided, comprising:

[0031] The first main control module is used to build a hierarchical timed colored Petri net model to simulate different scenarios of subway fire emergency response using CPN Tools;

[0032] The second main control module is used to establish a multi-objective optimization model for subway station fire emergency response with the goal of minimizing emergency response time and the number of firefighters;

[0033] The third main control module is used to solve the multi-objective optimization model based on the simulation data set using the Skyline Operator algorithm to obtain the Pareto front, and adopt the fuzzy ideal point method to obtain the optimal subway station emergency response plan.

[0034] According to another aspect of the present invention, there is also provided an electronic device, comprising:

[0035] At least one processor, at least one memory and a communication interface; wherein,

[0036] The processor, memory and communication interface communicate with each other;

[0037] The memory stores program instructions that can be executed by the processor, and the processor calls the program instructions to perform the above method.

[0038] According to another aspect of the present invention, a non-transitory computer-readable storage medium is provided. The non-transitory computer-readable storage medium stores computer instructions, and the computer instructions enable the computer to execute the above method.

[0039] In general, the above technical solutions conceived by the present invention have the following technical advantages compared with the existing technology:

[0040] 1. This invention can simultaneously optimize the time control and resource allocation of subway station fire emergency response, providing scientific guidance for real-time decision-making in subway fire emergency response.

[0041] 2. In this invention, due to the complex environment of subway stations, high population density, and high mobility, many departments and personnel collaborate to respond to emergencies. Emergency commanders must coordinate and organize personnel from different departments to jointly participate in the emergency response process, resulting in a complex and large emergency system. When the system is large, the Petri net structure becomes extremely complex. In this case, the use of HTCPN can make the Petri net model clearer and more logical.

[0042] 3. The multi-objective optimization in this invention can simultaneously optimize two or more conflicting objectives. Currently, many Pareto front-based algorithms, such as MOGA, NSGA-II, and SPEA, have been proposed to solve multi-objective optimization problems. However, these algorithms must first establish a relationship between inputs and outputs, rather than working with a dataset. The Skyline Operator algorithm can solve multi-objective optimization problems for datasets and is suitable for simulation-based multi-objective optimization and decision-making. BRIEF DESCRIPTION OF THE DRAWINGS

[0043] Figure 1 A flow chart of a method for designing and evaluating a subway station emergency response plan under fire conditions provided by an embodiment of the present invention;

[0044] Figure 2 A schematic diagram of the structure of a subway station emergency response plan design and evaluation system under fire conditions provided by an embodiment of the present invention;

[0045] Figure 3 A schematic diagram of the physical structure of an electronic device provided by an embodiment of the present invention;

[0046] Figure 4 An emergency rescue management system for a subway station fire provided by an embodiment of the present invention;

[0047] Figure 5 A schematic diagram of the HTCPN model provided in an embodiment of the present invention;

[0048] Figure 6 Initial parameter settings for the HTCPN model provided in the embodiment of the present invention;

[0049] Figure 7 An explanation of the HTCPN model library and its changes provided by the embodiments of the present invention;

[0050] Figure 8 The color meaning of the HTCPN model provided by the embodiment of the present invention;

[0051] Figure 9 The model structure of the T2 subsystem provided in the embodiment of the present invention;

[0052] Figure 10 The model structure of the T6 subsystem provided in the embodiment of the present invention;

[0053] Figure 11 The model structure of the T7 subsystem provided in the embodiment of the present invention;

[0054] Figure 12 The emergency response time distribution histogram, its fitting curve, and cumulative probability distribution provided by the embodiment of the present invention;

[0055] Figure 13 (a) to (d) are steady-state probabilities of each marker provided by an embodiment of the present invention;

[0056] Figure 14 (a) and (b) are the calculation results of the place busy rate and transition utilization rate provided by the embodiment of the present invention respectively;

[0057] Figure 15The actual scatter plot and the LightGBM-based metamodel prediction results provided by the embodiment of the present invention. DETAILED DESCRIPTION

[0058] In order to make the objectives, technical solutions and advantages of the present invention more clearly understood, the present invention is further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely for the purpose of explaining the present invention and are not intended to limit the present invention. In addition, the technical features involved in the various embodiments of the present invention described below may be combined with each other as long as they do not conflict with each other.

[0059] like Figure 1 As shown, an embodiment of the present invention provides a method for optimizing a subway station emergency response plan under fire conditions, including: constructing a hierarchical timed colored Petri net (HTCPN) model to simulate different scenarios of subway fire emergency response using CPN Tools; establishing a multi-objective optimization model for subway station fire emergency response, which considers minimizing emergency response time and minimizing the number of firefighters; based on a simulated data set, using the Skyline Operator to solve the multi-objective optimization problem to obtain the Pareto front, and using a fuzzy-based ideal point method to obtain the best compromise solution (i.e., the optimal solution).

[0060] In the present invention, an HTCPN system and its subsystems are established according to the preset emergency response process of the subway station. Specifically, the information collected by the sensor can be put into the Petri net through tokens and color states. When the token changes, it means that the state has changed and will enter the next state. The process occurring in the middle is represented by transitions. The execution time of the transition activity is set according to the actual situation. The activation and triggering of the transition and the corresponding tag changes are defined by the activation and triggering rules of the actual Petri net respectively. When the required number of tags and color information are present at the input of the transition, the transition will be activated. The activated transition can trigger an event, obtain a specified number of tags from the input, and store a specified number of tags at the output. The subway station fire emergency response process is relatively complex, so a hierarchical subsystem is established, and the main structure of the Petri net is clarified through multiple layers to establish an HTCPN model.

[0061] like Figure 4As shown, in one embodiment of the present invention, a subway fire emergency management system includes at least a control center, which uses a computing system to process and analyze collected data. It also includes sensors for panic detection, distance monitoring, and congestion detection, temperature sensors, and employee information data. These sensors and employee information data are connected to the control center via a communication component, transmitting collected and updated data to the control center. The control center, in turn, uses a computing system to process and analyze the collected data and execute actions via a feedback component. More specifically, the subway fire emergency management system relies on a sensor network to detect and locate hazards, locate personnel (including fire points, trapped personnel, and firefighters), and establish communication between the environment and the control center. In the HTCPN model, repositories represent active states in the system; transitions represent activity processes and can be assigned execution times; and color-coded tags, integrated with the Internet of Things, represent specific types of information. This allows the state of the HTCPN model to be transmitted based on real-time sensor data updates.

[0062] In one embodiment of the present invention, Figure 6 As shown, the parameters input to the HTCPN model include at least: the number of fire points with larger fire intensity identified by the sensor n1, the number of trapped passenger points identified by the sensor n2, the number of fire points with smaller fire intensity identified by the sensor n3, the number of evacuation congestion points identified by the sensor n4, the thermal radiation x of the fire points with larger fire intensity, the number of firefighters assigned to each group y, the distance d between trapped passengers and firefighters, the moving speed s1 of firefighters and trapped passengers, the moving speed s2 of firefighters, the number of firefighting teams G1 of firefighters, the number of firefighting teams G2 of firefighters, the number of firefighting teams of subway employees (staff) G3, the number of rescue teams of subway employees (staff) G4, the number of passengers on the subway q1, the number of passengers on the platform q2, the escalator capacity A1, the staircase capacity A2, the number of escalators B, and the total width N of the stairs. Figure 7 Figure 2 shows the place descriptions and corresponding place transitions in the HTCPN model. Transitions represent activity processes and are typically determined based on expert experience or fitted using statistical data.

[0063] In one embodiment of the present invention, the sensors transmit data to relevant departments, triggering the corresponding emergency response level and issuing an emergency response signal to each department. Upon receiving the signal, the fire brigade will immediately determine the deployment of firefighters based on the real-time data collected by the thermal radiation sensors. Simultaneously, the medical team should dispatch an appropriate number of medical personnel to the scene for rescue operations based on the number of people evacuated and the system's predicted casualties. Upon arrival, the fire brigade will primarily be responsible for extinguishing the fire and rescuing trapped passengers. Upon receiving the emergency response signal, the subway duty station will immediately notify the passenger notification department and guide passengers to evacuate in an orderly manner. Based on the thermal radiation sensor data, some personnel will be deployed to conduct preliminary firefighting to minimize the spread of the fire. Upon receiving the signal, the subway dispatch center will immediately adjust the relevant train schedules, modify the emergency plan based on the situation, and submit the updated passenger evacuation plan to the subway duty station to arrange and carry out passenger evacuation. Upon receiving the emergency response signal, the management team will immediately shut down the power supply to the relevant areas and suspend ticket sales. After each department completes its tasks, it will provide feedback to the Operations Control Center (OCC). Finally, the operation control center will check the situation and, if the fire hazard and evacuation are properly handled, terminate the emergency response system and prepare to resume the operation of the subway station. According to the above procedures, the subway fire emergency response model based on the HTCPN model is as follows Figure 5 As shown, the initial parameter settings are shown in Figure 6 . Figure 7 An explanation of places and transitions is given. Figure 8 The meaning of the colors in the model is given. T2, T6, and T7, namely the firefighting and rescue team, the firefighting team, and the passenger evacuation assistance team, are most affected by management and their behaviors are the most complex. Figure 9 、 Figure 10 as well as Figure 11 The HTCPN model has a three-subsystem model structure. The subsystems of the HTCPN model include a firefighting and rescue subsystem in the rescue section, which connects the location where the rescue team receives a signal and the location where the fire is successfully extinguished and the rescue is completed. The subsystems of the HTCPN model also include a staff firefighting subsystem, which connects the location where the staff assembles and the location where the staff extinguishes the fire. The subsystems of the HTCPN model also include a staff-assisting passenger evacuation subsystem, which connects the locations where the broadcast information is completed, the evacuation plan is generated, and the location where the passenger evacuation is completed.

[0064] More specifically, the emergency response management system's sensor network includes thermal radiation sensors, panic monitoring systems, and distance measurement systems. The thermal radiation sensors detect fires of any cause and monitor changes in thermal radiation in real time. These sensors transmit signals to the controller module, which processes the sensor signals through its input algorithms and outputs them to various departments, providing critical information for firefighters to make timely decisions. The panic monitoring system identifies passengers requiring emergency assistance. The distance measurement system provides the location and distance of passengers awaiting rescue. This information helps firefighters reach specific locations for rescue operations. The passenger monitoring system uses infrared sensors to monitor the number of passengers in the station in real time. The congestion monitoring system identifies crowding during evacuation and transmits this data to subway personnel, facilitating real-time response and preventing stampedes. Figure 4 It is an emergency response management system for subway station fire.

[0065] In one embodiment of the present invention, with the assistance of the Petri net modeling software CPN Tools, a scenario-based simulation of the established HTCPN model was performed to obtain emergency response time and firefighter allocation. This includes: In CPN Tools, the present invention uses tokens in places to represent the resources and information required during the emergency response process, uses transitions to represent the course of action, and adds time functions to transitions to represent the time required for the action. To obtain simulated data on emergency response time and the number of firefighters in each team, the present invention adds count monitors to key places and transitions before running the model.

[0066] More specifically, the HTCPN emergency response model will be simulated under four different scenarios: (1) low thermal radiation and a small number of fire points; (2) high thermal radiation and a small number of fire points; (3) low thermal radiation and a large number of fire points; (4) high thermal radiation and a large number of fire points. Thermal radiation greater than 6kW / m2 is considered high, and less than or equal to 6kW / m2 is considered low; more than 5 fire points are considered high, while less than or equal to 5 are considered low. Five test cases are generated under different scenarios with different thermal radiation intensities and different numbers of fire points, such as Figure 12 In order to fully consider the randomness in simulation optimization, the emergency response HTCPN process plan of each test case was executed 2000 times in CPN Tools.

[0067] In this paper, a process model based on HTCPN was established. Through model simulation, emergency response times for key sub-processes and the overall process were predicted and system performance indicators were analyzed. With the help of a machine learning-based metamodel, SHapley Additive exPlanations (SHAP) model interpretability analysis was performed to identify key factors influencing the performance of the emergency response model. The established HTCPN model was simulated with the Petri net modeling software CPNTools to estimate emergency response times. Monte Carlo (MC) simulation techniques were employed to simulate and mitigate the uncertainty inherent in the characteristics and measurements of input variables. The observed values of the uncertain input variables (including thermal radiation, firefighter walking speed, or the number of passengers) were assumed to follow a normal distribution, with a sampling distribution defined as having an expected value equal to the observed value and a coefficient of variance equal to a range of Bessian levels (5%). The established HTCPN model was executed 1000 times, taking into account different uncertainty levels. The mean and standard deviation of the emergency rescue times for the overall model and its sub-models were monitored and analyzed. Sample emergency response times were obtained, and a histogram of the emergency response time distribution was plotted. Emergency response time follows a log-normal distribution, with an average of 42.306 minutes and a standard deviation of 7.25141 minutes. There is a 74.3% probability that an emergency response can be completed within 46 minutes.

[0068] We further analyze the system's performance indicators and time characteristics to calculate the busyness and efficiency of the entire system or its transitions. We define the possible states of the Petri net, derive its reachable marking set, and establish an isomorphic Markov chain. Based on the stability probability of the Markov chain, we calculate the busyness and transition utilization of the entire system. Specifically, the formula for calculating the steady-state probability is as follows:

[0069]

[0070] in, is the steady-state probability of each tag, Q is the rate transition matrix, and For the diagonal elements in Q If in state and state There is an arc connection between them, and the value of C is at the arc transfer rate λ. Otherwise, the elements on the diagonal of the matrix Q The value is the status The inverse of the sum of the transfer rates on the output arcs of . The busy probability of a place represents the average number of tokens contained in the place, which corresponds to the proportion of the link in the total emergency rescue flow during the handling process.

[0071]

[0072] in, It is a library The stability probability of j tokens in the process is . The transition period utilization rate reflects the busyness of the transition period, which corresponds to the importance of the specific activities in the process. The calculation of the transition period utilization rate is given by the following formula.

[0073]

[0074] Among them, the utilization rate The higher it is, the more important the changes in the entire emergency rescue process will be. ,∀t∈T is equal to the sum of the stable probabilities of all identifiers that can implement t. E is the set of all reachable identifiers that make t implementable.

[0075] Calculations based on the steady-state probability, place busyness, and transition utilization of each marker indicate that P9 is the busiest location during the emergency response process. This indicates that the emergency state is most likely to occur during the response. Transition T2 has the highest utilization rate and takes the longest time, followed by transitions T6 and T7. The duration of firefighting operations is primarily influenced by factors such as the fire severity, accuracy of firefighting operations, firefighting resources, and the need for multi-faceted firefighting. Therefore, it is necessary to strengthen firefighting training, reduce firefighting errors, and conduct multi-faceted firefighting practice to improve firefighting efficiency.

[0076] Based on the content of the above method embodiment, as an optional embodiment, the contribution of different factors to the output of the Petri net model is analyzed according to Shapley's additive interpretation (SHAP). It includes: first, by incorporating the decision tree algorithm - Light Gradient Boosting Machine (LightGBM), a meta-model that can represent the complex relationship between influencing factors and targets is constructed. After constructing the meta-model based on LightGBM, Shapley's additive interpretation (SHAP) is integrated to explain and analyze the importance of influencing factors in the constructed meta-model. Let P be the set containing all influencing factor features, S be the set of all non-zero entries, and the SHAP value can be obtained from the following formula:

[0077]

[0078] where S is a subset of P, is the expected output of the constructed metamodel, whose features are in the set S.

[0079] Specifically, by adjusting the controllable key factors (x, d, s1, s2, q1 and q2) in the constructed meta-model, the model interpretability analysis of the constructed HTCPN system was carried out, and the values of the key factors were allowed to fluctuate by 5% to observe the overall performance of the system. The constructed meta-model was verified and tested, and a relatively high R 2, i.e., a score of 0.8490, and it was found that the constructed meta-model had a well-fitting line between the ground truth and the meta-model prediction results, indicating that the sensitivity analysis of the meta-model should be reliable. When the factors changed by 5%, d, s2, and s1 ranked in the top three, while q1, q2, and x ranked in the bottom three. This suggests that in order to improve the overall performance of the model, determining the location of passengers, especially trapped passengers, and training that may speed up firefighters may have a greater contribution to the efficiency of emergency rescue measures. On the other hand, thermal radiation and the number of people, on the contrary, play a less important role in emergency rescue events.

[0080] In a preferred embodiment of the present invention, the multi-objective optimization model includes:

[0081]

[0082]

[0083] Among them, F1 is the number of firefighters, F2 is the total emergency response time, and are the number of fire brigade and passenger rescue team respectively, is the number of firefighters in the nth fire brigade, is the number of firefighters in the mth passenger rescue team, and the objective function F2 is obtained by simulating the HTCPN model in CPN Tools;

[0084] The constraints are:

[0085]

[0086]

[0087] in, are the maximum and minimum capacities of the fire-fighting teams, respectively. They are the maximum and minimum capacities of the fire rescue team respectively.

[0088] Specifically, for each fire intensity, the number of firefighters per team is evenly distributed based on experience: for large fires, each team consists of eight firefighters; for medium fires, each team consists of six firefighters; and for small fires, each team consists of four firefighters. Upon arrival at the scene, firefighters are assigned to five teams. Three firefighting teams are responsible for extinguishing the fire, while two rescue teams are responsible for rescuing trapped passengers. After completing their current mission, the firefighters' teams remain unchanged and move on to the next one.

[0089] In a preferred embodiment of the present invention, the SkylineOperator algorithm is used to find the Pareto front of a multi-objective optimization problem based on the data set obtained by CPN Tools, including: given a set of n objectives , the Skyline Operator algorithm returns all objects ,make Not affected by any other object Domination, where i, j = 1,…, n.

[0090] Specifically, such as Figure 13 As shown in Figure 1, on the Pareto front, the more firefighters there are, the shorter the emergency response time. Human resource constraints prevent the number of firefighters assigned to a specific area from increasing arbitrarily, while insufficient firefighter allocation can also lead to excessively long emergency response times. Therefore, it is necessary to dispatch as few firefighters as possible within a reasonable timeframe.

[0091] In a preferred embodiment of the present invention, the fuzzy ideal point method is used to obtain the best subway station emergency response plan, specifically including: calculating the membership function The i-th objective function among the k Pareto solutions is expressed as:

[0092]

[0093] in, and are the maximum and minimum values of the i-th target respectively;

[0094] For non-dominated solutions, the normalized membership function μ k The optimization is performed using the following weighted sum model:

[0095]

[0096] Where M is the number of non-dominated solutions; W i is the weight of the i-th objective function; N is the total number of optimization objectives; then, μ k The maximum value max(μ 1 ,…,μ k ) is the optimal solution.

[0097] Specifically, the emergency response time is expected to be controlled within 40 minutes, and the number of firefighters is preferably no more than 30. In scenarios 1, 2, 3, and 4, the optimal Pareto front solution selected based on the fuzzy ideal point method can ensure that the emergency response is completed within 40 minutes. Figure 14 As shown in Figure 2, all 15 situations in Scenario I, Scenario II, and Scenario III can be well optimized, and the emergency response time and the number of firefighters are reduced compared to before optimization. Figure 15As shown, the average improvement levels for Scenario I, Scenario II, and Scenario III are -44.8%, -42.0%, and -41.3%, respectively. The percentage of reduced emergency response time is greater than the optimization of the number of firefighters. In Scenario I and Scenario II, when the number of firefighters is optimized by 10%, the emergency response time is reduced by 34.8 and 32.3%, respectively. In Scenario III, the number of firefighters is reduced by 15.8%, which is more than both Scenario I and Scenario II. However, the number of firefighters is contradictory to the emergency response, so when the number of firefighters is reduced, the emergency response time is only reduced by 25.4%. In the case of a large fire, in order to complete the subway fire emergency response within a reasonable time, the number of firefighters allocated will be sacrificed, thereby avoiding greater economic losses. By Figure 12 It can be seen that in order to ensure that the emergency response can be completed within 40 minutes, the Pareto frontier selects the leftmost point with the shortest time. Figure 14 It can be seen that the emergency response time in scenario 4 increased by 32.9% and the number of firefighters increased by 6%.

[0098] The method of the embodiment of the present invention is implemented by electronic devices, so it is necessary to introduce the relevant electronic devices. Based on this purpose, the embodiment of the present invention provides an electronic device, such as Figure 3 As shown, the electronic device includes: at least one processor, a communications interface, at least one memory, and a communications bus, wherein the at least one processor, the communications interface, and the at least one memory communicate with each other via the communications bus. The at least one processor can call logic instructions in the at least one memory to execute all or part of the steps of the methods provided in the aforementioned method embodiments.

[0099] In addition, the logic instructions in the at least one memory described above can be implemented in the form of a software functional unit and, when sold or used as an independent product, can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, or the portion that contributes to the prior art, or a portion of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions for causing a computer device (which can be a personal computer, server, or network device, etc.) to execute all or part of the steps of the method described in each method embodiment of the present invention. The aforementioned storage medium includes various media that can store program code, such as a USB flash drive, a mobile hard drive, a read-only memory (ROM), a random access memory (RAM), a magnetic disk, or an optical disk.

[0100] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate, and the components shown as units may or may not be physical units, i.e., they may be located in one location or distributed across multiple network units. Some or all of the modules may be selected based on actual needs to achieve the objectives of the present embodiment. Persons of ordinary skill in the art will be able to understand and implement the present invention without inventive effort.

[0101] Through the above description of the embodiments, those skilled in the art will clearly understand that each embodiment can be implemented using software plus a necessary general-purpose hardware platform, or of course, hardware. Based on this understanding, the essence of the above technical solution, or the portion that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, a magnetic disk, or an optical disk, and includes a number of instructions for causing a computer device (such as a personal computer, server, or network device) to execute the methods described in each embodiment or certain portions of the embodiments.

[0102] It will be easily understood by those skilled in the art that the above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. A method for optimizing emergency response plans for subway stations under fire conditions, characterized in that: include: S1 constructs a hierarchical timed colored Petri net model to simulate different scenarios of subway fire emergency response using CPN Tools; S2 establishes a multi-objective optimization model for subway station fire emergency response with the goal of minimizing emergency response time and the number of firefighters; Based on the simulated data set, S3 uses the Skyline Operator algorithm to solve the multi-objective optimization model, obtains the Pareto front, and uses the fuzzy ideal point method to obtain the optimal subway station emergency response plan; In step S1, an HTCPN model and its subsystems are established based on the subway station's preset emergency response process. The information collected by sensors in the subway fire emergency management system is placed in the Petri net via tokens and color states. When a token changes, it means that the state has changed and will enter the next state. The intermediate process is represented by a transition. The execution time of the transition activity is set according to the actual situation. The activation and triggering of the transition and the corresponding tag changes are defined by the activation and triggering rules of the actual Petri net. When the required number of tags and color information are present at the input of the transition, the transition will be activated. The activated transition can trigger an event, obtain a specified number of tags from the input, and store a specified number of tags at the output. The subsystems of the HTCPN model include a fire extinguishing and rescue subsystem in the rescue section, which connects the location where the rescue team receives the signal and the location where the fire is successfully extinguished and the rescue is completed; The subsystem of the HTCPN model also includes a staff fire extinguishing subsystem, which connects the warehouse representing the staff to assemble and the warehouse where the staff extinguish the fire; The subsystems of the HTCPN model also include a staff-assisting passenger evacuation subsystem, which is respectively connected to the library representing the completion of broadcast information, the generation of evacuation plan, and the library representing the completion of passenger evacuation.

2. The method for optimizing emergency response plans for subway stations under fire conditions according to claim 1 is characterized in that: The information detected by the sensors is transmitted to the control center through a heterogeneous network. The control center processes and analyzes the collected data and performs operations through feedback components. Specifically, the fire emergency management system in the subway system relies on sensor networks to detect and locate dangers, locate personnel, including fire points, trapped people, and firefighters, and establish communication between the environment and the control center. In the HTCPN model, the library is used to represent the active state in the fire emergency management system, the transition represents the active process of the state, and the execution time of the transition is assigned based on experience. The colored marks combined with the Internet of Things represent specific types of information. The state of the HTCPN model is transmitted in the form of real-time data updates from sensors.

3. The method for optimizing a subway station emergency response plan under fire conditions according to claim 2 is characterized in that: With the help of the Petri net modeling software CPN Tools, the established HTCPN model was simulated in different scenarios to obtain the emergency response time and the number of firefighters assigned. Specifically, tokens in the library were used to represent the resources and information required in the emergency response process, transitions were used to represent the action process, and time functions were added to the transitions to represent the time required for the action. In order to obtain simulation data on the emergency response time and the number of firefighters in each team, counting monitors were added to the key libraries and transitions before the model was run.

4. The method for optimizing a subway station emergency response plan under fire conditions according to claim 3 is characterized in that: In step S2, the multi-objective optimization model includes: min F2=f(x1,...,x n ,y1,...,y m )n∈{1,2,...,N f },m∈{1,2,...,N r } Among them, F1 is the number of firefighters, F2 is the total emergency response time, N f and N r are the number of fire brigade and passenger rescue team respectively, x n is the number of firefighters in the nth fire brigade, y m is the number of firefighters in the mth passenger rescue team, and the objective function F2 is obtained by simulating the HTCPN model in CPNTools; The constraints are: X min ≤x n ≤X max n∈{1,2,...,N f } AND min ≤y m ≤Y max m∈{1, 2,..., N r } Among them, X max 、X min are the maximum and minimum capacity of the fire fighting team, Y max 、Y min They are the maximum and minimum capacities of the fire rescue team respectively.

5. The method for optimizing a subway station emergency response plan under fire conditions according to any one of claims 1 to 4, characterized in that: In step S3, based on the data set obtained by CPN Tools, the Skyline Operator algorithm is used to find the Pareto front of the multi-objective optimization problem, including: given a set of n objectives {k1, k2, ..., k n }, the Skyline Operator algorithm returns all objects k i , so that k i Not affected by any other object k j Domination, where i, j = 1, ..., n.

6. The method for optimizing a subway station emergency response plan under fire conditions according to claim 5, characterized in that: The fuzzy ideal point method is used to obtain the best subway station emergency response plan, which specifically includes: calculating the membership function The i-th objective function among the k Pareto solutions is expressed as: in, and are the maximum and minimum values of the i-th target respectively; For non-dominated solutions, the normalized membership function μ k The optimization is performed using the following weighted sum model: Where M is the number of non-dominated solutions; W i is the weight of the i-th objective function; N is the total number of optimization objectives; then, μ k The maximum value max(μ 1 ,...,μ k ) is the optimal solution.

7. A subway station emergency response plan optimization system under fire conditions, characterized by: include: The first main control module is used to build a hierarchical timed colored Petri net model to simulate different scenarios of subway fire emergency response using CPN Tools; Based on the subway station's preset emergency response process, an HTCPN model and its subsystems were established. The information collected by sensors in the subway fire emergency management system was placed in a Petri net using tokens and color states. When a token changes, it indicates a state change, transitioning to the next state. The intermediate process is represented by a transition, and the execution time of the transition activity is set based on actual conditions. The activation and triggering of the transition, as well as the corresponding tag changes, are defined by the activation and triggering rules of the actual Petri net. When the required number of tags and color information are present at the transition input, the transition is activated. The activated transition triggers an event, obtaining a specified number of tags from the input and storing a specified number of tags at the output. The subsystems of the HTCPN model include a fire extinguishing and rescue subsystem in the rescue section, which connects the location where the rescue team receives the signal and the location where the fire is successfully extinguished and the rescue is completed; The subsystem of the HTCPN model also includes a staff fire extinguishing subsystem, which connects the warehouse representing the staff to assemble and the warehouse where the staff extinguish the fire; The subsystems of the HTCPN model also include a staff-assisting passenger evacuation subsystem, which is connected to the library representing the completion of broadcast information, the generation of evacuation plan, and the library representing the completion of passenger evacuation; The second main control module is used to establish a multi-objective optimization model for subway station fire emergency response with the goal of minimizing emergency response time and the number of firefighters; The third main control module is used to solve the multi-objective optimization model based on the simulation data set using the Skyline Operator algorithm to obtain the Pareto front, and to obtain the optimal subway station emergency response plan using the fuzzy ideal point method.

8. An electronic device, characterized in that: include: At least one processor, at least one memory and a communication interface; wherein, The processor, memory and communication interface communicate with each other; The memory stores program instructions that can be executed by the processor, and the processor calls the program instructions to execute the method for optimizing the emergency response plan of a subway station under fire conditions according to any one of claims 1 to 6.

9. A non-transitory computer-readable storage medium, characterized in that The non-transitory computer-readable storage medium stores computer instructions, and the computer instructions enable the computer to execute the method for optimizing a subway station emergency response plan under fire conditions according to any one of claims 1 to 6.

Images