Compressed air foam fire extinguishing assessment method and system under influence of multiple environmental factors

Abstract

The invention discloses a compressed air foam fire extinguishing assessment method and system under the influence of multiple environmental factors. The method comprises the following steps: collecting environmental parameters of an extra-high voltage station in a fire scene to form a fire environment parameter library; constructing a three-dimensional model of the extra-high voltage station, performing environmental parameter analogue simulation on the extra-high voltage station, and selecting a simulated nested physical model to simulate a fire scene to form a simulated extra-high voltage station; respectively simulating the influence of a single environment parameter and multiple environment parameters on the fire extinguishing performance index in the simulation extra-high voltage station, obtaining simulation data associated with the fire extinguishing performance index, and recording extra-high voltage station fire test data; constructing a data set by using the simulation data and the extra-high voltage station fire test data, performing data expansion to obtain a virtual sample, and using the virtual sample and the data set as a database; a fire extinguishing performance prediction model is constructed and trained, environmental parameters of the extra-high voltage station are collected in real time, and fire extinguishing performance indexes are predicted; the method has the advantages of being accurate in evaluation result, short in calculation time and sufficient in data sample.

Description

Technical Field

[0001] The present invention relates to the technical field of fire safety in UHV substations, and particularly to an evaluation method and system for compressed air foam fire extinguishing under the influence of multiple environmental factors. Background Art

[0002] Due to the huge amount of oil in the main transformer of the UHV substation, its fire presents the characteristics of "rapid combustion, fast spread, and often accompanied by explosion shock". Coupled with the difficulty of manual fire extinguishing, it is easy to cause significant economic losses and safety threats. This poses higher requirements for fire extinguishing technology. The compressed air foam fire extinguishing system (Compressed Air Foam Fire Extinguishing System, abbreviated as CAFS) has become an important fire extinguishing means due to its wide coverage and high fire extinguishing efficiency. However, China has a vast territory, a large north-south span, and significant east-west differences. The complex and changeable environmental conditions pose new challenges to the performance of CAFS. For example, the difference in altitude leads to uneven distribution of environmental pressure: the environmental pressure of Ganzi UHV substation is only 66 kPa, while that of Changji-Guquan UHV substation can reach 101 kPa; seasonal changes bring significant temperature and humidity fluctuations, with an annual temperature difference of up to 50 °C and a humidity difference of 25% at the same location. In winter, the temperature and humidity differences between the north and south and between the east and west are further exacerbated. These factors directly affect the delivery performance and fire extinguishing effect of compressed air foam: changes in environmental pressure will affect the foam expansion ratio and spraying distance, temperature and humidity differences may change the foam stability and coverage performance, while wind speed and wind field interference significantly affect the foam delivery path and fire extinguishing efficiency.

[0003] UHV substations are distributed in different geographical environments, and their pressure, temperature, humidity and other conditions significantly affect the combustion characteristics of large power transformer fires and the fire extinguishing efficiency of the foam fire extinguishing system (CAFS). However, current research mainly focuses on the analysis of the fire extinguishing performance under a single environmental factor, lacking systematic research on the coupling effect of multiple environmental factors. For example, a method for evaluating the design parameters of a compressed air foam fire extinguishing system for UHV substations disclosed in Chinese Patent Publication No. CN118761546A only considers the environmental wind field to test the fire extinguishing performance of compressed air foam. At the same time, the complex interaction between different environmental boundary conditions (such as pressure, temperature, humidity and wind speed) further exacerbates the difficulty of fire extinguishing performance evaluation, making it difficult to provide comprehensive guidance for practical applications. Although experimental research can intuitively evaluate the performance of CAFS, its high cost and external environmental interference limit the accuracy and reliability of test results. In this case, numerical simulation technology can, to a certain extent, make up for the deficiencies of experiments, significantly reduce costs and avoid the limitations of experimental conditions. However, numerical simulation involves complex physical processes such as multiphase flow, mass transfer, heat transfer and turbulent combustion, which takes a long time to calculate, and due to insufficient data sample size, it is difficult to quickly cover a variety of environmental combinations. Summary of the Invention

[0004] The technical problem to be solved by the present invention is that the existing compressed air foam fire extinguishing evaluation method only studies the analysis of the influence of a single environmental factor on the fire extinguishing performance, the evaluation result is not accurate enough, and the calculation takes a long time and the data sample is insufficient.

[0005] The present invention solves the above technical problems by the following technical means: A compressed air foam fire extinguishing evaluation method under the influence of multiple environmental factors, including:

[0006] S1. Collect the environmental parameters in the fire scene of the UHV substation to form a fire environmental parameter library; the environmental parameters include environmental pressure, environmental temperature and humidity, and environmental wind speed;

[0007] S2. Construct a three-dimensional model of the UHV substation, reproduce the UHV substation 1:1 to obtain a simulated UHV substation, and select a physical model with nested simulation to simulate the fire scene;

[0008] S3. Based on the fire environmental parameter library, simulate the influence of a single environmental parameter and multiple environmental parameters on the fire extinguishing performance index in the simulated UHV substation respectively, obtain the simulated data associated with the fire extinguishing performance index and record the UHV substation fire test data; the fire extinguishing performance index includes fire extinguishing time, foam injection distance, foam coverage range and fire field temperature distribution;

[0009] S4. Use the simulated data and the UHV substation fire test data to construct a data set, use a generative adversarial network to expand the data set to obtain virtual samples, and use the virtual samples and the data set together as a database;

[0010] S5. Construct a fire extinguishing performance prediction model, the input layer of the fire extinguishing performance prediction model is the environmental parameter, the output layer is the predicted fire extinguishing performance index, use the database to train the fire extinguishing performance prediction model, obtain a trained fire extinguishing performance prediction model, collect the environmental parameters of the UHV substation in real time, predict the fire extinguishing performance index, and guide the engineering operation.

[0011] Beneficial effects: The present invention simulates the influence of a single environmental parameter and multiple environmental parameters on the fire extinguishing performance index, so as to consider not only a single environmental factor but also the coupling influence of multiple environmental factors, the evaluation result is relatively accurate, and uses a generative adversarial network to expand the data set to avoid local convergence of the model caused by insufficient data samples, further improving the accuracy of the model. It also builds a three-dimensional model, conducts environmental parameter simulation on the UHV substation and selects a physical model with nested simulation to simulate the fire scene, and calculates and simulates the corresponding data based on the simulated UHV substation, without repeatedly conducting actual experiments, thus taking a short time and saving a large amount of manpower and material costs.

[0012] Further, S2 includes:

[0013] Build a 3D model of the UHV substation to achieve a 1:1 reproduction of the actual test conditions. Set the corresponding environmental parameters according to the actual environmental data of the UHV substation. Determine the heat release rate of the fire source based on the test data, set the flow rate of the foam injection, and ensure that the injection range covers the fire source area. The nested physical models for simulation include the VOF model, the EDM eddy dissipation model, and the air age model. The VOF model simulates the flow process of multiple immiscible fluids. The EDM eddy dissipation model simulates the turbulent dissipation process of combustion in the fire scene. The air age model characterizes the transmission efficiency of the foam in the pipeline.

[0014] Furthermore, S2 also includes:

[0015] Set thermocouples and radiometers in the UHV substation and the simulated UHV substation to monitor the temperature distribution and the change of radiative heat flux in the combustion area. Select the fuel pool surface, the flame core area, and the fire extinguishing agent injection area, and set the temperature and radiative heat flux thresholds respectively. If the temperature in the flame area is lower than the set threshold, or the radiative heat flux at the measurement point decreases to the environmental background level, or the change rate of the radiative heat flux tends to zero, it is considered that the flame has extinguished, thus simulating the process of fire extinguishment.

[0016] Furthermore, the formula of the air age model is as follows:

[0017]

[0018] Among them, represents the partial derivative, describing the change rate of τ air with respect to t air ; t air is the real-time time during the fluid flow process, is the gradient operator; ρ is the fluid density; τ air is the air age, that is, the transmission time of the foam in the pipeline; y is the fluid velocity vector; Γ τ is the diffusion coefficient of the air age; S τ is the air age source term, that is, the local air age growth rate.

[0019] Furthermore, the nested physical models for simulation also include the flame thermal radiation intensity model, which is used to reflect the rate of fire energy release; the formula of the flame thermal radiation intensity model is as follows:

[0020]

[0021] Among them, represents the total radiant energy released by the flame, represents the external radiation per unit area, l and w are the length and width of the rectangular fire source respectively, N represents the number of point sources, ω j represents the weight of the jth point source, satisfying χr represents the radiation fraction, represents the mass loss rate per unit area, x0 represents the vertical distance from the radiometer to the flame surface, S j represents the distance between the j-th point source and the radiometer, ΔH c represents the calorific value of combustion of the combustible substance, θ represents the flame tilt angle, H t is the vertical height of the radiometer, H j represents the vertical height of the point source j, H is the flame height, L v is the latent heat of evaporation of the liquid at room temperature, k represents the heat conduction constant, T F represents the flame temperature, T l is the liquid temperature, h is the convective heat transfer coefficient, σ is the Stefan - Boltzmann constant, Φ F is the configuration factor, ε f is the emissivity of the high - temperature gas and the flame, ε l is the emissivity of the liquid, c pl ,p l and T0 are the specific heat capacity, density and initial temperature of the liquid respectively, γ and α are both constants, g is the acceleration due to gravity, ρ a and ρ1 are the densities of air and gas products respectively.

[0022] The air age model proposed by the present invention breaks through the limitations of the traditional pressure - flow transport model by introducing a time - scale variable, combining a custom source term and a diffusion coefficient, quantifies the transport time of foam in a complex pipe network, makes the prediction of the fire - fighting response time more accurate, and further optimizes the response strategy of the fire - fighting system. The flame thermal radiation model of the present invention adopts a mass loss rate coupling method, multi - point source radiation theory, and combines aerodynamic parameters to achieve high - precision calculation of the flame thermal radiation characteristics. Compared with the traditional single - thermal - radiation calculation method, the present invention can more comprehensively evaluate the impact of flame thermal radiation on the fire - fighting system and improve the adaptability of the fire - fighting system to different fire scenarios.

[0023] Further, S3 also includes: comparing the simulation data with the test data, and for the cases exceeding the preset error, adjusting the simulation boundary conditions or adjusting the parameters of the simulated UHV substation to make the deviation between the test data and the simulation data within the preset range.

[0024] Further, S5 also includes constructing a flame extinction determination model. The flame extinction determination model is based on a convolutional neural network. The binary flame image is input, and the determination result of whether the flame is extinguished is output. The convolutional neural network includes an input layer, a convolutional layer, and a fully connected layer. The convolutional layer includes a 3×3 convolutional kernel, a ReLU activation function, and a max pooling layer. The input layer receives the binary flame image. After the output result of the input layer passes through the 3×3 convolutional kernel, the ReLU activation function, and the max pooling layer, the determination result of whether the flame is extinguished is output through the fully connected layer.

[0025] Furthermore, the binary flame image is obtained by converting the original flame image into a grayscale image and then into a binary image, and this binary image is the binary flame image.

[0026] The present invention also provides a compressed air foam fire extinguishing evaluation system under the influence of multiple environmental factors, including:

[0027] A data acquisition module, which is used to collect environmental parameters in the fire scene of the UHV substation to form a fire environmental parameter library; the environmental parameters include environmental pressure, environmental temperature and humidity, and environmental wind speed;

[0028] A simulation module, which is used to construct a three-dimensional model of the UHV substation, reproduce the UHV substation 1:1 to obtain a simulated UHV substation, and select a physical model for simulation nesting to simulate the fire scene;

[0029] A parameter analysis module, which is used to simulate the influence of single environmental parameters and multiple environmental parameters on the fire extinguishing performance indicators in the simulated UHV substation based on the fire environmental parameter library, obtain the corresponding simulation data and record the test data corresponding to the UHV substation; the fire extinguishing performance indicators include fire extinguishing time, foam injection distance, foam coverage range, and fire field temperature distribution;

[0030] A training set construction module, which is used to construct a data set using the test data and the simulation data, perform data augmentation on the data set using a generative adversarial network to obtain virtual samples, and use the virtual samples and the data set together as a database;

[0031] A fire extinguishing performance prediction module, which is used to construct a fire extinguishing performance prediction model. The input layer of the fire extinguishing performance prediction model is the environmental parameters, and the output layer is the predicted fire extinguishing performance indicators. The database is used to train the fire extinguishing performance prediction model to obtain a trained fire extinguishing performance prediction model, and the environmental parameters of the UHV substation are collected in real time to predict the fire extinguishing performance indicators.

[0032] Further, the simulation module is also used for:

[0033] Build a 3D model of the UHV substation to achieve a 1:1 reproduction of the actual test conditions. Set the corresponding environmental parameters according to the actual environmental data of the UHV substation. Determine the heat release rate of the fire source based on the test data, set the flow rate of the foam injection, and ensure that the injection range covers the fire source area. The nested physical models for simulation include the VOF model, the EDM eddy dissipation model, and the air age model. The VOF model simulates the flow process of multiple immiscible fluids. The EDM eddy dissipation model simulates the turbulent dissipation process of combustion in the fire scene. The air age model characterizes the transmission efficiency of the foam in the pipeline.

[0034] Furthermore, the simulation module is also used for:

[0035] Install thermocouples and radiometers in the UHV substation and the simulated UHV substation to monitor the temperature distribution and the change of radiant heat flux in the combustion area. Select the fuel pool surface, the flame core area, and the fire extinguishing agent injection area, and set the temperature and radiant heat flux thresholds respectively. If the temperature in the flame area is lower than the set threshold, or the radiant heat flux at the measurement point decreases to the environmental background level, or the change rate of the radiant heat flux tends to zero, it is considered that the flame has extinguished, thus simulating the process of fire extinguishment.

[0036] Furthermore, the formula of the air age model is as follows:

[0037]

[0038] Where, represents the partial derivative, describing the change rate of τ air with respect to t air ; t air is the real-time time during the fluid flow process, is the gradient operator; ρ is the fluid density; τ air is the air age, that is, the transmission time of the foam in the pipeline; u is the fluid velocity vector; Γ τ is the diffusion coefficient of the air age; S τ is the air age source term, that is, the local air age growth rate.

[0039] Furthermore, the nested physical models for simulation also include the flame thermal radiation intensity model, which is used to reflect the rate of fire energy release; the formula of the flame thermal radiation intensity model is as follows:

[0040]

[0041] Where, represents the total radiant energy released by the flame, represents the external radiation per unit area, l and w are the length and width of the rectangular fire source respectively, N represents the number of point sources, ω j represents the weight of the jth point source, satisfying χ r represents the radiation fraction, represents the mass loss rate per unit area, x0 represents the vertical distance from the radiometer to the flame surface, S j represents the distance between the j-th point source and the radiometer, ΔH c represents the calorific value of combustion of the combustible substance, θ represents the flame tilt angle, H t is the vertical height of the radiometer, H j represents the vertical height of the point source j, H is the flame height, L v is the latent heat of evaporation of the liquid at room temperature, k represents the heat conduction constant, T F represents the flame temperature, T l is the liquid temperature, h is the convective heat transfer coefficient, σ is the Stefan-Boltzmann constant, Φ F is the configuration factor, ε F is the emissivity of the hot gas and the flame, ε l is the emissivity of the liquid, c pl ,p l and T0 are the specific heat capacity, density and initial temperature of the liquid respectively, γ and α are both constants, g is the acceleration due to gravity, ρ a and ρ1 are the densities of air and gas products respectively.

[0042] Furthermore, the parameter analysis module is also used for: comparing the simulation data with the test data, and for cases where the preset error is exceeded, adjusting the simulation boundary conditions or adjusting the parameters of the simulated UHV substation so that the deviation between the test data and the simulation data is within the preset range.

[0043] Furthermore, the fire extinguishing performance prediction module further includes constructing a flame extinction determination model. The flame extinction determination model is based on a convolutional neural network. The binary flame image is input, and the determination result of whether the flame extinguishes is output. The convolutional neural network includes an input layer, a convolutional layer and a fully connected layer. The convolutional layer includes a 3×3 convolutional kernel, a ReLU activation function and a max pooling layer. The input layer receives the binary flame image, and the output result of the input layer passes through the 3×3 convolutional kernel, the ReLU activation function and the max pooling layer and then is output by the fully connected layer as the determination result of whether the flame extinguishes.

[0044] Even further, the binary flame image is obtained by converting the original flame image into a grayscale image and then into a binary image, and this binary image is the binary flame image.

[0045] The advantages of the present invention are:

[0046] (1) The present invention simulates the influence of single environmental parameters and multiple environmental parameters on fire extinguishing performance indicators, thereby considering not only single environmental factors but also the coupled influence of multiple environmental factors. The evaluation results are relatively accurate. Moreover, the generative adversarial network is used to expand the dataset, avoiding local convergence of the model caused by insufficient data samples, and further improving the accuracy of the model. Additionally, by building a 3D model, environmental parameter simulation of the UHV substation is carried out, and a physical model of simulation nesting is selected to simulate the fire scene. Based on the simulated UHV substation, corresponding data calculation and simulation are performed, eliminating the need for repeated actual experiments, thus consuming less time.

[0047] (2) The present invention designs an air age model to simulate the transmission efficiency of foam in the pipeline, better reflecting the transport behavior of foam in the pipeline, thereby reflecting the response speed and coverage uniformity of the fire extinguishing system, making the simulation results closer to the real scene, and enabling the obtained test data to be better used for subsequent training, further improving the accuracy of the fire extinguishing performance prediction model.

[0048] (3) The present invention combines the temperature threshold method and the thermal radiation flux method, uses measuring points such as thermocouples and radiometers to monitor the temperature and thermal radiation flux changes in the combustion area, and combines binary image processing for flame extinction determination, enhancing the model's automatic recognition ability for the fire extinguishing end point and ensuring accurate judgment of fire extinction.

[0049] (4) The present invention designs a flame thermal radiation intensity model. The flame thermal radiation intensity model directly reflects the rate of fire energy release, which is a key parameter for evaluating the fire suppression ability of the fire extinguishing system. Through this model, the fire suppression ability of the substation can be better reflected, making the prediction results more accurate.

[0050] (5) The air age model proposed by the present invention breaks through the limitations of the traditional pressure-flow transport model by introducing a time scale variable, combining a custom source term and diffusion coefficient, quantifying the transmission time of foam in complex pipe networks, making the prediction of the fire extinguishing response time more accurate, and then optimizing the response strategy of the fire extinguishing system. The flame thermal radiation model of the present invention adopts a mass loss rate coupling method, multi-point source radiation theory, and combines aerodynamic parameters to achieve high-precision calculation of flame thermal radiation characteristics. Compared with traditional single thermal radiation calculation methods, the present invention can more comprehensively evaluate the influence of flame thermal radiation on the fire extinguishing system and improve the adaptability of the fire extinguishing system to different fire scenes. BRIEF DESCRIPTION OF THE DRAWINGS

[0051] Figure 1 is a flowchart of a compressed air foam fire extinguishing evaluation method under the influence of multiple environmental factors disclosed in an embodiment of the present invention;

[0052] Figure 2Schematic diagram of machine learning to expand data in a compressed air foam fire extinguishing evaluation method affected by multiple environmental factors disclosed in an embodiment of the present invention;

[0053] Figure 3 Schematic diagram of the interface button of the fire extinguishing system formed by a compressed air foam fire extinguishing evaluation method affected by multiple environmental factors disclosed in an embodiment of the present invention. Detailed implementation manners

[0054] To make the objectives, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the embodiments of the present invention. Apparently, the described embodiments are part of the embodiments of the present invention, rather than all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

[0055] Embodiment 1

[0056] As Figure 1 shown, Embodiment 1 of the present invention provides a compressed air foam fire extinguishing evaluation method affected by multiple environmental factors, including:

[0057] S1. Collect environmental parameters in the fire scenario of the UHV substation to form a fire environmental parameter library; the environmental parameters include environmental pressure, environmental temperature and humidity, and environmental wind speed; the specific process is as follows:

[0058] First, investigate and analyze the environmental parameters, test fire source parameters and pipeline network parameters in the substation area, and clarify their change ranges. The fire source parameters include heat release rate (500kW - 10MW) and fire source scale (1 - 10m 2 ). The pipeline network parameters cover information such as the foam injection port diameter (10 - 50mm), pipe length (10 - 100m), pipe diameter (25 - 100mm), flow rate (10 - 500L / min), etc. The environmental parameters include environmental pressure (50 - 100kPa), temperature and humidity (-30°C to 50°C, 10% - 90%), and wind speed (0 - 30m / s), and establish a comprehensive fire environmental parameter library.

[0059] S2. Build a 3D model of the UHV substation, perform environmental parameter simulation on the UHV substation, and select a physical model for simulation nesting to simulate the fire scenario, so as to reproduce the UHV substation 1:1 and obtain a simulated UHV substation; the specific process is as follows:

[0060] S21. Simulation reproduction

[0061] Based on the computational fluid dynamics (CFD) method, and achieve a 1:1 reproduction of the actual test conditions. According to the actual environmental data of the substation, set the corresponding environmental parameters, including environmental pressure (50 - 100 kPa), temperature and humidity (-30°C to 50°C, 10% - 90%), and wind speed (0 - 30 m / s). Determine the heat release rate of the fire source (500 kW - 10 MW) according to the test data, set the flow rate of foam injection (10 - 500 L / min), and ensure that the injection range covers the fire source area.

[0062] S22. Model construction

[0063] Construct a three-dimensional model of the substation, including the fire source, foam injection device, and environmental conditions, and strive to restore the real fire scene. Adopt the method of gradually refining the grid. Start from a relatively coarse initial grid, and optimize the calculation accuracy by gradually adjusting the grid density, and finally determine the optimal grid configuration.

[0064] S23. Selection of physical model

[0065] Select a suitable physically nested model. The multiphase flow model is one of the most important and widely applied models. There are three selectable multiphase flow models:

[0066] ① The VOF model is suitable for simulating the flow process of multiple immiscible fluids. During the calculation process, the flow of all fluids shares a single momentum equation, and the volume percentage of each part of the fluid can be recorded within the entire flow field calculation unit.

[0067] ② The Euler model is one of the most complex multiphase flow models. This model establishes a momentum equation containing n parameters to solve each item in the multiphase fluid.

[0068] ③ The mixture model is a simplified model in the multiphase flow model. During the calculation process, each item in the multiphase fluid is regarded as being continuously coupled with each other, and the overall momentum equation after coupling is used for solution.

[0069] Combined with the content requirements of this study, involving foam injection, it is necessary to clearly observe the foam injection situation. In the multiphase flow model, the VOF model is suitable for simulating the flow process of multiple immiscible fluids. During the calculation process, the flow of all fluids shares a single momentum equation, and the volume percentage of each part of the fluid can be recorded within the entire flow field calculation unit. Therefore, the VOF model of the multiphase flow model is selected. The calculation formula is as follows:

[0070]

[0071] In the formula, is the mass transfer from phase q to phase p; is the mass transfer from the p-phase to the q-phase. By default, the source term S on the right side of the equation is zero, but the user can specify a constant or a custom mass source for each phase. αq

[0072] During the simulation process, it is necessary to study the wind resistance performance of the compressed air foam gun system. The foam is simplified as a ternary two-phase fluid, regarded as a separated phase composed of a gas-liquid mixture, and its density and viscosity are calculated by the following formulas:

[0073] ρ f = ρ g + ρ l (1 - α l )

[0074] μ f = μ g + μ l (1 - α l )

[0075] In the formula, ρ g , ρ l and ρ f are the densities of the gas phase, liquid phase and foam phase respectively, μ g , μ l and μ f are the viscosities in the liquid phase, gas phase and foam phase respectively, and α l is the volume fraction of the liquid phase in the foam phase.

[0076] is to describe the turbulent dissipation process of combustion in the fire field. The EDM eddy dissipation model. In this model, the net production rate of the substance generated by the reaction is controlled by the smaller formula in the following formula:

[0077]

[0078] Among them, ν j ′ ,r is the stoichiometric number of reactant i, ν j ′,′ r is the stoichiometric number of product i, Y R is the mass fraction of the reactant, Y P is the mass fraction of the product, and A and B are empirical constants, which are 4 and 0.5 respectively.

[0079] ④ The present invention also designs an air age model. The air age characterizes the transmission efficiency of the foam in the pipeline and directly affects the response speed and coverage uniformity of the fire extinguishing system. The air age model is introduced, and its transport equation is:

[0080]

[0081] Among them, Denotes the partial derivative and describes τ air with respect to t air ; t air is the real-time time during the fluid flow process, is the gradient operator; ρ is the fluid density (kg / m 3 ); τ is the air age (s), that is, the transport time of the foam in the pipeline; u is the fluid velocity vector (m / s); Γ τ is the diffusion coefficient of the air age (m 2 / s), usually related to the turbulent viscosity; S τ is the air age source term (s -1 ), used to define the local air age growth rate.

[0082] The air age is solved by defining the source term and the diffusion coefficient to accurately track the foam flow time. The boundary condition is set as τ air = 0, and the outlet is the free outflow condition. To ensure the reasonable propagation of the air age within the computational domain. At the same time, the physical properties parameters are dynamically updated by combining the foam phase fraction calculated by the VOF model to improve the accuracy of the air age calculation and make it better reflect the transport behavior of the foam in the pipeline.

[0083] The present invention also designs a flame thermal radiation intensity model. The flame thermal radiation intensity directly reflects the rate of fire energy release and is a key parameter for evaluating the fire suppression ability of the fire extinguishing system. As an indirect parameter, it can be determined according to the following model formula by correlating the direct parameter (fire source size). The fire source thermal radiation intensity formula is as follows:

[0084]

[0085]

[0086] Wherein, represents the total radiant energy released by the flame, represents the external radiation per unit area, l and w are the length and width of the rectangular fire source respectively, N represents the number of point sources, ω j represents the weight of the j-th point source, satisfying χ r represents the radiation fraction, represents the mass loss rate per unit area, x0 represents the vertical distance from the radiometer to the flame surface, S j represents the distance between the j-th point source and the radiometer, ΔH c represents the calorific value of the combustible substance, θ represents the flame tilt angle, H t is the vertical height of the radiometer, H j represents the vertical height of the point source j, H is the flame height, L vis the latent heat of evaporation of the liquid at room temperature, k represents the heat conduction constant, T F represents the flame temperature, T l is the liquid temperature, h is the convective heat transfer coefficient, σ is the Stefan - Boltzmann constant, Φ F is the configuration factor, ε F is the emissivity of the high - temperature gas and the flame, ε l is the emissivity of the liquid, c pl ,p l and T0 are the specific heat capacity, density and initial temperature of the liquid respectively, γ and α are both constants, g is the gravitational acceleration, ρ a and ρ1 are the densities of air and gas products respectively.

[0087] S24. Model Validation

[0088] Verify the predictive ability of the model by comparing the simulation results and experimental data, especially indicators such as extinguishing time, foam injection distance, foam coverage, and fire field temperature distribution.

[0089] Analyze the error between the simulation results and experimental data, and evaluate whether the error is within a reasonable range. The error of the extinguishing time should be controlled within an acceptable range (usually less than 10%). If the error is too large, it is necessary to adjust the model parameters or select other more suitable physical models.

[0090] S25. Fire Extinguishing Criterion

[0091] Combined with measuring points such as thermocouples and radiometers, monitor the temperature distribution and the change of radiant heat flux in the combustion area to judge whether the flame is extinguished. Select key measuring points such as the fuel pool surface, the flame core area, and the fire extinguishing agent injection area, and set temperature and radiant heat flux thresholds respectively. If the temperature in the flame area is lower than the set threshold, or the radiant heat flux of the measuring point drops to the ambient background level, or its change rate tends to zero, it is considered that the flame is extinguished.

[0092] S3. Based on the fire environment parameter library, simulate the influence of single environmental parameters and multiple environmental parameters on the fire extinguishing performance indicators in the simulated UHV substation respectively, obtain the corresponding simulation data and record the corresponding experimental data of the UHV substation; the fire extinguishing performance indicators include extinguishing time, foam injection distance, foam coverage, and fire field temperature distribution; the specific process is as follows:

[0093] S31. Single Environmental Parameter Analysis

[0094] By simulating the influence of environmental parameters (ambient pressure, temperature and humidity, and wind speed) on the processes such as fire extinguishing time, foam coverage range, foam injection distance, and fire field temperature distribution. Set boundary conditions according to the data in actual tests, including environmental parameters, fire source parameters, pipe network parameters, etc., to ensure that the simulation conditions can truly reflect the actual fire scenario. Use the method of controlling variables, and separately adjust a single environmental parameter (such as ambient pressure) for simulation, while keeping other parameters fixed, to study its influence on the fire extinguishing process. Taking ambient pressure as an example:

[0095] Simulate the combustion process of the fire and the foam coverage under different pressure conditions (such as 80 kPa, 100 kPa, 120 kPa);

[0096] Obtain fire extinguishing parameters such as fire extinguishing time, foam coverage range, foam injection distance, and fire field temperature distribution;

[0097] In the analysis of other environmental parameters such as temperature and humidity, and wind speed, conduct one-by-one simulations and data collection as well.

[0098] By analyzing the errors between the simulation and the test (usually controlled within 5%-10%), especially for key indicators such as fire extinguishing time, foam injection distance, and fire field temperature distribution, evaluate the accuracy and rationality of the numerical model. For parameters with large errors, it is necessary to trace back the simulation boundary conditions or model settings and further optimize them.

[0099] S32. Simulation of Multiple Environmental Parameters

[0100] Extract fire extinguishing parameters affected by different environmental parameters (ambient pressure, temperature and humidity, and wind speed), including fire extinguishing time t, foam injection distance, foam coverage range, and fire field temperature distribution. Under different fire source scenarios, analyze the contribution degree of each environmental parameter to the fire extinguishing performance respectively. Combine the simulation data, calculate the influence weights of each environmental factor, and construct a weight index system:

[0101] Ambient pressure weight (w p ): Reflects the influence degree of ambient pressure on the fire extinguishing time.

[0102] Ambient temperature weight (w T ): Used to evaluate the effect of temperature change on foam fluidity and fire extinguishing efficiency.

[0103] Ambient humidity weight (w H ): Represents the influence of humidity on foam stability and injection performance.

[0104] Wind speed weight (w v ): Considers the disturbance effect of wind speed on the foam injection trajectory and coverage range.

[0105] Establish a time prediction model under the influence of multiple environmental parameters:

[0106] t = aw p + bw T + cw H + dw v + e

[0107] t: Fire extinguishing time (s)

[0108] p: Ambient pressure (Pa)

[0109] v: Wind speed (m / s)

[0110] T: Ambient temperature (°C)

[0111] H: Ambient humidity (%)

[0112] a, b, c, d, e: Undetermined coefficients, determined by regression analysis of simulation data.

[0113] S4. Construct a data set using experimental data and simulation data, use a generative adversarial network to augment the data set to obtain virtual samples, and use the virtual samples and the data set together as a database; the specific process is as follows:

[0114] This step combines experimental data and simulation results, uses machine learning methods to augment and optimize the data, generates virtual data samples covering more environmental conditions to make up for the lack of data volume. In addition, combined with binary image processing, it is used to train the neural network model to automatically identify the flame extinction state in the follow-up, improving the objectivity of fire extinguishing determination. Select PyTorch as the machine learning framework, combined with its powerful data processing and model training functions, as Figure 2 shown, the specific steps are as follows:

[0115] S41. Data preprocessing:

[0116] Use the Dataset and DataLoader classes of PyTorch to encapsulate the simulation results into iterable training samples, including environmental factors (pressure, temperature, humidity, wind speed) and corresponding fire extinguishing performance indicators (fire extinguishing time, foam coverage, foam injection distance, fire field temperature distribution). Use the NumPy class of PyTorch to normalize or standardize numerical data to ensure that the scales of different variables are consistent. Split the data into training set, validation set and test set, and the ratio is usually 70%:20%:10%.

[0117] S42. Data augmentation:

[0118] ① Virtual data augmentation (GAN-generated data)

[0119] Augment the simulation data using a Generative Adversarial Network (GAN). The GAN consists of a Generator and a Discriminator. The Generator learns the existing data distribution and generates realistic virtual data, while the Discriminator is used to distinguish between real data and generated data, thereby improving the authenticity of the generated samples. Build a GAN model using PyTorch, including the network structures of the Generator and the Discriminator. Input environmental parameters (such as pressure, temperature, humidity, wind speed, etc.) as conditional variables, and generate virtual fire extinguishing data that meets specific environmental conditions through a conditional GAN.

[0120] Randomly sample the latent variable z in the Generator, combine it with the conditional variables (environmental parameters), and input them into the Generator to generate virtual fire extinguishing performance data. The Discriminator is responsible for distinguishing between the generated data and the real data, and trains the Generator and the Discriminator through backpropagation to make the generated data gradually approach the real data distribution. GAN belongs to the prior art and will not be elaborated here.

[0121] Screen the generated data and eliminate virtual data that does not conform to physical laws (such as the fire extinguishing time significantly exceeding the reasonable range) or deviates from the experimental results.

[0122] ② Image data augmentation (flame binary determination)

[0123] Accurately judge whether the flame is completely extinguished through visual images, avoid subjective errors, and improve the objectivity of the fire extinguishing end point. Use image processing technology to convert the original flame image into a grayscale image, and then convert it into a binary image. Here, 1 represents the existence of a flame, and 0 represents the background. By averaging each pixel of the binary image, the intermittent contour of the flame is obtained.

[0124] S43. Generate a database:

[0125] Integrate the preprocessed and augmented data to generate a comprehensive database. This database will serve as the data source for model training, and the data needs to be appropriately indexed and optimized for efficient access during the model training process. For each flame image, attach a binary label indicating whether it is extinguished (0 = extinguished, 1 = burning); combine information such as the fire extinguishing time and temperature distribution to create a multi-modal dataset (numerical + image).

[0126] S5. Build a fire extinguishing performance prediction model. The input layer of the fire extinguishing performance prediction model is the environmental parameters, and the output layer is the predicted fire extinguishing performance indicators. Use the database to train the fire extinguishing performance prediction model to obtain a trained fire extinguishing performance prediction model, and collect the environmental parameters of the UHV substation in real time to predict the fire extinguishing performance indicators. The specific process is as follows:

[0127] S51. MLP model architecture:

[0128] Using a multi-layer perceptron (MLP), a fire extinguishing performance prediction model based on conditional variables is constructed:

[0129] Input layer: environmental parameters p, T, H, v

[0130] Hidden layer: Set 2 - 3 layers, each layer contains 64 - 128 neurons, and the ReLU activation function is used.

[0131] Output layer: Predict fire extinguishing performance indicators (extinguishing time, foam jet distance, foam coverage, fire field temperature distribution).

[0132] Training and optimization:

[0133] Loss function: The mean squared error (MSE) is used to measure the error between the predicted value and the true value:

[0134]

[0135] n: The total number of samples, that is, the number of data points in the dataset.

[0136] y i : The true value (actual value) of the i-th sample.

[0137] The predicted value of the i-th sample (model output).

[0138] The squared error of the i-th sample, indicating the degree of deviation between the predicted value and the true value.

[0139] The mean operation, normalizing the sum of squared errors to each sample.

[0140] Optimizer: Select the Adam optimizer, and the initial learning rate is set to 0.001. By adjusting the parameters of the model (such as the weights and biases of the neural network), the loss function (such as the mean squared error MSE) is minimized to improve the performance of the model.

[0141] Use augmented data and experimental data for training simultaneously to improve the generalization ability of the model.

[0142] S52, CNN model architecture:

[0143] Flame extinction determination is based on a convolutional neural network (CNN). After inputting a binary flame image and extracting flame features, classification is performed. The model structure is as follows:

[0144] Input layer: Binary flame image (size 64×64);

[0145] Convolutional layer: 3×3 convolutional kernel, ReLU activation, max pooling (2×2);

[0146] Fully connected layer: 128 neurons, ReLU activation, output 2 classes (0 = extinguished, 1 = burning);

[0147] Training and optimization:

[0148] Use the cross - entropy loss function (CrossEntropyLoss) to calculate the classification error:

[0149]

[0150] Optimizer: Select the Adam optimizer (learning rate 0.001) to optimize the network parameters and improve the classification accuracy.

[0151] Dataset: Use real fire - extinguishing images + augmented data for training.

[0152] Finally, evaluate the performance of the above - mentioned method, and the evaluation process is as follows:

[0153] Use the validation set to evaluate the model performance, and the metrics include mean squared error (MSE), mean absolute error (MAE) and coefficient of determination (R 2 )

[0154] Divide the data into a training set and a validation set. Continuously optimize the model during the training process, and at the same time monitor the model performance through the validation set, recording the MSE, MAE, and R 2 after each round of training. Select the optimal parameters or model structure according to the evaluation results of the validation set.

[0155] Apply the selected model to the test set and calculate the evaluation metrics. If the performance of the test set is close to the results of the validation set, it indicates that the model has good generalization ability for unknown data.

[0156] Use an independent test set to evaluate the flame - extinguishing recognition model, using the following metrics: Among them, TP (true positive) represents the number of samples correctly identified as flames; FP (false positive) represents the number of samples misjudged as flames. FN (false negative) represents the number of samples misjudged as extinguished. Reflects the balance between the precision and recall of the model.

[0157] As Figure 3 shown, based on the above research results, the present invention proposes a comprehensive fire - extinguishing efficiency evaluation system, including functions such as parameter input and setting, scene simulation, and result export. The system functions include:

[0158] 1. Parameter setting:

[0159] Support the input of fire source parameters: fire source length, fire source width;

[0160] Support the input of foam parameters: foam flow rate, foam density;

[0161] Provide the input of pipeline network parameters: including pipeline network injection diameter, pipe length, pipe diameter;

[0162] Environmental parameter setting: Allow to set environmental pressure, temperature, humidity and wind speed, etc., to meet the custom requirements of multi-scenario parameters.

[0163] 2. Simulation calculation: Built-in efficient numerical simulation calculation function, based on CFD technology, analyze the foam fire extinguishing performance in real time; Output key performance indicators, such as fire extinguishing time, foam injection distance, foam coverage range and fire field temperature distribution; Support dynamic adjustment of input parameters to realize performance evaluation under multi-condition conditions.

[0164] 3. Result export: Support to export a detailed fire extinguishing performance report, covering key data such as fire field temperature change, foam coverage range, foam injection distance and fire extinguishing time.

[0165] Finally, by integrating experimental and simulation data, embed the fire extinguishing performance prediction model into the system, provide an optimized suggestion plan, and improve the system's ability to adapt to multi-environmental factors.

[0166] Through the above technical solutions, the present invention first integrates open-source experimental data, uses the experimental results to verify the simulation model, clarifies the influence law of environmental parameters on the fire extinguishing performance, and ensures that the model can truly reflect the fire scene. Further combined with machine learning technology, through learning the experimental data and simulation results, generate virtual data samples covering more environmental conditions, expand the scale of the simulation data set, construct a comprehensive fire extinguishing performance database, and further establish a compressed air foam fire extinguishing evaluation method and system based on the influence of multiple environmental factors.

[0167] Embodiment 2

[0168] Based on Embodiment 1, Embodiment 2 of the present invention further provides a compressed air foam fire extinguishing evaluation system under the influence of multiple environmental factors, including:

[0169] A data acquisition module, used to collect environmental parameters in the fire scene of the UHV substation to form a fire environmental parameter library; The environmental parameters include environmental pressure, environmental temperature and humidity, and environmental wind speed;

[0170] A simulation module, used to build a three-dimensional model of the UHV substation, conduct environmental parameter simulation on the UHV substation and select a physical model for simulation nesting to simulate the fire scene, so as to reproduce the UHV substation 1:1 and obtain a simulated UHV substation;

[0171] A parameter analysis module, which is used to simulate the influence of single environmental parameters and multiple environmental parameters on the fire extinguishing performance indexes respectively in a simulated UHV substation based on a fire environment parameter library, obtain corresponding test data and record the test data corresponding to the UHV substation; the fire extinguishing performance indexes include fire extinguishing time, foam injection distance, foam coverage range and fire field temperature distribution;

[0172] A training set construction module, which is used to construct a data set by using test data and simulation data, perform data augmentation on the data set by using a generative adversarial network to obtain virtual samples, and use the virtual samples and the data set together as a database;

[0173] A fire extinguishing performance prediction module, which is used to construct a fire extinguishing performance prediction model. The input layer of the fire extinguishing performance prediction model is environmental parameters, and the output layer is predicted fire extinguishing performance indexes. Use the database to train the fire extinguishing performance prediction model to obtain a trained fire extinguishing performance prediction model, and collect the environmental parameters of the UHV substation in real time to predict the fire extinguishing performance indexes.

[0174] Specifically, the simulation module is also used for:

[0175] Construct a 3D model of the UHV substation to achieve a 1:1 reproduction of the actual test conditions. Set corresponding environmental parameters according to the actual environmental data of the UHV substation, determine the heat release rate of the fire source according to the test data, set the flow rate of foam injection, and ensure that the injection range covers the fire source area; the physically nested simulation models include the VOF model, the EDM eddy dissipation model and the air age model. The VOF model simulates the flow process of multiple immiscible fluids, the EDM eddy dissipation model simulates the turbulent dissipation process of combustion in the fire field, and the air age model characterizes the transmission efficiency of foam in the pipeline.

[0176] More specifically, the simulation module is also used for:

[0177] Set thermocouples and radiometers in the UHV substation and the simulated UHV substation to monitor the temperature distribution and the change of radiant heat flux in the combustion area. Select the fuel pool surface, the flame core area and the fire extinguishing agent injection area, and set temperature and radiant heat flux thresholds respectively. If the temperature in the flame area is lower than the set threshold, or the radiant heat flux at the measuring point decreases to the environmental background level, or the change rate of the radiant heat flux tends to zero, it is considered that the flame has extinguished, thus simulating the process of fire extinguishment.

[0178] More specifically, the formula of the air age model is as follows:

[0179]

[0180] Where, represents the partial derivative, describing the change rate of τ air with respect to t air ; t airis the real-time during the fluid flow process; is the gradient operator; ρ is the fluid density; τ air is the air age, i.e., the transport time of the foam in the pipeline; u is the fluid velocity vector; Γ τ is the diffusion coefficient of the air age; S τ is the air age source term, i.e., the local air age growth rate.

[0181] More specifically, the nested physical model of the simulation further includes a flame thermal radiation intensity model, which is used to reflect the fire energy release rate; the formula of the flame thermal radiation intensity model is as follows:

[0182]

[0183] Among them, represents the total radiant energy released by the flame, represents the external radiation per unit area, l and w are respectively the length and width of the rectangular fire source, N represents the number of point sources, ω j represents the weight of the j-th point source, satisfying χ r represents the radiation fraction, represents the mass loss rate per unit area, x0 represents the vertical distance from the radiometer to the flame surface, S j represents the distance between the j-th point source and the radiometer, ΔH c represents the calorific value of combustion of the combustible substance, θ represents the flame tilt angle, H is the flame height, H t is the vertical height of the radiometer, H j represents the vertical height of the point source j, L v is the latent heat of evaporation of the liquid at room temperature, k represents the heat conduction constant, T F represents the flame temperature, T l is the liquid temperature, h is the convective heat transfer coefficient, σ is the Stefan-Boltzmann constant, Φ F is the configuration factor, ε F is the emissivity of the high-temperature gas and the flame, ε l is the emissivity of the liquid, c pl ,p l and T0 are respectively the heat capacity, density and initial temperature of the liquid, γ and α are both constants, g is the gravitational acceleration, ρ a and ρ1 are respectively the densities of air and gas products.

[0184] Specifically, the parameter analysis module is also used to: compare the simulation data with the test data, and for the cases exceeding the preset error, adjust the simulation boundary conditions or adjust the parameters of the simulated UHV substation so that the deviation between the test data and the simulation data is within the preset range.

[0185] Specifically, the fire extinguishing performance prediction module further includes constructing a flame extinction determination model. The flame extinction determination model is based on a convolutional neural network, inputs a binary flame image, and outputs a determination result of whether the flame has extinguished. The convolutional neural network includes an input layer, a convolutional layer, and a fully connected layer. The convolutional layer includes a 3×3 convolutional kernel, a ReLU activation function, and a max pooling layer. The input layer receives the binary flame image, and the output result of the input layer passes through the 3×3 convolutional kernel, the ReLU activation function, and the max pooling layer and then is output by the fully connected layer as the determination result of whether the flame has extinguished.

[0186] More specifically, the binary flame image is obtained by converting the original flame image into a grayscale image and then into a binary image, and this binary image is the binary flame image.

[0187] The above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit them. Although the present invention has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that they can still modify the technical solutions described in the foregoing embodiments, or perform equivalent replacements for some of the technical features; and these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for evaluating compressed air foam fire extinguishing under the influence of multiple environmental factors, characterized in that, Including: S1. Collect the environmental parameters in the fire scenarios of UHV substations to form a fire environmental parameter library; The environmental parameters include environmental pressure, environmental temperature and humidity, and environmental wind speed; S2. Build a 3D model of the UHV substation, reproduce the UHV substation at a scale of 1:1 to obtain a simulated UHV substation, and select a physically nested simulation model for simulating fire scenarios; S3. Based on the fire environmental parameter library, simulate the influence of single environmental parameters and multiple environmental parameters on the fire extinguishing performance indicators in the simulated UHV substation respectively, obtain the simulated data associated with the fire extinguishing performance indicators, and record the corresponding test data of the UHV substation; The fire extinguishing performance indicators include fire extinguishing time, foam injection distance, foam coverage area, and fire field temperature distribution; S4. Use the test data and simulated data to construct a data set, use a generative adversarial network to perform data augmentation on the data set to obtain virtual samples, and use the virtual samples and the data set together as a database; S5. Build a fire extinguishing performance prediction model. The input layer of the fire extinguishing performance prediction model is the environmental parameters, and the output layer is the predicted fire extinguishing performance indicators. Use the database to train the fire extinguishing performance prediction model to obtain a trained fire extinguishing performance prediction model, and collect the environmental parameters of the UHV substation in real time to predict the fire extinguishing performance indicators.

2. The compressed air foam fire extinguishing evaluation method under the influence of multiple environmental factors according to claim 1, wherein S2 Including: Build a 3D model of the UHV substation to achieve 1:1 reproduction of the actual test conditions. Set the corresponding environmental parameters according to the actual environmental data of the UHV substation. Determine the heat release rate of the fire source according to the test data, set the flow rate of foam injection, and ensure that the injection range covers the fire source area; The physically nested simulation model includes the VOF model, the EDM eddy dissipation model, and the air age model. The VOF model simulates the flow process of multiple immiscible fluids, the EDM eddy dissipation model simulates the turbulent dissipation process of combustion in the fire field, and the air age model characterizes the transmission efficiency of foam in the pipeline.

3. The method for evaluating the fire extinguishing of compressed air foam under the influence of multiple environmental factors according to claim 2, characterized in that S2 also includes: Set thermocouples and radiometers in the UHV substation and the simulated UHV substation to monitor the temperature distribution and the change of radiant heat flux in the combustion area. Select the fuel pool surface, the flame core area, and the fire extinguishing agent injection area, and set the temperature and radiant heat flux thresholds respectively. If the temperature in the flame area is lower than the set threshold, or the radiant heat flux at the measurement point decreases to the environmental background level, or the change rate of the radiant heat flux tends to zero, it is considered that the flame has extinguished, thus simulating the process of fire extinguishment.

4. A method for evaluating the fire extinguishing of compressed air foam under the influence of multiple environmental factors according to claim 2, characterized in that, The formula of the air age model is as follows: Among them, represents the partial derivative and describes τ air changing with respect to t air ; t air is the real-time time during the fluid flow process; is the gradient operator; ρ is the fluid density; τ air is the air age, that is, the transport time of the foam in the pipeline; u is the fluid velocity vector; Γ τ is the diffusion coefficient of the air age; S τ is the air age source term, that is, the local air age growth rate.

5. The compressed air foam fire extinguishing evaluation method under the influence of multiple environmental factors according to claim 2, characterized in that, The physically nested simulation model also includes a flame thermal radiation intensity model, which is used to reflect the fire energy release rate; The formula of the flame thermal radiation intensity model is as follows: Among them, represents the total radiant energy released by the flame, represents the external radiation per unit area, l and w are the length and width of the rectangular fire source respectively, N represents the number of point sources, and ω j represents the weight of the j-th point source, satisfying χ r represents the radiation fraction, represents the mass loss rate per unit area, x0 represents the vertical distance from the radiometer to the flame surface, S j represents the distance between the j-th point source and the radiometer, ΔH c represents the calorific value of combustion of the combustible substance, θ represents the flame tilt angle, H t is the vertical height of the radiometer, H j represents the vertical height of the point source j, H is the flame height, L v is the latent heat of evaporation of the liquid at room temperature, k represents the heat conduction constant, T F represents the flame temperature, T l is the liquid temperature, h is the convective heat transfer coefficient, σ is the Stefan-Boltzmann constant, Φ F is the configuration factor, ε F is the emissivity of the high-temperature gas and the flame, ε l is the emissivity of the liquid, c pl ,p l and T0 are the heat capacity, density and initial temperature of the liquid respectively, γ and α are both constants, g is the gravitational acceleration, ρ a and ρ1 are the densities of air and gas products respectively.

6. The method for evaluating the compressed air foam fire extinguishing under the influence of multiple environmental factors according to claim 1, wherein S3 also includes: Compare the simulated data with the test data. For cases exceeding the preset error, adjust the simulation boundary conditions or adjust the parameters of the simulated UHV substation to make the deviation between the test data and the simulated data within the preset range.

7. A method for evaluating the fire extinguishing of compressed air foam under the influence of multiple environmental factors according to claim 1, characterized in that, S5 further includes constructing a flame extinction determination model. The flame extinction determination model is based on a convolutional neural network. The binary flame image is input, and the determination result of whether the flame is extinguished is output. The convolutional neural network includes an input layer, a convolutional layer, and a fully connected layer. The convolutional layer includes a 3×3 convolutional kernel, a ReLU activation function, and a max pooling layer. The input layer receives the binary flame image. After the output result of the input layer passes through the 3×3 convolutional kernel, the ReLU activation function, and the max pooling layer, the determination result of whether the flame is extinguished is output through the fully connected layer.

8. A method for evaluating the fire extinguishing performance of compressed air foam under the influence of multiple environmental factors according to claim 7, characterized in that, The binary flame image is obtained by converting the original flame image into a grayscale image and then into a binary image, and this binary image is the binary flame image.

9. A compressed air foam fire extinguishing evaluation system under the influence of multiple environmental factors, characterized in that, It includes: A data acquisition module for collecting environmental parameters in the fire scene of the UHV substation to form a fire environmental parameter library; The environmental parameters include environmental pressure, environmental temperature and humidity, and environmental wind speed; A simulation module for constructing a three-dimensional model of the UHV substation, replicating the UHV substation at a scale of 1:1 to obtain a simulated UHV substation, and selecting a physical model for simulation nesting to simulate the fire scene; A parameter analysis module for, based on the fire environmental parameter library, simulating the influence of single environmental parameters and multiple environmental parameters on the fire extinguishing performance indicators in the simulated UHV substation respectively, obtaining simulation data associated with the fire extinguishing performance indicators and recording the test data corresponding to the UHV substation; the fire extinguishing performance indicators include fire extinguishing time, foam spraying distance, foam coverage, and fire field temperature distribution; A training set construction module for using the test data and simulation data to construct a data set, using a generative adversarial network to perform data augmentation on the data set to obtain virtual samples, and using the virtual samples and the data set together as a database; A fire extinguishing performance prediction module for constructing a fire extinguishing performance prediction model. The input layer of the fire extinguishing performance prediction model is the environmental parameters, and the output layer is the predicted fire extinguishing performance indicators. Using the database to train the fire extinguishing performance prediction model to obtain a trained fire extinguishing performance prediction model, and collecting the environmental parameters of the UHV substation in real time to predict the fire extinguishing performance indicators.

10. The compressed air foam fire extinguishing evaluation system under the influence of multiple environmental factors according to claim 9, characterized in that, The simulation module is further used for: Constructing a three-dimensional model of the UHV substation to achieve a 1:1 replication of the actual test conditions, setting corresponding environmental parameters according to the actual environmental data of the UHV substation, determining the heat release rate of the fire source according to the test data, setting the flow rate of the foam spraying, and ensuring that the spraying range covers the fire source area; the physical models for simulation nesting include the VOF model, the EDM eddy dissipation model, and the air age model. The VOF model simulates the flow process of multiple immiscible fluids, the EDM eddy dissipation model simulates the turbulent dissipation process of combustion in the fire field, and the air age model characterizes the transmission efficiency of the foam in the pipeline.

Images