A multi-field coupled emergency dynamic simulation evaluation method for fire safety resilience of large and complex public buildings
By employing multi-field coupled simulation and data-driven methods, the accuracy of dynamic fire assessment in large and complex public buildings was addressed, providing timely decision support and improving the building's fire safety management level.
Patent Information
- Application Number
- CN202611500194.0
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2026-09-24
- Publication Date
- 2026-11-17
AI Technical Summary
Existing technologies are insufficient to effectively address the dynamic evolution of fires in large and complex public buildings. The lack of automated interactive coupling and standardized calculation of multi-source data leads to inaccurate fire risk assessment results, making it impossible to guide the upgrading of smoke control and exhaust systems or the dynamic diversion of personnel evacuation in a timely manner. Buildings also struggle to provide early warnings or respond to emergencies.
By constructing a multi-field coupled emergency dynamic simulation evaluation method, multiple sets of FDS input data for different scenarios are generated using data on building floors, fire compartments, and evacuation distances. Combined with Python scripts and large eddy simulation, multi-field coupled simulation of fire evacuation is carried out to obtain simulation data of the entire process, such as fire temperature field, smoke concentration, and visibility distribution. A resilience evaluation model is constructed, and three types of resilience indicators are extracted for pre-disaster, during-disaster, and post-disaster periods. The CRITIC and TOPSIS algorithms are used to sort them and generate the optimal plan.
It enables dynamic quantitative assessment of fire risks in large and complex public buildings, provides timely and accurate decision support, improves fire safety management, and reduces the impact and losses caused by fires.
Smart Images

Figure CN123116633A_ABST
Abstract
Description
Technical Field
[0001] This invention specifically relates to a multi-field coupled emergency dynamic simulation evaluation method for fire safety resilience of large and complex public buildings. Background Technology
[0002] Modern large-scale commercial complexes are important carriers of urban economic development, but due to their large building volume, dense population, and complex spatial structure, they pose significant challenges in controlling fire accidents and causing severe consequences. Current static fire risk assessment methods are insufficient to effectively address the extremely complex dynamic evolution of disasters. While fire dynamics and evacuation simulation software is used for building safety analysis, it still has many shortcomings. These include: current simulations heavily rely on manual modeling and parameter modification, which is time-consuming and prone to errors with unclear causes, making analysis and tracing difficult; fire evolution and evacuation scenarios are often separated into isolated modules, lacking automated interaction and coupling of underlying data; furthermore, not only is the manual extraction of massive amounts of simulation data extremely inefficient and prone to information distortion due to transient high-temperature peaks and instantaneous crowding extremes, but existing methods also lack unified automated and standardized extraction pipelines and calculation methods for multi-dimensional heterogeneous physical and behavioral data such as time, temperature, distance, and number of people, resulting in the inability to directly align and integrate multi-source data. The quality of data input for subsequent evaluation models is difficult to guarantee. From the perspective of safety evaluation indicators and models, most existing research is still at the level of static risk assessment. Research on quantitative decision-making and feedback optimization driven by massive simulation data is still lacking. More research focuses on single-dimensional fire smoke diffusion patterns or simple evacuation path exploration. There is a lack of standardized and dynamic and accurate processing methods for objective comprehensive evaluation after acquiring massive amounts of underlying data, such as introducing multi-criteria objective decision-making algorithms to eliminate information redundancy. The evaluation results cannot be timely, accurately, and intuitively transformed into engineering intervention strategies for upgrading the effectiveness of smoke control systems or dynamically guiding personnel evacuation, resulting in significant delays. Buildings are unable to provide early warnings or make timely decisions to protect themselves in response to emergencies. Summary of the Invention
[0003] To overcome the shortcomings of existing technologies, a multi-field coupled emergency dynamic simulation evaluation method for fire safety resilience of large and complex public buildings is provided to solve the above problems.
[0004] A multi-field coupled emergency dynamic simulation evaluation method for fire safety resilience of large and complex public buildings is proposed. This method involves constructing a building scenario model using acquired data on building floors, fire compartments, and evacuation distances. The building scenario model automatically generates multiple sets of FDS input data for different scenarios using a Python batch script. Building geometry and fire scenario data are extracted from these FDS input data. A fire evacuation multi-field coupled simulation model is then constructed using this data. This model provides full-process simulation data for fire temperature field, smoke concentration, visibility distribution, evacuation time, evacuation routes, and casualties in each scenario. A resilience evaluation model is then constructed using this full-process simulation data. This model extracts three types of resilience indicators: pre-disaster resistance, in-disaster absorption, and post-disaster recovery, and constructs a standardized decision data matrix. This standardized decision data matrix undergoes CRITIC weighting and TOPSIS evaluation and ranking calculations to obtain a resilience ranking result. The optimal resilience plan for the corresponding building layout and fire protection configuration is then determined based on the resilience ranking result.
[0005] As a preferred approach: A building scene model is constructed using the acquired data on building floors, fire compartments, and evacuation distances. The process of automatically generating multiple sets of FDS input data for different scenarios using Python batch scripts involves obtaining basic building data from the large, complex public building under study. This basic data includes building floors, fire compartments, and evacuation distances. Complex spatial geometry, floor distribution, and material properties are extracted using BIM models or IFC format files to establish an initial 3D basic model for the building scene model. An automated script engine developed based on Python is used to read and parse the underlying disaster prevention simulation files. Python code scripts are used to read, identify, and modify the underlying information of fire source coordinates and environmental parameters. Regular expressions are used to automatically align multiple spatiotemporal data, completing the batch overwriting of underlying parameters and the automated generation of multiple fire location scenarios. For different fire occurrence characteristics, regular expressions and other methods are used to automatically identify and modify the initial environmental parameters of the fire source, thereby obtaining multiple sets of FDS input data for different scenarios.
[0006] As a preferred approach: Building geometry and fire scenario data are extracted from multiple sets of FDS input data for different scenarios. A multi-field coupled simulation model for fire evacuation is then constructed using this data. The process of obtaining full-process simulation data on fire temperature field, smoke concentration, visibility distribution, evacuation time, evacuation routes, and casualties under each scenario through this model involves parametric modeling and batch scripting using Python. This process iterates through variable parameters such as building floors, fire compartments, evacuation distance, fire source location, and fire source power to generate multiple sets of FDS input data for different scenarios. These FDS input data are presented as multiple differentiated FDS input text files, each corresponding to an independent building fire simulation scenario. The parsing program is then started, and each FDS input file is read in a loop. After reading the text data line by line from the input file, the corresponding valid parameters are retained. The key FDS data is identified from the valid parameters. The identified key FDS data is organized into a building geometry structured dataset. After matching and parsing the building geometry structured dataset, fire scene parameters are extracted. After parameter verification and unit normalization of the fire scene parameters, building geometry and fire scene data are formed.
[0007] After obtaining the building geometry and fire scene data, the building geometry and fire scene data are combined with the Large Eddy Simulation (LES) method to complete the high-fidelity calculation of the fire dynamics inside the building. That is, the simulation mesh resolution and physical boundary conditions are set, and based on the mass conservation equation, momentum conservation equation, energy conservation equation and state equation, transient calculations are performed on the flow, diffusion rate and toxic gas concentration distribution of high-temperature smoke generated by the fire to construct a fire evacuation multi-field coupled simulation model. In the fire evacuation multi-field coupled simulation model, the eddy current effect of the atrium structure, the smoke stratification phenomenon in the corridor area and the local backflow characteristics under complex partitions are accurately captured. The calculation is then completed to generate three-dimensional slice and mesh feature data including temperature, visibility and carbon monoxide concentration.
[0008] After acquiring three-dimensional slice and mesh feature data including temperature, visibility and carbon monoxide concentration, the Steering mode is used as the control algorithm for personnel evacuation path planning in the fire evacuation multi-field coupled simulation model. The personnel density and personnel behavior parameters of each area are set. The Steering mode is used to accurately calculate the micro-trajectory of personnel when avoiding obstacles and other escapees, and accurately identify the dynamic congestion bottleneck, peak number of people queuing and staying in the corresponding stairwell or other relevant information.
[0009] Using a code interface, the time-dependent spatial smoke data generated by the fire simulation is synchronously mapped into the fire evacuation multi-field coupled simulation model. This ensures that virtual evacuees exhibit velocity attenuation or path replanning behavior when encountering visibility below the safety threshold or high-temperature hazard areas. After completing the spatiotemporal deep interactive coupling calculation of the physical field and the microscopic behavioral field, the required safe evacuation time RSET comprehensive evolution data is output. The required safe evacuation time RSET comprehensive evolution data is the full-process simulation data of fire temperature field, smoke concentration, visibility distribution, personnel evacuation time, personnel evacuation path, and personnel casualties in each scenario.
[0010] As a preferred approach: A resilience evaluation model is constructed using full-process simulation data. This model then extracts three types of resilience indicators—pre-disaster resistance, in-disaster absorption, and post-disaster recovery—and builds a standardized decision data matrix. The process involves first executing a full-process multi-field coupled simulation, which imports the building model and generates different fire scenarios in batches. The LES algorithm solves the smoke dynamics, and the Steering algorithm simulates evacuation, completing the spatiotemporal interaction between the smoke hazard field and evacuation behavior. Low-level heterogeneous simulation data, such as temperature, visibility, evacuation congestion, and evacuation time margin, are collected. This massive amount of simulation output data serves as the model input foundation, building the CRITIC system. -TOPSIS Resilience Assessment Model: First, the original simulation data is cleaned and normalized; the CRITIC algorithm calculates the standard deviation and Pearson correlation coefficient of the indicators, calculates the information content of the indicators to complete objective weighting, and eliminates subjective scoring and information redundancy; then, a weighted decision matrix is constructed, positive and negative ideal solutions are solved, and the Euclidean distance and relative closeness of each scenario are calculated to obtain a 0-1 comprehensive resilience evaluation value, thus completing the resilience assessment model construction process. After the resilience assessment model is constructed, three types of resilience indicators are extracted from the resilience assessment model: pre-disaster resistance, disaster absorption, and post-disaster recovery. A standardized decision data matrix is constructed through these three resilience indicators.
[0011] The standardized decision data matrix is a data matrix consisting of twenty-seven indicators;
[0012] The pre-disaster resilience indicators in the standardized decision-making data matrix include the time it takes for smoke to reach the stairwell, the time it takes for the critical temperature to reach the threshold, the evacuation start time, and the distribution of evacuation distances for all personnel.
[0013] The disaster absorption indicators in the standardized decision data matrix include the proportion of smoke diffusion area, the efficiency index of personnel movement, the evacuation time margin, the maximum heat release rate, the congestion of evacuation routes, and the efficiency of personnel evacuation actions.
[0014] The post-disaster recovery indicators in the standardized decision-making data matrix include the proportion of structural high-temperature damage areas, the recoverability of evacuation routes, the balance of evacuation route utilization, and the smoke removal time.
[0015] As a preferred option: the calculation process for the smoke arrival time index in the stairwell is as follows: the stairwell is a vertical evacuation shaft in the building, and the timing of smoke intrusion directly determines the survival period of the safety defense line. The visibility function of the stairwell node is V(t), and the critical time t when it first drops to the safety threshold is tracked. stair Combined with the baseline safety time limit t required by the specifications baseline The relative abundance formula is as follows: The total number of stairwells N will be extracted. total and visibility N fail_vis Substitute it into the global score formula as a variable The calculation is performed to obtain the time index for smoke to reach the stairwell;
[0016] The calculation process for the critical temperature arrival time index is as follows: when the ambient temperature of the evacuation route exceeds the human body's tolerance limit, people will lose their ability to move. The critical temperature arrival time index characterizes the building's time buffering capacity in high-temperature environments. The three-dimensional temperature field data T(x,y,t) at a height of Z=2.0m in the evacuation network is extracted, and the time t for the first breach of the critical temperature is calculated. temp Combined with the baseline safety time limit t required by the specifications baseline The relative sufficiency formula is used to measure the status of a single line of defense: Extract the variable, specifically the total number of stairwells N. total and limiting temperature N fail_temp Substituting variables into the formula The calculation is performed to obtain the critical temperature arrival time index;
[0017] The evacuation start time index is calculated by extracting the average time t during the agent's pre-action phase. pre Total required safe evacuation time t RSET Then, the average time t during the agent's pre-action phase is... pre Total required safe evacuation time t RSET Construct the basic determination formula: ; Extracting the number of floors and areas that successfully triggered the visibility alarm and performing Z-division. act and the total number Z of all monitored areas total Substituting variables into the formula The calculations were performed to obtain the evacuation start time index;
[0018] The calculation process for the overall evacuation distance distribution index involves assessing the rationality of the building's internal spatial layout and evacuation network topology. The closer the actual evacuation distance is to the standard limit, the stronger the redundancy resistance of the spatial layout. The total actual evacuation paths ΣD of all agents are calculated. actual The sum of the ideal shortest paths ΣDideal Based on the sum of the actual evacuation paths of all agents ΣD actual The sum of the ideal shortest paths ΣD ideal Construct the basic determination formula: Extract the standard distance limit D limit And the average evacuation distance D for all personnel. mean Substituting variables into the formula The calculations were performed to obtain the distribution index of evacuation distances for all personnel.
[0019] As a preferred option: the calculation process for the flue gas diffusion area ratio index involves integrating the slice to obtain the instantaneous hazardous area and locking in the extreme value A. max_smoke With total area A total Construct the basic determination formula: Extract the peak proportion R of visibility dropping below the 10m threshold in the built area. max Substituting variables into the formula Calculations were performed to obtain the percentage of flue gas diffusion area.
[0020] The calculation process of the personnel mobility efficiency index is a process of assessing the constraints and degree of soft disruption on the behavioral dynamics of a crowd in a harsh fire environment, and statistically analyzing the free movement speed of agents in a fire-free state. Actual deceleration in dangerous areas Construct the basic determination formula: Extract the lowest recorded individual effective speed V min and the critical failure threshold V for severe congestion limit Substitute it into the formula as a variable The personnel mobility efficiency index was obtained through calculation.
[0021] The evacuation time margin index is calculated by extracting the available safe evacuation time t based on the ASET-RSET theory. ASET With the necessary safe evacuation time t RSET As a variable, the basic determination formula is constructed: ; then extract the available safe evacuation time t ASET and the required safe evacuation time t REST Substituting variables into the formula The evacuation time margin index is calculated.
[0022] The calculation process for the maximum heat release rate index involves determining the ultimate thermal shock pressure that the entire smoke control system needs to absorb based on the fire source evolution intensity correspondence method, and then extracting the peak combustion power HRR. max With design limits HRR limit Construct the basic determination formula: Extract the peak HRR value Q after stabilization. max_smoothand the maximum heat capacity limit Q specified in the structural smoke exhaust design. limit Substituting variables into the formula The maximum heat release rate index is obtained through calculation;
[0023] The calculation process for the congestion index of evacuation routes involves determining the route's capacity to handle sudden large crowds by quantifying the physical stagnation and stampede risk caused by excessively high local crowd density, and then integrating the total congestion time Σt. jam With total time t evac_total Construct the basic determination formula: Extract the maximum value of the total number of people (Occ) in the most crowded stairwell. max And the theoretical maximum load capacity C calculated based on the staircase area and extreme congestion density. max Substituting variables into the formula The congestion index of evacuation routes is calculated in the process.
[0024] The calculation process for the personnel evacuation efficiency index is based on the overall evacuation network's macroscopic population flow efficiency in the face of a fire. This involves calculating the minimum evacuation time t under conditions of no fire obstruction. RSET_ideal Compared with the actual time t RSET_actual Construct the basic determination formula: Extract the average efficiency of all personnel. and the predefined free-flow efficiency threshold E free_flow Substituting variables into the formula The efficiency index of personnel evacuation operations was obtained through calculation.
[0025] As a preferred option: the calculation process for the structural high-temperature damage area ratio index is to assess the damage range of the building entity under high temperature exposure, determine the difficulty of restoring the building's functionality after a disaster, and extract the surface area A whose surface temperature exceeds a threshold due to fire. damage With total area A total Construct the basic determination formula: ; Extract the peak value ratio R of concrete damage threshold at temperatures exceeding 300°C from the surface area of the structure. dmg_max Substituting variables into the formula The proportion of high-temperature damage area in the structure is obtained through calculation.
[0026] The calculation process for the recoverability index of evacuation routes is based on assessing the flow recovery and carrying capacity of physical routes under the impact of pedestrian traffic. The recoverability index is determined by using the extreme value of the recovery time t. max t min and channel recovery time t recover_channel Construct the basic determination formula: Extract the calculated browsing limit F based on the step width. theoretical and the peak flow F dynamically observed on all bottleneck sections actual_peak Substituting variables into the formula The recoverability index of evacuation routes is obtained through calculation.
[0027] The calculation process for the evacuation route utilization balance index is based on the Gini coefficient theory to assess whether the population can adaptively and evenly utilize all the route networks to determine the evacuation route utilization balance index, with x representing the number of people passing through n exits. i The mean is Construct the basic determination formula: Extract the Gini coefficient G, calculated based on the cumulative distribution of evacuees along all available evacuation routes, and substitute it into the formula. The utilization balance index of evacuation routes is obtained through calculation.
[0028] The calculation process for the smoke removal time index is based on quantifying the smoke control and exhaust system's exhaust efficiency in the later stages of the system to determine the system's agility in restoring the spatial environment to a tolerable state, and tracking the time t required for visibility restoration. clear and extreme values t max t min Construct the basic determination formula: Extract the precise timestamp t when the overall spatial tolerance drops below the 95% tolerance threshold. breach and total simulation duration t total_sim Substituting variables into the formula The flue gas removal time index is calculated in the process.
[0029] As a preferred option, pre-disaster resilience indicators also include building fire resistance rating, average fire resistance limit of structural components, compliance rate of smoke control and exhaust system configuration, compliance rate of safety exit spacing, and reliability of fire-resistant doors and windows.
[0030] Post-disaster recovery indicators include structural repairability index, fire protection facility availability rate index, building functional zoning replaceability index, and structural redundancy index.
[0031] As a preferred option: the standardized decision data matrix is sequentially weighted by CRITIC and ranked by TOPSIS to obtain the resilience ranking results. The process of determining the optimal resilience plan for the corresponding building layout and fire protection configuration based on the resilience ranking results is as follows:
[0032] In the resilience evaluation model, the standardized decision data matrix is sequentially processed through CRITIC weighting and TOPSIS evaluation sorting to obtain the CRITIC-TOPSIS quantitative evaluation model. The calculation process of the CRITIC-TOPSIS quantitative evaluation model is as follows: the standardized index data is imported into the evaluation algorithm kernel. First, CRITIC objective weighting is performed. By calculating the standard deviation of each index under different scenarios, the contrast strength index is obtained. The Pearson correlation coefficient between the indices is calculated to obtain the quantitative conflict index. The contrast strength index and the quantitative conflict index are combined to calculate the comprehensive information content. Then, the weight distribution data of each index after objectively removing redundant information is obtained. Then, the system resilience score is calculated. The corresponding calculation process is as follows: based on the obtained objective weights, a weighted standardized decision matrix is constructed; the optimal and worst values of each index are extracted to generate positive ideal solutions and negative ideal solutions; then, the Euclidean geometric distance between each simulation scenario and the positive and negative ideal solutions is calculated; finally, through the relative closeness formula, the centroid method is used to determine a definite precise value for the output quantity, and the comprehensive system resilience score between 0 and 1 is calculated.
[0033] The system receives the overall resilience score output by the evaluation module and draws corresponding conclusions based on the weighting of each indicator. When the system identifies that the overall available safe evacuation time (ASET) is approaching its limit or the total resilience score is below the safety threshold, it automatically triggers a high-risk warning.
[0034] When the cumulative weight of macro-indicators such as smoke exhaust time or smoke diffusion ratio exceeds 68%, the system generates optimization suggestions for upgrading the smoke control system efficiency, guiding the re-deployment and optimization of passive smoke exhaust and mechanical active smoke exhaust in the building space.
[0035] When the arrival time of smoke in the stairwell or the congestion index of the evacuation route shows high sensitivity, a corresponding treatment plan is generated to improve the early multi-mode sensor warning and the configuration of smoke control pressure differential in the core evacuation shaft.
[0036] When the evacuation congestion is highly sensitive, a dynamic flow management solution based on digital twin space is output;
[0037] When the system calculates and finds that the overall available safe evacuation time is approaching the limit or the resilience score is extremely low, it sends an extremely high risk warning and automatically outputs the corresponding optimal intervention plan.
[0038] The beneficial effects of this invention are as follows:
[0039] This invention is essentially a building fire safety resilience evaluation method based on multi-field coupled simulation and data-driven approach. It aims to standardize and accurately improve the fire safety management process for complex buildings such as large commercial complexes, eliminating the subjectivity and lag of static assessments and achieving dynamic quantification of fire risk throughout its entire lifecycle, thus enabling a standardized and timely decision-making process. This invention constructs a building scenario model using acquired data on building floors, fire compartments, and evacuation distances. This model automatically generates multiple sets of FDS input data for different scenarios using Python batch scripts. Building geometry and fire scenario data are extracted from these FDS input data. A multi-field coupled simulation model for fire evacuation is then constructed using this data. This model provides full-process simulation data for fire temperature field, smoke concentration, visibility distribution, evacuation time, evacuation routes, and casualties in each scenario. This full-process simulation data is then used to construct… This invention establishes a resilience evaluation model, extracting three types of resilience indicators—pre-disaster resistance, in-disaster absorption, and post-disaster recovery—and constructing a standardized decision data matrix. This matrix is then subjected to CRITIC weighting and TOPSIS evaluation and ranking calculations to obtain resilience ranking results. These ranking results are used to determine the optimal resilience plan for the corresponding building layout and fire protection configuration. This invention constructs an evaluation system from a multi-step dynamic resilience perspective, encompassing resistance, absorption, and recovery. It incorporates smoke clearance time, stairwell smoke arrival time, and other relevant dynamic physical indicators into a comprehensive evaluation process. The evaluation results are accurate and timely, guiding the safety protection of actual buildings and improving their proactive safety protection performance. This invention is particularly suitable for early warning assessment of large and complex public buildings, improving the overall fire safety management level of these buildings throughout their entire lifecycle and minimizing the impact and losses caused by fire evolution. Attached Figure Description
[0040] Figure 1 This is a schematic diagram of the overall process of the present invention;
[0041] Figure 2 This is a schematic diagram of the process logic of the present invention;
[0042] Figure 3 This is a schematic diagram of the fire safety resilience evaluation index system.
[0043] Figure 4 This is a schematic diagram of the simulation and underlying data extraction framework in this invention.
[0044] Figure 5 This is a schematic diagram of the toughness evaluation system and calculation process in this invention.
[0045] Figure 6 This is a case application model diagram of the building fire safety resilience evaluation method in this invention. Detailed Implementation
[0046] The following specific examples illustrate the implementation of the present invention. Those skilled in the art can easily understand other advantages and effects of the present invention from the content disclosed in this specification. The present invention can also be implemented or applied through other different specific embodiments, and various details in this specification can also be modified or changed based on different viewpoints and applications without departing from the spirit of the present invention.
[0047] Specific implementation method one: Combining Figures 1 to 6 This implementation method describes a multi-field coupled simulation evaluation method that utilizes acquired data on building floors, fire compartments, and evacuation distances to construct a building scenario model. This model automatically generates multiple sets of FDS input data for different scenarios using a Python batch script. Building geometry and fire scenario data are extracted from these FDS input data. A fire evacuation multi-field coupled simulation model is then constructed using this data. This model provides full-process simulation data for fire temperature field, smoke concentration, visibility distribution, evacuation time, evacuation routes, and casualties in each scenario. A resilience evaluation model is then constructed using this full-process simulation data. This model extracts three types of resilience indicators: pre-disaster resistance, in-disaster absorption, and post-disaster recovery, and constructs a standardized decision data matrix. This standardized decision data matrix undergoes CRITIC weighting and TOPSIS evaluation and ranking calculations to obtain resilience ranking results. The process of determining the optimal resilience plan for the corresponding building layout and fire protection configuration based on the resilience ranking results is derived through an auxiliary decision-making and scheme optimization model.
[0048] The implementation of this invention involves four models: a building scene model, a fire evacuation multi-field coupled simulation model, a resilience evaluation model, and a decision support and scheme optimization model.
[0049] Combination Figure 1As shown, this invention involves several main models, namely, a building scene model, a fire evacuation multi-field coupled simulation model, and a resilience evaluation model. The building scene model takes parameters such as building floors, fire compartments, and evacuation distances as input. Through parametric modeling and Python batch scripts, it automatically generates multiple sets of FDS input data for different scenarios, transmitting the building geometry and fire scene data to the fire evacuation multi-field coupled simulation model. The fire evacuation multi-field coupled simulation model takes these FDS files as input and runs fire dynamics large eddy simulation and personnel evacuation group behavior models, respectively. Through Python, it automatically drives the simulation process, achieving bidirectional dynamic data flow coupling between fire smoke data and personnel visibility and pathfinding behavior. It outputs a full-process simulation of fire temperature field, smoke concentration, visibility distribution, and personnel evacuation time, path, and casualties under each scenario. The data and resilience evaluation model takes these simulation result files as input, automatically extracts three types of resilience indicators: pre-disaster resistance, disaster absorption, and post-disaster recovery, constructs a standardized data matrix, calculates the objective weights of each indicator using the CRITIC method, and then uses the TOPSIS method to comprehensively score and rank the fire resilience of commercial complexes in multiple scenarios. The decision support and scheme optimization model takes the resilience ranking results as input, compares the score differences of different scenarios, identifies specific weak links in low-scoring scenarios such as excessively long evacuation time and excessively rapid decline in visibility, and finally determines which building layout and fire protection configuration has the best resilience, and proposes targeted optimization suggestions, which are fed back to the building scenario model to adjust building parameters and re-simulate, forming an iterative closed loop.
[0050] Specific Implementation Method Two: This implementation method further defines Specific Implementation Method One. In this method, a building scene model is constructed using the acquired building floor, fire compartment, and evacuation distance data. The process of automatically generating multiple sets of FDS input data for different scenarios using Python batch scripts involves obtaining basic building data from the building of the research object. The basic building data includes building floor, fire compartment, and evacuation distance data. Complex spatial geometry, floor distribution, and material properties are extracted using BIM models or IFC format files to establish an initial three-dimensional basic model of the building scene model. An automated script engine developed based on Python is used to read and parse the underlying disaster prevention simulation file. The underlying information of fire source coordinates and environmental parameters is read, identified, and modified using Python code scripts. Regular expressions are used to automatically align multiple spatiotemporal data to complete the batch overwriting of underlying parameters and the automated generation of multiple fire occurrence location scenarios. For different fire occurrence characteristics, the fire source coordinates and initial environmental parameters are automatically identified and modified using regular expressions and other means to obtain multiple sets of FDS input data for different scenarios.
[0051] Specific Implementation Method Three: This implementation method is a further limitation of Specific Implementation Method One or Two. In this implementation method, building geometry and fire scene data are extracted from multiple sets of FDS input data of different scenarios. A multi-field coupled simulation model for fire evacuation is constructed using this data. The process of obtaining full-process simulation data of fire temperature field, smoke concentration, visibility distribution, personnel evacuation time, personnel evacuation path, and personnel casualties under each scenario through the multi-field coupled simulation model is as follows: Through parametric modeling and Python batch scripting, variable parameters such as building floors, fire compartments, evacuation distance, fire source location, and fire source power are traversed to generate multiple sets of FDS input data for different scenarios. These multiple sets of FDS input data are multiple differentiated scenario FDS input text files. Each FDS input file corresponds to an independent building fire simulation scenario. The parsing program is started, and each FDS input file is read in a loop. After reading the text data line by line from the input file, the corresponding valid parameters are retained. The key FDS data is identified from the valid parameters. The identified key FDS data is organized into a building geometry structured dataset. After matching and parsing the building geometry structured dataset, fire scene parameters are extracted. After parameter verification and unit normalization of the fire scene parameters, building geometry and fire scene data are formed.
[0052] After obtaining the building geometry and fire scene data, the building geometry and fire scene data are combined with the Large Eddy Simulation (LES) method to complete the high-fidelity calculation of the fire dynamics inside the building. That is, the simulation mesh resolution and physical boundary conditions are set, and based on the mass conservation equation, momentum conservation equation, energy conservation equation and state equation, transient calculations are performed on the flow, diffusion rate and toxic gas concentration distribution of high-temperature smoke generated by the fire to construct a fire evacuation multi-field coupled simulation model. In the fire evacuation multi-field coupled simulation model, the eddy current effect of the atrium structure, the smoke stratification phenomenon in the corridor area and the local backflow characteristics under complex partitions are accurately captured. The calculation is then completed to generate three-dimensional slice and mesh feature data including temperature, visibility and carbon monoxide concentration.
[0053] After acquiring three-dimensional slice and mesh feature data including temperature, visibility and carbon monoxide concentration, the Steering mode is used as the control algorithm for personnel evacuation path planning in the fire evacuation multi-field coupled simulation model. The personnel density and personnel behavior parameters of each area are set. The Steering mode is used to accurately calculate the micro-trajectory of personnel when avoiding obstacles and other escapees, and accurately identify the dynamic congestion bottleneck, peak number of people queuing and staying in the corresponding stairwell or other relevant information.
[0054] Using a code interface, the time-dependent spatial smoke data generated by the fire simulation is synchronously mapped into the fire evacuation multi-field coupled simulation model. This ensures that virtual evacuees exhibit velocity attenuation or path replanning behavior when encountering visibility below the safety threshold or high-temperature hazard areas. After completing the spatiotemporal deep interactive coupling calculation of the physical field and the microscopic behavioral field, the required safe evacuation time RSET comprehensive evolution data is output. The required safe evacuation time RSET comprehensive evolution data is the full-process simulation data of fire temperature field, smoke concentration, visibility distribution, personnel evacuation time, personnel evacuation path, and personnel casualties in each scenario.
[0055] Specific Implementation Method Four: This implementation method is a further limitation of Specific Implementation Methods One, Two, or Three. In this implementation method, a resilience evaluation model is constructed using full-process simulation data. The process of extracting three types of resilience indicators—pre-disaster resistance, disaster absorption, and post-disaster recovery—from the resilience evaluation model and then constructing a standardized decision data matrix is as follows:
[0056] In this implementation, a full-process multi-field coupled simulation process is first executed, i.e., the building model is imported and different fire scenarios are generated in batches; LES solves the smoke dynamics, and the Steering algorithm simulates personnel evacuation, completing the spatiotemporal interaction between the smoke hazard field and personnel evacuation behavior. The underlying simulation heterogeneous raw data of temperature, visibility, evacuation congestion, and evacuation time margin are collected. The massive simulation output data is used as the model input base to build the CRITIC-TOPSIS resilience evaluation model: First, the raw simulation data is cleaned and normalized; the CRITIC algorithm calculates the standard deviation of the indicators and the Pearson correlation coefficient, calculates the information content of the indicators to complete the objective weighting, and eliminates subjective scoring and information redundancy; then, a weighted decision matrix is constructed, the positive and negative ideal solutions are solved, the Euclidean distance and relative proximity of each scenario are calculated, and a 0-1 comprehensive resilience evaluation value is obtained, thus completing the resilience evaluation model construction process. After the resilience evaluation model is built, three types of resilience indicators are extracted from the resilience evaluation model: pre-disaster resistance, in-disaster absorption, and post-disaster recovery. A standardized decision data matrix is constructed through the three resilience indicators.
[0057] The standardized decision data matrix is a data matrix consisting of twenty-seven indicators;
[0058] The pre-disaster resilience indicators in the standardized decision-making data matrix include the time it takes for smoke to reach the stairwell, the time it takes for the critical temperature to reach the threshold, the evacuation start time, and the distribution of evacuation distances for all personnel.
[0059] The disaster absorption indicators in the standardized decision data matrix include the proportion of smoke diffusion area, the efficiency index of personnel movement, the evacuation time margin, the maximum heat release rate, the congestion of evacuation routes, and the efficiency of personnel evacuation actions.
[0060] The disaster absorption indicators in the standardized decision data matrix include the proportion of structural high-temperature damage areas, the recoverability of evacuation routes, the balance of evacuation route utilization, and the flue gas removal time.
[0061] Specific Implementation Method Five: This implementation method is a further limitation of Specific Implementation Method One. In this implementation method, the calculation process for the smoke arrival time index in the stairwell is as follows: the stairwell is a vertical evacuation shaft of the building, and the timing of smoke intrusion directly determines the survival period of the safety defense line. The visibility function of the stairwell node is V(t), and the critical time t when it first drops to the safety threshold is tracked. stair Based on the baseline safety time limit t required by the specifications. baseline The relative abundance formula is as follows: The total number of stairwells N will be extracted. total and visibility N fail_vis Substitute it into the global score formula as a variable The calculation is performed to obtain the time index for smoke to reach the stairwell;
[0062] The calculation process for the critical temperature arrival time index is as follows: when the ambient temperature of the evacuation route exceeds the human body's tolerance limit, people will lose their ability to move. The critical temperature arrival time index characterizes the building's time buffering capacity in high-temperature environments. The three-dimensional temperature field data T(x,y,t) at a height of Z=2.0m in the evacuation network is extracted, and the time t for the first breach of the critical temperature is calculated. temp Combined with the baseline safety time limit t required by the specifications baseline The relative sufficiency formula is used to measure the status of a single line of defense: Extract the variable, specifically the total number of stairwells N. total and limiting temperature N fail_temp Substituting variables into the formula The calculation is performed to obtain the critical temperature arrival time index;
[0063] The evacuation start time index is calculated by extracting the average time t during the agent's pre-action phase. pre Total required safe evacuation time t RSET Then, the average time t during the agent's pre-action phase is... pre Total required safe evacuation time t RSET Construct the basic determination formula: ; Extracting the number of floors and areas that successfully triggered the visibility alarm and performing Z-division. act and the total number Z of all monitored areas total Substituting variables into the formula The calculations were performed to obtain the evacuation start time index;
[0064] The calculation process for the overall evacuation distance distribution index involves assessing the rationality of the building's internal spatial layout and evacuation network topology. The closer the actual evacuation distance is to the standard limit, the stronger the redundancy resistance of the spatial layout. The total actual evacuation paths ΣD of all agents are calculated. actual The sum of the ideal shortest paths ΣD ideal Based on the sum of the actual evacuation paths of all agents ΣD actual The sum of the ideal shortest paths ΣD ideal Construct the basic determination formula: Extract the standard distance limit D limit And the average evacuation distance D for all personnel. mean Substituting variables into the formula The calculations were performed to obtain the distribution index of evacuation distances for all personnel.
[0065] Specific Implementation Method Six: This implementation method is a further limitation of Specific Implementation Method One. In this implementation method, the calculation process of the flue gas diffusion area ratio index is to integrate the slice to obtain the instantaneous dangerous area and lock the extreme value A. max_smoke With total area A total Construct the basic determination formula: Extract the peak proportion R of visibility dropping below the 10m threshold in the built area. max Substituting variables into the formula Calculations were performed to obtain the percentage of flue gas diffusion area.
[0066] The calculation process of the personnel mobility efficiency index is a process of assessing the constraints and degree of soft disruption on the behavioral dynamics of a crowd in a harsh fire environment, and statistically analyzing the free movement speed of agents in a fire-free state. Actual deceleration in dangerous areas Construct the basic determination formula: Extract the lowest recorded individual effective speed V min and the critical failure threshold V for severe congestion limit Substitute it into the formula as a variable The personnel mobility efficiency index was obtained through calculation.
[0067] The evacuation time margin index is calculated by extracting the available safe evacuation time t based on the ASET-RSET theory. ASET With the necessary safe evacuation time t RSET As a variable, the basic determination formula is constructed: ; then extract the available safe evacuation time t ASET and the required safe evacuation time t REST Substituting variables into the formula The evacuation time margin index is calculated.
[0068] The calculation process for the maximum heat release rate index involves determining the ultimate thermal shock pressure that the entire smoke control system needs to absorb based on the fire source evolution intensity correspondence method, and then extracting the peak combustion power HRR. max With design limits HRR limit Construct the basic determination formula: Extract the peak HRR value Q after stabilization. max_smooth and the maximum heat capacity limit Q specified in the structural smoke exhaust design. limit Substituting variables into the formula The maximum heat release rate index is obtained through calculation;
[0069] The calculation process for the congestion index of evacuation routes involves determining the route's capacity to handle sudden large crowds by quantifying the physical stagnation and stampede risk caused by excessively high local crowd density, and then integrating the total congestion time Σt. jam With total time t evac_total Construct the basic determination formula: Extract the maximum value of the total number of people (Occ) in the most crowded stairwell. max And the theoretical maximum load capacity C calculated based on the staircase area and extreme congestion density. max Substituting variables into the formula The congestion index of evacuation routes is calculated in the process.
[0070] The calculation process for the personnel evacuation efficiency index is based on the overall evacuation network's macroscopic population flow efficiency in the face of a fire. This involves calculating the minimum evacuation time t under conditions of no fire obstruction. RSET_ideal Compared with the actual time t RSET_actual Construct the basic determination formula: Extract the average efficiency of all personnel. and the predefined free-flow efficiency threshold E free_flow Substituting variables into the formula The efficiency index of personnel evacuation operations was obtained through calculation.
[0071] Specific Implementation Method Seven: This implementation method is a further limitation of Specific Implementation Method One. In this implementation method, the calculation process for the structural high-temperature damage area ratio index is to assess the damage range of the building entity under high-temperature baking, determine the difficulty of restoring the building's functional structure after the disaster, and extract the surface area A whose surface temperature exceeds the threshold due to fire. damage With total area A total Construct the basic determination formula: ; Extract the peak value ratio R of concrete damage threshold at temperatures exceeding 300°C from the surface area of the structure. dmg_max Substituting variables into the formula The proportion of high-temperature damage area in the structure is obtained through calculation.
[0072] The calculation process for the recoverability index of evacuation routes is based on assessing the flow recovery and carrying capacity of physical routes under the impact of pedestrian traffic. The recoverability index is determined by using the extreme value of the recovery time t. max t min and channel recovery time t recover_channel Construct the basic determination formula: Extract the calculated browsing limit F based on the step width. theoretical and the peak flow F dynamically observed on all bottleneck sections actual_peak Substituting variables into the formula The recoverability index of evacuation routes is obtained through calculation.
[0073] The calculation process for the evacuation route utilization balance index is based on the Gini coefficient theory to assess whether the population can adaptively and evenly utilize all the route networks to determine the evacuation route utilization balance index, with x representing the number of people passing through n exits. i With a mean of µ, construct the basic determination formula: Extract the Gini coefficient G, calculated based on the cumulative distribution of evacuees along all available evacuation routes, and substitute it into the formula. The utilization balance index of evacuation routes is obtained through calculation.
[0074] The calculation process for the smoke removal time index is based on quantifying the smoke control and exhaust system's exhaust efficiency in the later stages of the system to determine the system's agility in restoring the spatial environment to a tolerable state, and tracking the time t required for visibility restoration. clear and extreme values t max t min Construct the basic determination formula: Extract the precise timestamp t when the overall spatial tolerance drops below the 95% tolerance threshold. breach and total simulation duration t total_sim Substituting variables into the formula The flue gas removal time index is calculated in the process.
[0075] Specific Implementation Method Eight: This implementation method is a further limitation of Specific Implementation Method One. In this implementation method, the process of the automatic extraction and evaluation module of resilience index is based on the underlying data output by multi-field coupling simulation. Fourteen quantitative indicators characterizing the resilience of building fire safety are constructed, and customized calculation and scoring formulas are configured for each of the fourteen indicators for automatic conversion; the specific process is as follows:
[0076] Step A: This step involves cleaning multi-source heterogeneous data. The Python code engine is used to automatically parse the fire evolution grid file and personnel evacuation trajectory text, and extract the transient time variable t, spatial environmental variables, and personnel micro-behavioral variables of each monitoring point.
[0077] Step B involves calculating customized indicator formulas. Using fourteen built-in customized calculation formulas, the extracted underlying variables are transformed into resilience quantitative indicators. These fourteen quantitative indicators and their corresponding calculation formulas specifically include:
[0078] 1) The formula for calculating the first indicator, smoke arrival time in the stairwell (B10), is as follows: ;
[0079] 2) The calculation formula for the second indicator, the critical temperature arrival time (B11), is as follows: ;
[0080] 3) The calculation formula for the third indicator, evacuation start time (B12), is as follows: ;
[0081] 4) The calculation formula for the fourth indicator, the distribution of evacuation distances for all personnel (B13), is as follows: ;
[0082] 5) The formula for calculating the fifth indicator, the percentage of flue gas diffusion area (B20), is as follows: ;
[0083] 6) The calculation formula for the sixth indicator, Personnel Mobility Effectiveness Index (B21), is as follows: ;
[0084] 7) The formula for calculating the seventh indicator, evacuation time margin (B22), is as follows: ;
[0085] 8) The formula for calculating the eighth indicator, maximum heat release rate (B23), is as follows: ;
[0086] 9) The calculation formula for the ninth indicator, evacuation route congestion (B24), is as follows: ;
[0087] 10) The formula for calculating the tenth indicator, personnel evacuation efficiency (B25), is as follows: ;
[0088] 11) The calculation formula for the proportion of high-temperature damage area in the eleventh indicator (B30) is as follows: ;
[0089] 12) Regarding the twelfth indicator, the recoverability of evacuation routes (B31), the calculation formula is as follows: ;
[0090] 13) The calculation formula for the thirteenth indicator, the balance of evacuation route utilization (B32), is as follows: ;
[0091] 14) The calculation formula for the fourteenth indicator, flue gas removal time (B33), is as follows: ;
[0092] Step C: This step is a comprehensive evaluation calculation. After constructing a standardized index data matrix through the above 14 calculation formulas, the system calls the CRITIC-TOPSIS multi-criteria decision algorithm to perform dimensionality reduction and geometric space ranging on the matrix, and finally outputs an objective comprehensive score of system resilience.
[0093] In this implementation method, the principle of resilience evolution throughout the entire process is carried out in three steps: pre-disaster resistance, disaster absorption, and post-disaster recovery. The determination of the remaining thirteen indicators out of the twenty-seven is based on benchmark data in relevant standards.
[0094] 1) Building Basic Compliance and Physical Isolation Indicators (B14-B19): For the building fire resistance rating (B14) and the average fire resistance limit of structural components (B16), this invention relies on the International Organization for Standardization's "Test Methods for Fire Resistance of Building Components" (ISO 834) and "Standard for Types of Building Structures" (NFPA 220) for selection; for the compliance rate of fire compartment area (B15), the compliance rate of safety exit spacing (B18), and the reliability of fire-resistant doors and windows (B19), it strictly conforms to the core mandatory clauses in China's "Code for Fire Protection Design of Buildings" (GB50016); and the compliance rate of smoke control and exhaust system configuration (B17) is directly extracted based on the "Technical Standard for Smoke Control and Exhaust Systems in Buildings" (GB 51251).
[0095] 2) Initial Response and Situation Simulation Indicators (B10-B13): For evacuation activation time (B12) and total evacuation distance distribution (B13), parameters are calibrated according to the assessment criteria for crowd reaction and action time in China's "Fire Safety Engineering Part 9: Guidelines for Personnel Evacuation Assessment" (GB / T 31593.9); For smoke arrival time in stairwell (B10) and critical temperature arrival time (B11), the selection is mainly based on the judgment logic of the ASET (Available Safe Evacuation Time) pre-stage in the "SFPE Fire Engineering Manual".
[0096] (2) Selection of indicators for the disaster absorption stage: The indicators for this stage are the core of the system’s resilience in absorbing disaster energy. They mainly quantify the dynamic hazard state of fire, dynamic evacuation efficiency, and the tolerance limit of building structures.
[0097] 1) Fire dynamics and smoke control simulation indicators (B20, B23): For the maximum heat release rate (B23), the physical parameters of the fire source are calibrated according to the Chinese standard "Fire Safety Engineering Part 4: Setting Fire Scenarios and Selection of Setting Fires" (GB / T 31593.4); for the smoke diffusion area ratio (B20), the core reference is the performance evaluation requirements of the US "Smoke Control and Exhaust System Standard" (NFPA92), and it is mapped to the provisions of Chinese GB 51251 on the effectiveness of dynamic smoke control.
[0098] 2) Crowd evacuation resistance and efficiency indicators (B21, B22, B24, B25): For evacuation time margin (B22), personnel movement efficiency index (B21), evacuation route congestion (B24), and personnel evacuation action efficiency (B25), the indicators are extracted based on the evacuation width index for 100 people in the business hall in China's "Code for Fire Protection Design of Buildings" (GB 50016) and combined with the guiding principles for dynamic crowd trampling risk and bottleneck management in the UK's "Code for Fire Safety Design, Management and Use of Buildings" (BS 9999), to accurately quantify the risk of space failure during a disaster.
[0099] 3) Disaster-resistant structural and facility tolerance indicators (B26-B28): For the compliance rate of fire compartment integrity (B26), the compliance rate of net width of evacuation passage (B27), and the high-temperature strength retention rate of structural materials (B28), the indicators are constructed based on the description of the high-temperature decay characteristics of materials by the ISO 834 standard curve and the rigid requirements of GB 50016 for the net width of evacuation under fire conditions.
[0100] (3) Selection of indicators for the post-fire recovery phase: The indicators for this phase mainly assess the self-healing efficiency of the building's internal environment and the recovery efficiency of its structure and function after the disaster energy dissipates.
[0101] 1) Simulation Indicators for Physical Environment and Evacuation Network Recovery (B30-B33): For flue gas removal time (B33), the parameters are set according to the post-disaster flue gas removal technical guidelines in NFPA 92 and the specifications for exhaust ventilation rate in GB 51251; for the proportion of high-temperature damaged areas of structures (B30), the recoverability of evacuation routes (B31), and the balance of evacuation route utilization (B32), the parameters are quantified according to the evaluation theory of post-disaster network topology recovery and function restart in BS 9999 standard.
[0102] 2) Structural repairability and redundancy building indices (B34-B37): For structural repairability index (B34), fire protection facility availability rate (B35), building functional zoning replaceability (B36), and structural redundancy (B37), this invention follows the "SFPE Fire Protection Engineering Manual" and engineering reliability design specifications, and incorporates the reversible repair capability of buildings after extreme fires and equipment redundancy into the evaluation system, thus completing the building-level resilience closed loop.
[0103] In summary, the evaluation system constructed by this invention, which includes the above 27 indicators, has been supported by both international authoritative fire protection guidelines and current Chinese national mandatory standards for each of its physical thresholds, simulation logic, and building parameters, ensuring the legality and reliability of the evaluation framework in actual engineering applications of commercial complexes.
[0104] In this implementation method, the pre-disaster resistance indicators also include building fire resistance rating, fire compartment area compliance rate, average fire resistance limit of structural components, smoke control system configuration compliance rate, safety exit spacing compliance rate, and fire door and window closure reliability.
[0105] Disaster absorption indicators also include the fire compartment integrity compliance rate, the evacuation route net width compliance rate, and the high-temperature strength retention rate of structural materials.
[0106] Post-disaster recovery indicators also include structural repairability index, fire protection facility availability index, building functional zoning replaceability index, and structural redundancy index.
[0107] Furthermore, pre-disaster resilience indicators include: building fire resistance rating (B14), fire compartment area compliance rate (B15), average fire resistance limit of structural components (B16), smoke control system configuration compliance rate (B17), safety exit spacing compliance rate (B18), and fire door and window closing reliability (B19).
[0108] Disaster absorption indicators include: fire compartment integrity compliance rate (B26), evacuation route net width compliance rate (B27), and structural material high temperature strength retention rate (B28).
[0109] Post-disaster recovery indicators include structural repairability index (B34), fire protection facility availability rate (B35), building functional zoning replaceability (B36), and structural redundancy (B37).
[0110] Specific Implementation Method Nine: This implementation method is a further limitation of Specific Implementation Method One. In this implementation method, the standardized decision data matrix in the resilience evaluation model is sequentially processed by CRITIC weighting and TOPSIS evaluation sorting to obtain the CRITIC-TOPSIS quantitative evaluation model. The calculation process of the CRITIC-TOPSIS quantitative evaluation model is as follows: the standardized index data is imported into the evaluation algorithm kernel. First, CRITIC objective weighting is performed. By calculating the standard deviation of each index under different scenarios, the contrast strength index is obtained. The Pearson correlation coefficient between the indicators is calculated to obtain the quantitative conflict index. The contrast strength index and the quantitative conflict index are combined to calculate the comprehensive information content. Then, the weight distribution data of each index after objectively removing redundant information is obtained. Then, the system resilience score is calculated. The corresponding calculation process is as follows: based on the obtained objective weights, a weighted standardized decision matrix is constructed; the optimal and worst values of each index are extracted to generate positive ideal solutions and negative ideal solutions; then, the Euclidean geometric distance between each simulation scenario and the positive and negative ideal solutions is calculated; finally, through the relative closeness formula, the centroid method is used to determine a definite precise value for the output quantity, and the comprehensive system resilience score between 0 and 1 is calculated.
[0111] The system receives the overall resilience score output by the evaluation module and draws corresponding conclusions based on the weighting of each indicator. When the system identifies that the overall available safe evacuation time (ASET) is approaching its limit or the total resilience score is below the safety threshold, it automatically triggers a high-risk warning.
[0112] When the cumulative weight of macro-indicators such as smoke exhaust time or smoke diffusion ratio exceeds 68%, the system generates optimization suggestions for upgrading the smoke control system efficiency, guiding the re-deployment and optimization of passive smoke exhaust and mechanical active smoke exhaust in the building space.
[0113] When the arrival time of smoke in the stairwell or the congestion index of the evacuation route shows high sensitivity, a corresponding treatment plan is generated to improve the early multi-mode sensor warning and the configuration of smoke control pressure differential in the core evacuation shaft.
[0114] When the evacuation congestion is highly sensitive, a dynamic flow management solution based on digital twin space is output;
[0115] When the system calculates and finds that the overall available safe evacuation time is approaching the limit or the resilience score is extremely low, it sends an extremely high risk warning and automatically outputs the corresponding optimal intervention plan.
Claims
1. A multi-field coupled emergency dynamic simulation evaluation method for fire safety resilience of large and complex public buildings, characterized in that: The multi-field coupled simulation evaluation method constructs a building scenario model using acquired data on building floors, fire compartments, and evacuation distances. This model automatically generates multiple sets of FDS input data for different scenarios using Python batch scripts. Building geometry and fire scenario data are extracted from these FDS input data. A multi-field coupled simulation model for fire evacuation is then constructed using this data. This model provides full-process simulation data on fire temperature field, smoke concentration, visibility distribution, evacuation time, evacuation routes, and casualties in each scenario. A resilience evaluation model is then constructed using this full-process simulation data. This model extracts three types of resilience indicators: pre-disaster resistance, in-disaster absorption, and post-disaster recovery, and constructs a standardized decision data matrix. This standardized decision data matrix undergoes CRITIC weighting and TOPSIS evaluation and ranking calculations to obtain a resilience ranking result. The optimal resilience plan for the corresponding building layout and fire protection configuration is then determined based on the resilience ranking result.
2. The multi-field coupled emergency dynamic simulation evaluation method for fire safety resilience of large and complex public buildings according to claim 1, characterized in that: The process of constructing a building scene model using acquired building floor, fire compartment, and evacuation distance data, and then automatically generating multiple sets of FDS input data for different scenarios using Python batch scripts, involves obtaining basic building data from the large, complex public building under study. This basic data includes building floor, fire compartment, and evacuation distance data. Complex spatial geometry, floor distribution, and material properties are extracted from BIM models or IFC format files to establish an initial 3D basic model for the building scene model. An automated script engine developed based on Python is used to read and parse the underlying disaster prevention simulation files. Python code scripts are used to read, identify, and modify the underlying information of fire source coordinates and environmental parameters. Regular expressions are used to automatically align multiple spatiotemporal data, completing the batch overwriting of underlying parameters and the automated generation of multiple fire occurrence location scenarios. For different fire occurrence characteristics, regular expressions and other methods are used to automatically identify and modify the initial environmental parameters of the fire source, thereby obtaining multiple sets of FDS input data for different scenarios.
3. The multi-field coupled emergency dynamic simulation evaluation method for fire safety resilience of large and complex public buildings according to claim 1, characterized in that: Building geometry and fire scene data are extracted from multiple sets of FDS input data for different scenarios. A multi-field coupled simulation model for fire evacuation is then constructed using this data. The process of obtaining full-process simulation data on fire temperature field, smoke concentration, visibility distribution, evacuation time, evacuation path, and casualties under each scenario through this model involves parametric modeling and batch scripting using Python. This process iterates through variable parameters such as building floors, fire compartments, evacuation distance, fire source location, and fire source power to generate multiple sets of FDS input data for different scenarios. These FDS input data are presented as multiple differentiated FDS input text files, each corresponding to an independent building fire simulation scenario. The parsing program is then started, and each FDS input file is read in a loop. After reading the text data line by line from the input file, the corresponding valid parameters are retained. The key FDS data is identified from the valid parameters. The identified key FDS data is organized into a building geometry structured dataset. After matching and parsing the building geometry structured dataset, fire scene parameters are extracted. After parameter verification and unit normalization of the fire scene parameters, building geometry and fire scene data are formed. After obtaining the building geometry and fire scene data, the building geometry and fire scene data are combined with the Large Eddy Simulation (LES) method to complete the high-fidelity calculation of the fire dynamics inside the building. That is, the simulation mesh resolution and physical boundary conditions are set, and based on the mass conservation equation, momentum conservation equation, energy conservation equation and state equation, transient calculations are performed on the flow, diffusion rate and toxic gas concentration distribution of high-temperature smoke generated by the fire to construct a fire evacuation multi-field coupled simulation model. In the fire evacuation multi-field coupled simulation model, the eddy current effect of the atrium structure, the smoke stratification phenomenon in the corridor area and the local backflow characteristics under complex partitions are accurately captured. The calculation is then completed to generate three-dimensional slice and mesh feature data including temperature, visibility and carbon monoxide concentration. After acquiring three-dimensional slice and mesh feature data including temperature, visibility and carbon monoxide concentration, the Steering mode is used as the control algorithm for personnel evacuation path planning in the fire evacuation multi-field coupled simulation model. The personnel density and personnel behavior parameters of each area are set. The Steering mode is used to accurately calculate the micro-trajectory of personnel when avoiding obstacles and other escapees, and accurately identify the dynamic congestion bottleneck, peak number of people queuing and staying in the corresponding stairwell or other relevant information. Using a code interface, the time-dependent spatial smoke data generated by the fire simulation is synchronously mapped into the fire evacuation multi-field coupled simulation model. This ensures that virtual evacuees exhibit velocity attenuation or path replanning behavior when encountering visibility below the safety threshold or high-temperature hazard areas. After completing the spatiotemporal deep interactive coupling calculation of the physical field and the microscopic behavioral field, the required safe evacuation time RSET comprehensive evolution data is output. The required safe evacuation time RSET comprehensive evolution data is the full-process simulation data of fire temperature field, smoke concentration, visibility distribution, personnel evacuation time, personnel evacuation path, and personnel casualties in each scenario.
4. A multi-field coupled emergency dynamic simulation evaluation method for fire safety resilience of large and complex public buildings according to claim 1, 2 or 3, characterized in that: The process of constructing a resilience evaluation model using full-process simulation data, and then extracting three types of resilience indicators—pre-disaster resistance, disaster absorption, and post-disaster recovery—from the resilience evaluation model to build a standardized decision data matrix is as follows: First, a full-process multi-field coupled simulation process is executed, namely, importing the building model and generating different fire scenarios in batches; LES solves the smoke dynamics, and the Steering algorithm simulates personnel evacuation, completing the spatiotemporal interaction between the smoke hazard field and personnel evacuation behavior, and collecting heterogeneous raw data from the underlying simulation, such as temperature, visibility, evacuation congestion, and evacuation time margin. Using the massive simulation output data as the model input base, a CRITIC-TOPSIS resilience evaluation model is built: first, the raw simulation data is cleaned and normalized; the CRITIC algorithm calculates the standard deviation of the indicators and the Pearson correlation coefficient, calculates the information content of the indicators to complete objective weighting, and eliminates subjective scoring and information redundancy; Next, a weighted decision matrix is constructed, the positive and negative ideal solutions are solved, and the Euclidean distance and relative proximity of each scenario are calculated to obtain a 0-1 comprehensive resilience evaluation value, thus completing the resilience evaluation model construction process. After the resilience evaluation model is constructed, three types of resilience indicators are extracted from the resilience evaluation model: pre-disaster resistance, disaster absorption, and post-disaster recovery. A standardized decision data matrix is then constructed using these three resilience indicators. The standardized decision data matrix is a data matrix consisting of twenty-seven indicators; The pre-disaster resilience indicators in the standardized decision-making data matrix include the time it takes for smoke to reach the stairwell, the time it takes for the critical temperature to reach the peak, the evacuation start time, and the distribution of evacuation distances for all personnel. The disaster absorption indicators in the standardized decision data matrix include the proportion of smoke diffusion area, the efficiency index of personnel movement, the evacuation time margin, the maximum heat release rate, the congestion of evacuation routes, and the efficiency of personnel evacuation actions. The standardized decision-making data matrix includes post-disaster recovery indicators such as the proportion of structural high-temperature damage areas, the recoverability of evacuation routes, the balance of evacuation route utilization, and the smoke removal time.
5. The multi-field coupled emergency dynamic simulation evaluation method for fire safety resilience of large and complex public buildings according to claim 4, characterized in that: The calculation process for the smoke arrival time index in the stairwell is as follows: As a vertical evacuation shaft in a building, the timing of smoke intrusion directly determines the survival period of the safety defense line. The visibility function of the stairwell node is V(t), and the critical time when it first drops to the safety threshold is tracked. Combined with the benchmark safety time limit required by the specifications The relative abundance formula is as follows: The total number of stairwells will be extracted. and visibility Substitute it into the global score formula as a variable The calculation is performed to obtain the time index for smoke to reach the stairwell; The calculation process for the critical temperature arrival time index is as follows: when the ambient temperature of the evacuation route exceeds the human body's tolerance limit, people will lose their ability to move. The critical temperature arrival time index characterizes the building's time buffering capacity in high-temperature environments. The three-dimensional temperature field data T(x,y,t) at a height of Z=2.0m in the evacuation network is extracted, and the time to first exceed the critical temperature is calculated. Combined with the benchmark safety time limit required by the specifications The relative sufficiency formula is used to measure the status of a single line of defense: ; Extracting variables will extract the total number of stairwells. and extreme temperatures Substituting variables into the formula The calculation is performed to obtain the critical temperature arrival time index; The evacuation initiation time index is calculated by extracting the average time consumed during the agent's pre-action phase. Total required safe evacuation time Then, the average time consumed during the agent's pre-action phase. Total required safe evacuation time Construct the basic determination formula: The number of floors and areas that successfully triggered the visibility alarm was extracted. and the total number of all monitored areas Substituting variables into the formula The calculations were performed to obtain the evacuation start time index; The calculation process for the overall evacuation distance distribution index involves assessing the rationality of the building's internal spatial layout and evacuation network topology. The closer the actual evacuation distance is to the standard limit, the stronger the redundancy resistance of the spatial layout. The total actual evacuation paths ΣD of all agents are calculated. actual The sum of the ideal shortest paths ΣD ideal Based on the sum of the actual evacuation paths of all agents ΣD actual The sum of the ideal shortest paths ΣD ideal Construct the basic determination formula: Extract standard distance limits And to calculate the average evacuation distance of all personnel. Substituting variables into the formula The calculations were performed to obtain the distribution index of evacuation distances for all personnel.
6. The multi-field coupled emergency dynamic simulation evaluation method for fire safety resilience of large and complex public buildings according to claim 5, characterized in that: The calculation process for the flue gas diffusion area ratio index involves integrating the slice to obtain the instantaneous hazardous area and locking in the extreme value A. max_smoke With total area A total Construct the basic determination formula: ; Extract the peak proportion R of visibility dropping below the 10m threshold in the built area. max Substituting variables into the formula Calculations were performed to obtain the percentage of flue gas diffusion area. The calculation process of the personnel mobility efficiency index is a process of assessing the constraints and degree of soft disruption on the behavioral dynamics of a crowd in a harsh fire environment, and statistically analyzing the free movement speed of the agent in a fire-free state. Actual deceleration in dangerous areas Construct the basic determination formula: Extract the lowest recorded individual effective speed V min and the critical failure threshold V for severe congestion limit Substitute it into the formula as a variable The personnel mobility efficiency index was obtained through calculation. The evacuation time margin index is calculated by extracting the available safe evacuation time t based on the ASET-RSET theory. ASET With the necessary safe evacuation time t RSET As a variable, the basic determination formula is constructed: ; Re-extract available safe evacuation time t ASET and the required safe evacuation time t REST Substituting variables into the formula The evacuation time margin index is calculated. The calculation process for the maximum heat release rate index involves determining the ultimate thermal shock pressure that the entire smoke control system needs to absorb based on the fire source evolution intensity correspondence method, and then extracting the peak combustion power HRR. max With design limits HRR limit Construct the basic determination formula: Extract the peak HRR value Q after stabilization. max_smooth and the maximum heat capacity limit Q specified in the structural smoke exhaust design. limit Substituting variables into the formula The maximum heat release rate index is obtained through calculation; The calculation process for the congestion index of evacuation routes involves determining the route's capacity to handle sudden large crowds by quantifying the physical stagnation and stampede risk caused by excessively high local crowd density, and then integrating the total congestion time Σt. jam With total time t evac_total Construct the basic determination formula: Extract the maximum value of the total number of people (Occ) in the most crowded stairwell. max And the theoretical maximum load capacity C calculated based on the staircase area and extreme congestion density. max Substituting variables into the formula The congestion index of evacuation routes is calculated in the process. The calculation process for the personnel evacuation efficiency index is based on the overall evacuation network's macroscopic population flow efficiency in the face of a fire. This involves calculating the minimum evacuation time t under conditions of no fire obstruction. RSET_ideal Compared with the actual time t RSET_actual Construct the basic determination formula: Extract the average efficiency of all personnel. and the predefined free-flow efficiency threshold E free_flow Substituting variables into the formula The efficiency index of personnel evacuation operations was obtained through calculation.
7. The multi-field coupled emergency dynamic simulation evaluation method for fire safety resilience of large and complex public buildings according to claim 6, characterized in that: The calculation process for the structural high-temperature damage area ratio index involves assessing the extent of damage to the building structure under high-temperature exposure, determining the difficulty of restoring the building's functionality after a disaster, and extracting the surface area A whose fire temperature exceeds a threshold. damage With total area A total Construct the basic determination formula: ; Extract the peak value ratio R of concrete damage threshold at temperatures exceeding 300°C from the surface area of the structure. dmg_max Substituting variables into the formula The proportion of high-temperature damage area in the structure is obtained through calculation. The calculation process for the recoverability index of evacuation routes is based on assessing the flow recovery and carrying capacity of physical routes under the impact of pedestrian traffic. The recoverability index is determined by using the extreme value of the recovery time t. max t min and channel recovery time t recover_channel Construct the basic determination formula: Extract the calculated browsing limit F based on the step width. theoretical and the peak flow F dynamically observed on all bottleneck sections actual_peak Substituting variables into the formula The recoverability index of evacuation routes is obtained through calculation. The calculation process for the evacuation route utilization balance index is based on the Gini coefficient theory to assess whether the population can adaptively and evenly utilize all the route networks to determine the evacuation route utilization balance index, with x representing the number of people passing through n exits. i With a mean of µ, construct the basic determination formula: Extract the Gini coefficient G, calculated based on the cumulative distribution of evacuees along all available evacuation routes, and substitute it into the formula. The utilization balance index of evacuation routes is obtained through calculation. The calculation process for the smoke removal time index is based on quantifying the smoke control and exhaust system's exhaust efficiency in the later stages of the system to determine the system's agility in restoring the spatial environment to a tolerable state, and tracking the time t required for visibility restoration. clear and extreme values t max t min Construct the basic determination formula: Extract the precise timestamp t when the overall spatial tolerance drops below the 95% tolerance threshold. breach and total simulation duration t total_sim Substituting variables into the formula The flue gas removal time index is calculated in the process.
8. The multi-field coupled emergency dynamic simulation evaluation method for fire safety resilience of large and complex public buildings according to claim 7, characterized in that: Pre-disaster resilience indicators also include building fire resistance rating, fire compartment area compliance rate, average fire resistance limit of structural components, smoke control system configuration compliance rate, safety exit spacing compliance rate, and fire door and window closure reliability. Disaster absorption indicators also include the fire compartment integrity compliance rate, the evacuation route net width compliance rate, and the high-temperature strength retention rate of structural materials. Post-disaster recovery indicators also include structural repairability index, fire protection facility availability index, building functional zoning replaceability index, and structural redundancy index.
9. The multi-field coupled emergency dynamic simulation evaluation method for fire safety resilience of large and complex public buildings according to claim 1, characterized in that: The standardized decision data matrix is sequentially weighted by CRITIC and ranked by TOPSIS to obtain the resilience ranking results. The process of determining the optimal resilience plan for the corresponding building layout and fire protection configuration based on the resilience ranking results is as follows: In the resilience evaluation model, the standardized decision data matrix is sequentially processed through CRITIC weighting and TOPSIS evaluation sorting to obtain the CRITIC-TOPSIS quantitative evaluation model. The calculation process of the CRITIC-TOPSIS quantitative evaluation model is as follows: the standardized index data is imported into the evaluation algorithm kernel. First, CRITIC objective weighting is performed. By calculating the standard deviation of each index under different scenarios, the contrast strength index is obtained. The Pearson correlation coefficient between the indices is calculated to obtain the quantitative conflict index. The contrast strength index and the quantitative conflict index are combined to calculate the comprehensive information content. Then, the weight distribution data of each index after objectively removing redundant information is obtained. Then, the system resilience score is calculated. The corresponding calculation process is as follows: based on the obtained objective weights, a weighted standardized decision matrix is constructed; the optimal and worst values of each index are extracted to generate positive ideal solutions and negative ideal solutions; then, the Euclidean geometric distance between each simulation scenario and the positive and negative ideal solutions is calculated; finally, through the relative closeness formula, the centroid method is used to determine a definite precise value for the output quantity, and the comprehensive system resilience score between 0 and 1 is calculated. The system receives the overall resilience score output by the evaluation module and draws corresponding conclusions based on the weighting of each indicator. When the system identifies that the overall available safe evacuation time (ASET) is approaching its limit or the total resilience score is below the safety threshold, it automatically triggers a high-risk warning. When the cumulative weight of macro-indicators such as smoke exhaust time or smoke diffusion ratio exceeds 68%, the system generates optimization suggestions for upgrading the smoke control system efficiency, guiding the re-deployment and optimization of passive smoke exhaust and mechanical active smoke exhaust in the building space. When the arrival time of smoke in the stairwell or the congestion index of the evacuation route shows high sensitivity, a corresponding treatment plan is generated to improve the early multi-mode sensor warning and the smoke control pressure differential configuration of the core evacuation shaft. When the evacuation congestion is highly sensitive, a dynamic flow management solution based on digital twin space is output. When the system calculates and finds that the overall available safe evacuation time is approaching the limit or the resilience score is extremely low, it sends an extremely high risk warning and automatically outputs the corresponding optimal intervention plan.
