Extreme environment fire spread rate prediction method, device, equipment and medium

Abstract

The application provides an extreme environment fire spread rate prediction method, device, equipment and medium, which can be applied to the technical field of forest fire prevention and prediction, and the method comprises the following steps: using a pre-constructed combustion experiment platform to perform combustion experiments under different environmental conditions, and obtaining environmental parameters, fuel parameters and actual measured values of fire spread rates under different environmental conditions; a fire spread rate prediction model is constructed, and the prediction model is expressed as a benchmark fire spread rate multiplied by multiple correction terms; the environmental parameters, the fuel parameters and the actual measured values of the fire spread rates obtained through the combustion experiments are used to fit the benchmark fire spread rate in the prediction model and the to-be-fitted parameters contained in each correction term, so that a fitted prediction model is obtained; and the environmental parameters and the fuel parameters of a target region are input into the fitted prediction model, so as to determine a fire spread rate prediction value of the target region.

Description

Technical Field

[0001] This application relates to the field of forest fire prevention and prediction technology, and more specifically to a method, device, equipment, and medium for predicting the fire spread rate in extreme environments. Background Technology

[0002] Global warming has led to more frequent extreme heat and drought events, significantly increasing the intensity and frequency of forest fires. Currently, widely used international forest fire behavior prediction models include the Rothermel model from the United States, the Canadian Forest Fire Hazard Assessment System, and the McArthur Index from Australia. These models are essentially statistical relationships established through regression analysis of experimental data under historical, normal environmental conditions, possessing high computational efficiency and being widely adopted by mainstream operational software such as WRF-SFIRE and FlamMap. However, the parameters of these models are derived from regression analysis of experimental data under historical, normal temperature and humidity conditions, failing to account for changes under extreme heat conditions, resulting in significant prediction errors when directly applied to extreme climate scenarios. Summary of the Invention

[0003] In view of the above problems, this application provides a method, apparatus, device, medium and program product for predicting fire spread rate in extreme environments.

[0004] According to the first aspect of this application, a method for predicting fire spread rate in extreme environments is provided, comprising:

[0005] Using a pre-constructed combustion test platform, combustion experiments were conducted under different environmental conditions to obtain measured values ​​of environmental parameters, fuel parameters, and fire spread rate under these different environmental conditions.

[0006] A fire spread rate prediction model is constructed, which is represented by a baseline fire spread rate multiplied by multiple correction terms. The multiple correction terms include an atmospheric dryness correction term based on saturated water vapor pressure difference, a fuel moisture content correction term based on fuel moisture content, a volume density correction term based on fuel volume density, and an edge effect correction term based on fuel bed width-to-thickness ratio.

[0007] Using the environmental parameters, fuel parameters, and measured fire spread rate obtained from the combustion experiment, the baseline fire spread rate in the prediction model and the parameters to be fitted in each correction term are fitted to obtain the fitted prediction model.

[0008] The environmental and fuel parameters of the target area are input into the fitted prediction model to determine the predicted fire spread rate of the target area.

[0009] According to embodiments of this application, it also includes:

[0010] Obtain the ambient temperature and relative humidity under the current environmental conditions;

[0011] The saturated vapor pressure difference is calculated based on the ambient temperature and relative humidity, and the saturated vapor pressure difference is used to characterize the coupling effect of ambient temperature and relative humidity on the fire spread rate.

[0012] The first ratio is obtained by dividing the saturated water vapor pressure difference by the reference saturated water vapor pressure difference under the reference conditions.

[0013] Using the saturated water vapor pressure difference power exponent as the exponent, the first ratio is exponentially calculated to obtain the atmospheric dryness correction term;

[0014] The saturated water vapor pressure difference power exponent is a parameter to be fitted, used to reflect the sensitivity of fire spread rate to atmospheric dryness.

[0015] According to embodiments of this application, it also includes:

[0016] The saturated vapor pressure was calculated using the Magnus-Tetens formula with the ambient temperature as input.

[0017] Multiply the saturated water vapor pressure by the ratio of relative humidity to 100 to obtain the actual water vapor pressure;

[0018] The saturated vapor pressure difference is obtained by subtracting the actual vapor pressure from the saturated vapor pressure.

[0019] According to embodiments of this application, it also includes:

[0020] Obtain the fuel water content under current environmental conditions;

[0021] Obtain a pre-set critical fuel moisture content, which represents the fuel moisture content threshold at which the fire spread rate theoretically drops to zero;

[0022] Subtract the ratio of the fuel moisture content to the critical fuel moisture content from the first ratio to obtain the second ratio.

[0023] Using the power exponent of fuel moisture content as the exponent, the second ratio is exponentially calculated to obtain the fuel moisture content correction term.

[0024] The power exponent of fuel moisture content is a parameter to be fitted, used to reflect the sensitivity of fire spread rate to changes in fuel moisture content.

[0025] According to embodiments of this application, it also includes:

[0026] Obtain the fuel bed surface density and fuel bed thickness under current environmental conditions;

[0027] Dividing the fuel bed surface density by the fuel bed thickness yields the fuel volume density, which is used to characterize the density of the fuel bed.

[0028] Divide the fuel volume density by the preset reference volume density to obtain the third ratio;

[0029] Using the volume density power exponent as the exponent, the third ratio is exponentially calculated to obtain the volume density correction term;

[0030] The volume density power exponent is a parameter to be fitted, used to reflect the sensitivity of the fire spread rate to changes in the density of the fuel bed.

[0031] According to embodiments of this application, it also includes:

[0032] Obtain the fuel bed width and fuel bed thickness under the current environmental conditions;

[0033] Divide the width of the fuel bed by the thickness of the fuel bed to obtain the width-to-thickness ratio of the fuel bed;

[0034] Obtain the edge effect coefficients to be fitted, where the edge effect coefficients are dimensionless constants;

[0035] Divide the fuel bed width-to-thickness ratio by the sum of the fuel bed width-to-thickness ratio and the edge effect coefficient to obtain the edge effect correction term;

[0036] The edge effect correction term satisfies the following physical limits: when the fuel bed width-to-thickness ratio approaches infinity, the edge effect correction term approaches one, indicating no edge heat loss; when the fuel bed width-to-thickness ratio approaches zero, the edge effect correction term approaches zero, indicating complete edge heat loss.

[0037] According to an embodiment of this application, the step of fitting the baseline fire spread rate and the parameters to be fitted in each correction term of the prediction model using the environmental parameters, fuel parameters, and measured fire spread rate values ​​obtained from the combustion experiment to obtain the fitted prediction model includes:

[0038] Multiple sets of experimental data were obtained from the combustion experiment. Each set of experimental data included environmental parameters, fuel parameters, and the corresponding measured value of fire spread rate.

[0039] The objective function is obtained by summing the squared residuals between the predicted fire spread rate calculated by the prediction model for each set of experimental data and the corresponding measured fire spread rate.

[0040] The objective function is minimized using the nonlinear least squares method to obtain the fitted prediction model.

[0041] A second aspect of this application provides a device for predicting the fire spread rate in extreme environments, comprising:

[0042] The acquisition module is used to conduct combustion experiments under different environmental conditions using a pre-constructed combustion experiment platform, and to acquire measured values ​​of environmental parameters, fuel parameters and fire spread rate under the different environmental conditions.

[0043] The construction module is used to construct a fire spread rate prediction model. The prediction model is represented as a baseline fire spread rate multiplied by multiple correction terms. The multiple correction terms include an atmospheric dryness correction term based on saturated water vapor pressure difference, a fuel moisture content correction term based on fuel moisture content, a volume density correction term based on fuel volume density, and an edge effect correction term based on fuel bed width-to-thickness ratio.

[0044] The module is used to fit the baseline fire spread rate and the parameters to be fitted in each correction term of the prediction model using the environmental parameters, fuel parameters and measured values ​​of fire spread rate obtained from the combustion experiment, so as to obtain the fitted prediction model.

[0045] The input module is used to input the environmental parameters and fuel parameters of the target area into the fitted prediction model to determine the predicted value of the fire spread rate of the target area.

[0046] A third aspect of this application provides an electronic device comprising: one or more processors; and a memory for storing one or more computer programs, wherein the one or more processors execute the one or more computer programs to implement the steps of the method described above.

[0047] A fourth aspect of this application also provides a computer-readable storage medium having a computer program or instructions stored thereon, which, when executed by a processor, implement the steps of the above-described method.

[0048] The fifth aspect of this application also provides a computer program product, including a computer program or instructions that, when executed by a processor, implement the steps of the above-described method.

[0049] According to the embodiments of this application, by introducing an atmospheric dryness correction term based on saturated water vapor pressure difference, the coupling effect of ambient temperature and relative humidity on fire spread rate is comprehensively characterized. At the same time, by introducing an edge effect correction term based on fuel bed width-to-thickness ratio, the attenuation effect of heat loss at the edge of a finite-width fuel bed on fire spread rate is quantitatively described, thereby achieving effective extrapolation from laboratory scale to field scale. Experimental data show that the prediction model constructed using this technical solution has an average absolute percentage error of only 7.09% under extreme environmental conditions, which is significantly better than existing models. Moreover, the model has fewer input parameters and higher computational efficiency, which can meet the rapid deployment requirements of operational forest fire prediction systems. Attached Figure Description

[0050] The above-mentioned contents, other objects, features and advantages of this application will become clearer from the following description of embodiments with reference to the accompanying drawings, in which:

[0051] Figure 1 The schematic diagram illustrates the structure and sensor layout of the environmentally controlled combustion experimental platform according to an embodiment of this application;

[0052] Figure 2 A flowchart illustrating an extreme environment fire spread rate prediction method according to an embodiment of this application is shown schematically.

[0053] Figure 3 The diagram illustrates a comparison between the theoretical limit and experimental fit of the edge effect coefficient according to an embodiment of this application.

[0054] Figure 4 The diagram illustrates the results of six sets of independent verification experiments according to embodiments of this application;

[0055] Figure 5 A schematic diagram illustrating the structure of an extreme environment fire spread rate prediction device according to an embodiment of this application is shown.

[0056] Figure 6 A block diagram schematically illustrates an electronic device suitable for implementing a method for predicting the fire spread rate in extreme environments, according to an embodiment of this application. Detailed Implementation

[0057] The embodiments of this application will now be described with reference to the accompanying drawings. However, it should be understood that these descriptions are exemplary only and are not intended to limit the scope of this application. In the following detailed description, numerous specific details are set forth to provide a thorough understanding of the embodiments of this application for ease of explanation. However, it will be apparent that one or more embodiments may be implemented without these specific details. Furthermore, descriptions of well-known structures and technologies are omitted in the following description to avoid unnecessarily obscuring the concepts of this application.

[0058] The terminology used herein is for the purpose of describing particular embodiments only and is not intended to limit the scope of this application. The terms “comprising,” “including,” etc., as used herein indicate the presence of the stated features, steps, operations, and / or components, but do not exclude the presence or addition of one or more other features, steps, operations, or components.

[0059] All terms used herein (including technical and scientific terms) have the meanings commonly understood by those skilled in the art, unless otherwise defined. It should be noted that the terms used herein are to be interpreted in a manner consistent with the context of this specification, and not in an idealized or overly rigid way.

[0060] When using expressions such as "at least one of A, B and C", they should generally be interpreted in accordance with the meaning that is commonly understood by those skilled in the art (e.g., "a system having at least one of A, B and C" should include, but is not limited to, a system having A alone, a system having B alone, a system having C alone, a system having A and B, a system having A and C, a system having B and C, and / or a system having A, B and C, etc.).

[0061] Figure 1 The schematic diagram illustrates the structure and sensor layout of the environmentally controlled combustion experimental platform according to an embodiment of this application.

[0062] like Figure 1 As shown, this embodiment provides an environmentally controllable combustion experimental platform for conducting forest combustible fuel bed combustion experiments under extreme temperature conditions, and obtaining measured data on fire spread rate under different environmental and fuel conditions.

[0063] The combustion chamber measures 6 meters × 4 meters × 4 meters and is constructed using insulated walls 1 to ensure a stable indoor temperature during the experiment. The combustion chamber is equipped with a temperature control system capable of precisely regulating the indoor temperature within a range of -30 degrees Celsius to 50 degrees Celsius, simulating environmental temperatures under various climatic conditions, from cold winters to extreme heat and drought. A push-rod controlled exhaust port 3 is installed at the top of the combustion chamber, connected to an exhaust fan 2. During the experiment, the exhaust port 3 remains closed to prevent heat loss and airflow disturbances from affecting the experimental results. After the experiment, the exhaust fan 2 is activated and the exhaust port 3 is opened to vent the combustion gases outdoors, ensuring the safety of the experimental personnel and a clean experimental environment.

[0064] The fuel bed 8 measures 1.2 meters by 1.2 meters and is square in shape, located in the center of the combustion chamber. The fuel bed 8 uses 5 mm thick refractory gypsum board as the support material. This material has a low thermal conductivity, which can reduce heat loss from the bottom of the fuel bed 8 to the support structure, ensuring that heat is mainly transferred forward and upward during the flame propagation process, meeting the thermal boundary condition requirements of laboratory-scale combustion experiments.

[0065] To facilitate the description of the spatial position during flame propagation, a three-dimensional coordinate system is established, with the origin 9 located at the starting point of the centerline of the fuel bed 8, i.e., near the ignition end. Figure 1 The origin of the coordinate system is 9 (0,0,0). The y-axis runs along the flame spread direction, the x-axis is perpendicular to the flame spread direction (i.e., the width of the fuel bed), and the z-axis points vertically upwards. In this coordinate system, the spatial position of any point on the fuel bed 8 can be uniquely determined by the coordinates (x, y, z).

[0066] Multiple thermocouple sensors 5, including TS1, TS2, and TS3, are arranged at 20-centimeter intervals along the centerline of the fuel bed 8 (i.e., the straight line x=0) to track the position of the flame front and calculate the fire spread rate. The first thermocouple sensor TS1 is located at coordinates (0, 10, 0), which is 10 centimeters away from the ignition end. When the flame front arrives at each thermocouple sensor 5 in sequence, the temperature measured by the sensor rises sharply. By recording the time of temperature rise at each sensor, the average propagation speed of the flame front in different intervals can be calculated, thus obtaining the measured value of the fire spread rate. Specifically, the fire spread rate is calculated by dividing the distance between two adjacent thermocouple sensors by the time difference between the arrival of the flame front at those two sensors.

[0067] Temperature and humidity sensors 6, including HS1 and HS2, are arranged around the fuel bed 8 to monitor the ambient temperature and relative humidity during the experiment. Specifically, the first humidity sensor HS1 is located at coordinates (65, 60, 10), and the second humidity sensor HS2 is located at coordinates (160, 60, 10). The ambient temperature is taken as the average of the temperature values ​​built into the two temperature and humidity sensors 6, and the ambient humidity is taken as the average of the measurements from the two temperature and humidity sensors 6. This multi-point measurement and averaging method can reduce the influence of local thermal disturbances during combustion on the measurement results and improve the accuracy of environmental parameter measurements.

[0068] An oxygen concentration sensor 7, or OS, is arranged around the fuel bed 8 to monitor changes in oxygen concentration during the experiment. Specifically, the oxygen concentration sensor OS is located at coordinates (65, 60, 10), in the same position as the first humidity sensor HS1. The oxygen concentration sensor 7 records the decrease in oxygen concentration during combustion, indirectly reflecting combustion intensity and efficiency, and providing data support for energy conservation verification.

[0069] A heat flux meter 4 is positioned above the end of the fuel bed 8 to measure the heat feedback during flame propagation. Specifically, the heat flux meter 4 is located at coordinates (0, 120, δ+1), where δ is the thickness of the fuel bed 8. The installation height of the heat flux meter 4 is 1 cm above the fuel bed 8, a position that effectively captures the total heat flux density, including radiative and convective heat flux, when the flame front reaches the end of the fuel bed 8. The measured total heat flux density can be used to calculate the normalized combustion efficiency, thereby verifying the rationality of the model's energy conservation and heat transfer mechanisms.

[0070] The experimental fuel used was Pinus massoniana needle from Yunnan Province, China, a typical representative of forest surface combustibles in southern China. The pine needles were dried to constant weight in an 80°C constant-temperature drying oven to remove initial moisture. After drying, the pine needles were humidified according to the moisture content required by the experimental design: the required mass of deionized water was evenly sprayed onto the dried pine needles, and they were sealed and left to stand for 48 hours, turning them several times during this period to ensure uniform moisture distribution. After humidification, samples were taken, weighed, and dried again to verify the deviation between the actual moisture content and the target moisture content, ensuring the accuracy of moisture content control.

[0071] Before the experiment, the combustion chamber temperature was first adjusted to the target temperature and maintained for at least 30 minutes to ensure that the fuel bed 8, the sensor, and the indoor air temperature all reached the target temperature uniformly. Then, the pine needle fuel, prepared according to the target moisture content, was evenly spread on the fuel bed 8. During the spreading process, it was ensured that the fuel thickness was uniform, the surface was flat, and the fuel bed surface density and thickness met the experimental design requirements. After spreading, the combustion chamber door was closed, and ignition began after the indoor temperature and humidity stabilized.

[0072] Ignition is achieved using an electric heating wire ignition method, with the ignition source located at the starting point of the centerline of fuel bed 8, i.e., the origin 9 (0,0,0). After ignition, the flame propagates from the ignition end along the positive y-axis (the direction of the centerline of fuel bed 8). Thermocouple sensors 5 record temperature data in real time, and the sampling frequency of the data acquisition system is set to 10 Hz. When the flame front reaches the position of each thermocouple sensor 5, the sensor temperature rises sharply, and the system automatically records the arrival time. The experiment continues until the flame propagates to the end of fuel bed 8 or the flame extinguishes itself.

[0073] After the experiment, start the exhaust fan 2 and open the exhaust port 3 to exhaust the flue gas from the combustion chamber. After the combustion chamber temperature drops to a safe range, clean the fuel bed 8 and prepare for the next experiment.

[0074] Data recorded for each experiment included: ambient temperature (degrees Celsius), relative humidity (percentage), fuel moisture content (percentage), fuel bed width (cm), fuel bed thickness (cm), fuel bed surface density (g / cm²), and the measured fire spread rate (cm / s) calculated from data from thermocouple sensor 5. All data were entered into the experimental database for subsequent model parameter fitting and model validation.

[0075] The aforementioned environmentally controlled combustion experimental platform can be used to obtain measured values ​​of fire spread rate under different combinations of environmental temperatures (0°C to 50°C), fuel moisture content (0% to 10%), fuel bed widths (20 cm to 100 cm), fuel bed thicknesses (1 cm to 3 cm), and fuel loads (0.04 g / cm² to 0.08 g / cm²), providing systematic experimental data support for the construction and parameter fitting of fire spread rate prediction models in extreme environments.

[0076] Figure 2 A flowchart illustrating an extreme environment fire spread rate prediction method according to an embodiment of this application is shown.

[0077] like Figure 2 As shown, the extreme environment fire spread rate prediction method of this embodiment includes operations S210 to S240.

[0078] In operating S210, combustion experiments were conducted under different environmental conditions using a pre-constructed combustion experimental platform to obtain measured values ​​of environmental parameters, fuel parameters, and fire spread rate under different environmental conditions.

[0079] According to an embodiment of this application, the combustion test platform includes an environmentally controllable combustion chamber equipped with a temperature control system capable of adjusting the chamber temperature within a predetermined range. In one specific embodiment, the combustion chamber has overall dimensions of 6 meters × 4 meters × 4 meters, is constructed with insulated walls, and has a temperature adjustment range of -30°C to 50°C. A fuel bed, measuring 1.2 meters × 1.2 meters, is placed at the center of the combustion chamber, supported by 5 mm thick refractory gypsum board. Multiple thermocouple sensors are arranged at 20 cm intervals along the centerline of the fuel bed to track the flame front position and calculate the fire spread rate. Temperature and humidity sensors and oxygen concentration sensors are arranged around the fuel bed to monitor environmental parameters during the experiment. A heat flux meter is placed above the end of the fuel bed to measure flame thermal feedback.

[0080] According to embodiments of this application, environmental conditions include a comprehensive combination of multiple factors such as ambient temperature, ambient humidity, fuel moisture content, fuel bed width, fuel bed thickness, and fuel load. Environmental parameters include ambient temperature and relative humidity, while fuel parameters include fuel moisture content, fuel bed surface density, fuel bed thickness, and fuel bed width. The measured fire spread rate refers to the flame propagation speed calculated from the time difference of the flame front reaching different locations recorded by thermocouple sensors.

[0081] According to an embodiment of this application, forest combustibles are used as experimental fuel in combustion experiments. First, the fuel is dried to constant weight in an 80°C constant-temperature drying oven. Then, the fuel is adjusted to the target moisture content according to different experimental conditions. Independent combustion experiments are conducted under different combinations of ambient temperature, fuel moisture content, fuel bed width, fuel bed thickness, and fuel load. The measured fire spread rate is recorded for each group of experiments. In one specific embodiment, the experimental range for ambient temperature is 0°C to 50°C, the experimental range for fuel moisture content is 0% to 10%, the experimental range for fuel bed width is 20 cm to 100 cm, the experimental range for fuel bed thickness is 1 cm to 3 cm, and the experimental range for fuel bed surface density is 0.04 g / cm² to 0.08 g / cm².

[0082] In operation S220, a fire spread rate prediction model is constructed, which is represented as a baseline fire spread rate multiplied by multiple correction terms.

[0083] According to embodiments of this application, the prediction model employs a product structure, meaning the fire spread rate equals the product of the baseline fire spread rate and multiple correction terms. The baseline fire spread rate refers to the fire spread rate measured under preset baseline environmental and fuel conditions, serving as the benchmark reference value for model calculation. The multiple correction terms include an atmospheric dryness correction term based on saturated vapor pressure difference, a fuel moisture content correction term based on fuel moisture content, a bulk density correction term based on fuel bulk density, and an edge effect correction term based on fuel bed width-to-thickness ratio. Each correction term quantitatively describes the influence of a specific factor on the fire spread rate. The values ​​of the correction terms are typically between 0 and 1; when a factor is under baseline conditions, the corresponding correction term has a value of 1.

[0084] According to embodiments of this application, the atmospheric dryness correction term is used to characterize the coupling effect of ambient temperature and relative humidity on the fire spread rate. Saturated vapor pressure difference (VPD) is a physical quantity that comprehensively reflects the influence of temperature and humidity on atmospheric dryness. Its physical meaning is the difference between saturated vapor pressure and actual vapor pressure; the larger the difference, the drier the air, and the greater the potential for absorbing moisture from fuel. Embodiments of this application can introduce saturated vapor pressure difference as the basis for constructing the atmospheric dryness correction term, thereby achieving a comprehensive characterization of the temperature-humidity coupling effect.

[0085] According to embodiments of this application, the fuel moisture content correction term is used to characterize the inhibitory effect of the moisture content in the fuel on the fire spread rate. Fuel moisture content refers to the percentage of the mass of water contained in the fuel to the mass of dry fuel. When the fuel moisture content is high, more heat is required to evaporate the moisture, thereby slowing down the pyrolysis and combustion process of the fuel and reducing the fire spread rate. When the fuel moisture content reaches a certain critical value, the fire spread rate theoretically drops to zero; this critical value is called the critical fuel moisture content.

[0086] According to embodiments of this application, the bulk density correction term is used to characterize the effect of fuel bed density on fire spread rate. Fuel bulk density is defined as the ratio of fuel bed surface density to fuel bed thickness, where fuel bed surface density refers to the mass of fuel per unit area of ​​the fuel bed, and fuel bed thickness refers to the vertical height of the fuel bed. Fuel bulk density reflects the compactness of the fuel bed packing; a higher bulk density indicates a denser fuel packing, more combustible material per unit volume, and thus affects the heat transfer efficiency and oxygen supply conditions for flame propagation.

[0087] According to embodiments of this application, the edge effect correction term is used to characterize the attenuation effect of edge heat loss on the fire spread rate of a finite-width fuel bed. In laboratory-scale fuel bed combustion experiments, the width of the fuel bed is finite, and the flame loses heat to both sides during propagation, resulting in a fire spread rate lower than the ideal case for an infinitely wide fuel bed. The fuel bed width-to-thickness ratio is defined as the ratio of the fuel bed width to the fuel bed thickness. The larger this ratio, the wider the fuel bed, the smaller the proportion of edge heat loss to total heat loss, and the closer the value of the edge effect correction term is to 1.

[0088] In operation S230, the environmental parameters, fuel parameters, and measured values ​​of fire spread rate obtained from combustion experiments are used to fit the baseline fire spread rate and the parameters to be fitted in each correction term in the prediction model, thereby obtaining the fitted prediction model.

[0089] According to embodiments of this application, the prediction model includes multiple parameters to be fitted, including the baseline fire spread rate, the saturated vapor pressure difference power exponent, the fuel water content power exponent, the bulk density power exponent, and the edge effect coefficient. The specific values ​​of these parameters cannot be directly obtained through theoretical derivation and need to be estimated using experimental data acquired in operation S110. Parameter fitting can be achieved by using an optimization algorithm to find a set of parameter values ​​that minimizes the difference between the predicted fire spread rate calculated by the prediction model and the experimentally measured values.

[0090] According to an embodiment of this application, the parameter fitting employs a nonlinear least squares method. Specifically, the difference between the predicted fire spread rate calculated by the prediction model for each set of experimental data and the corresponding measured fire spread rate is first used as the residual. Then, the squared residuals of all experimental data are summed to obtain the objective function. By minimizing this objective function, the optimal estimate of the parameters to be fitted is obtained. In a specific embodiment, the confidence region reflection algorithm is used to solve this nonlinear least squares problem. This algorithm determines the search direction by constructing a local quadratic approximation model at the current iteration point, sets the confidence region radius to constrain the iteration step size, and iteratively updates the parameter vector until the objective function converges.

[0091] According to an embodiment of this application, a predictive model with definite parameter values ​​is obtained after fitting. This model can calculate the corresponding predicted fire spread rate value based on the input environmental parameters and fuel parameters. In a specific embodiment, the model parameters obtained by fitting based on 48 sets of pine needle combustion experimental data are as follows: the baseline fire spread rate is 0.4901 cm / s, the saturated vapor pressure difference power exponent is 0.0789, the fuel moisture content power exponent is 0.9581, the bulk density power exponent is -0.3695, and the edge effect coefficient is 4.1486. ​​The model fit goodness R² is 0.6991, the root mean square error is 0.0464 cm / s, and the mean absolute percentage error is 7.09%.

[0092] In operation S240, the environmental and fuel parameters of the target area are input into the fitted prediction model to determine the predicted fire spread rate of the target area.

[0093] According to embodiments of this application, the target area refers to an actual forest fire scenario requiring fire spread rate prediction or a forest area to be assessed. For the target area, environmental and fuel parameters need to be obtained as inputs to the prediction model. Environmental parameters include ambient temperature and relative humidity, which can be obtained through meteorological monitoring stations, weather forecasts, or satellite remote sensing. Fuel parameters include fuel moisture content, fuel bed surface density, fuel bed thickness, and fuel bed width, which can be obtained through field surveys, remote sensing inversion, or historical data statistics.

[0094] According to an embodiment of this application, the acquired environmental and fuel parameters are input into the fitted prediction model, which automatically calculates the predicted fire spread rate using a product-like expression. Specifically, the saturated vapor pressure difference is first calculated based on the ambient temperature and relative humidity. Then, the values ​​of the atmospheric dryness correction term, fuel moisture content correction term, bulk density correction term, and edge effect correction term are calculated respectively. Finally, the baseline fire spread rate is multiplied by the four correction terms to obtain the predicted fire spread rate. This predicted value is output in centimeters per second and is used to assess the fire spread risk in the target area and formulate corresponding prevention and control strategies.

[0095] According to embodiments of this application, in actual forest fire scenarios, when the fuel bed width is much greater than the fuel bed thickness, the fuel bed width-to-thickness ratio approaches infinity, and the value of the edge effect correction term approaches 1. In this case, the prediction model can be simplified to the product of the baseline fire spread rate and the atmospheric dryness correction term, fuel moisture content correction term, and bulk density correction term, eliminating the need to input the fuel bed width parameter. This simplified model is suitable for predicting large-area continuous forest surface fires, demonstrating the ability of the laboratory results of this application embodiment to extrapolate to actual field scenarios.

[0096] According to the embodiments of this application, by introducing an atmospheric dryness correction term based on saturated water vapor pressure difference, the coupling effect of ambient temperature and relative humidity on fire spread rate is comprehensively characterized. At the same time, by introducing an edge effect correction term based on fuel bed width-to-thickness ratio, the attenuation effect of heat loss at the edge of a finite-width fuel bed on fire spread rate is quantitatively described, thereby achieving effective extrapolation from laboratory scale to field scale. Experimental data show that the prediction model constructed using this technical solution has an average absolute percentage error of only 7.09% under extreme environmental conditions, which is significantly better than existing models. Moreover, the model has fewer input parameters and higher computational efficiency, which can meet the rapid deployment requirements of operational forest fire prediction systems.

[0097] According to the embodiments of this application, in order to further improve the applicability of the prediction model in real complex field environments, in addition to the above-mentioned four correction terms—atmospheric dryness correction term, fuel moisture content correction term, bulk density correction term, and edge effect correction term—the fire spread rate prediction model can also selectively introduce wind speed correction term and terrain slope correction term.

[0098] In one alternative embodiment, the wind speed correction term is represented as a nonlinear function of the baseline wind speed influence factor and the current wind speed, used to characterize the enhancing effect of horizontal airflow on the preheating efficiency of unburned fuel in front of the flame.

[0099] In another alternative embodiment, the terrain slope correction term is represented as a combination of the tangent function of the slope angle and the slope coefficient to be fitted, used to characterize the increased spread rate caused by the flame-attachment effect on sloping terrain.

[0100] The product model framework provided in this application has good scalability. The aforementioned wind speed correction term and slope correction term can be directly incorporated into the prediction model as additional multiplicative factors without changing the existing model fitting logic and parameter system.

[0101] According to an embodiment of this application, the construction of the atmospheric dryness correction term may include: obtaining the ambient temperature and relative humidity under the current environmental conditions; calculating the saturated water vapor pressure difference based on the ambient temperature and relative humidity; dividing the saturated water vapor pressure difference by the reference saturated water vapor pressure difference under the reference conditions to obtain a first ratio; and performing a power operation on the first ratio using the power exponent of the saturated water vapor pressure difference as the exponent to obtain the atmospheric dryness correction term.

[0102] In some embodiments, ambient temperature and relative humidity are two key environmental factors affecting the fire spread rate. Under high temperature and low humidity conditions, the rate of moisture evaporation from the fuel surface increases significantly, making it easier for the fuel to reach its pyrolysis temperature, thereby accelerating flame propagation. However, the effects of temperature and humidity on the fire spread rate are not independent but rather exhibit a significant coupling effect. For example, under the same relative humidity conditions, the higher the ambient temperature, the greater the saturated vapor pressure of the air, and the greater the difference between the actual vapor pressure and the saturated vapor pressure, indicating a stronger ability of the air to absorb moisture from the fuel. Neither temperature nor humidity alone can fully characterize this coupling effect. Therefore, this invention introduces the saturated vapor pressure difference as a physical quantity to comprehensively characterize the temperature-humidity coupling effect.

[0103] In some embodiments, the saturated vapor pressure difference refers to the difference between the saturated vapor pressure and the actual vapor pressure, and its physical meaning is a key indicator for measuring the dryness of the air. A larger saturated vapor pressure difference indicates drier air, a greater potential for absorbing moisture from the fuel, and is more conducive to flame propagation and spread. The saturated vapor pressure difference comprehensively considers the effects of both temperature and humidity, and can more accurately reflect the combined effect of the atmospheric environment on fire spread behavior.

[0104] In some embodiments, the reference saturated vapor pressure difference is a value calculated under preset reference environmental conditions. The reference conditions should be selected such that all correction terms are equal to 1 under these conditions, so that the reference fire spread rate directly corresponds to the measured fire spread rate under these conditions. In one specific embodiment, the reference saturated vapor pressure difference is 1.17 kPa.

[0105] According to an embodiment of this application, the reference saturated vapor pressure difference is a saturated vapor pressure difference value calculated under preset reference environmental conditions. The reference conditions should be selected such that the values ​​of all correction terms are 1 under these conditions, so that the reference fire spread rate can directly correspond to the measured fire spread rate under these conditions. In a specific embodiment, the reference saturated vapor pressure difference is taken as 1.17 kPa.

[0106] According to an embodiment of this application, the atmospheric dryness correction term is expressed by the following expression:

[0107]

[0108] in, This is a correction term for atmospheric dryness, where VPD is the saturated water vapor pressure difference under current environmental conditions. The reference saturated water vapor pressure difference under reference conditions. This represents the power exponent of the saturated vapor pressure difference. When the current saturated vapor pressure difference equals the baseline saturated vapor pressure difference, the ratio is 1, and the atmospheric dryness correction term is 1, indicating that the atmospheric dryness under current conditions is the same as the baseline conditions. When the current saturated vapor pressure difference is greater than the baseline saturated vapor pressure difference, the ratio is greater than 1, indicating that the current atmosphere is drier than the baseline conditions, and the expected fire spread rate is increased. When the current saturated vapor pressure difference is less than the baseline saturated vapor pressure difference, the ratio is less than 1, indicating that the current atmosphere is wetter than the baseline conditions, and the expected fire spread rate is decreased.

[0109] According to embodiments of this application, the aforementioned ratio is transformed using a power function, introducing the power exponent of the saturated vapor pressure difference as a parameter to be fitted. The power function form offers mathematical simplicity and flexibility, allowing the sensitivity of the fire spread rate to changes in the saturated vapor pressure difference to be reflected by adjusting the power exponent. When the power exponent is large, the fire spread rate is more sensitive to changes in the saturated vapor pressure difference; when the power exponent is small, the sensitivity is low. The specific value of this power exponent cannot be obtained through theoretical derivation and needs to be determined by parameter fitting using combustion experimental data. In one specific embodiment, the power exponent of the saturated vapor pressure difference obtained by fitting 48 sets of pine needle combustion experimental data is 0.0789, indicating that under these experimental conditions, the fire spread rate exhibits low sensitivity to changes in the saturated vapor pressure difference.

[0110] According to an embodiment of this application, the calculation of saturated water vapor pressure difference may include: using the Magnus-Tetens formula to calculate the saturated water vapor pressure with the ambient temperature as input; multiplying the saturated water vapor pressure by the ratio of relative humidity to 100 to obtain the actual water vapor pressure; and subtracting the actual water vapor pressure from the saturated water vapor pressure to obtain the saturated water vapor pressure difference.

[0111] The Magnus-Tetens formula is a widely used empirical formula for calculating saturated vapor pressure in meteorology. This formula establishes a functional relationship between ambient temperature and saturated vapor pressure. Saturated vapor pressure refers to the pressure of water vapor in the air at a given temperature when the air is saturated with water vapor; its value increases exponentially with increasing temperature. In other words, the higher the temperature, the greater the amount of water vapor the air can hold, and the greater the saturated vapor pressure. The specific form of the Magnus-Tetens formula is:

[0112]

[0113] in The saturated vapor pressure is T, measured in kilopascals (kPa), where T is the ambient temperature, measured in degrees Celsius.

[0114] According to embodiments of this application, actual water vapor pressure refers to the pressure generated by the actual water vapor present in the air under current environmental conditions. Actual water vapor pressure cannot be directly obtained from ambient temperature and relative humidity, but it can be calculated by multiplying saturated water vapor pressure by relative humidity. The specific expression is:

[0115]

[0116] Where ea is the actual water vapor pressure in kilopascals (kPa), and RH is the relative humidity in percentage. The physical basis of this calculation method is that relative humidity is defined as the percentage of actual water vapor pressure to saturated water vapor pressure; therefore, actual water vapor pressure equals saturated water vapor pressure multiplied by the percentage of relative humidity. For example, when the relative humidity is 50%, the actual water vapor pressure equals saturated water vapor pressure multiplied by 0.5.

[0117] According to an embodiment of this application, the saturated vapor pressure difference is defined as the difference between the saturated vapor pressure and the actual vapor pressure. The saturated vapor pressure difference is obtained by subtracting the actual vapor pressure from the saturated vapor pressure calculated in the above steps, and the specific expression is as follows:

[0118]

[0119] VPD stands for Vapor Pressure Difference, measured in kilopascals (kPa). This physical quantity reflects how far the current air moisture content is from saturation. When the actual vapor pressure equals the saturation vapor pressure, the vapor pressure difference is zero, indicating that the air is saturated and the relative humidity is 100%. When the actual vapor pressure is much lower than the saturation vapor pressure, the vapor pressure difference is large, indicating that the air is very dry and the relative humidity is low. Therefore, the vapor pressure difference is essentially a temperature-corrected expression of relative humidity, and it can more accurately reflect the impact of atmospheric dryness on fuel moisture evaporation and fire spread.

[0120] According to the embodiments of this application, by using the saturated vapor pressure difference calculated above as the input parameter for constructing the atmospheric aridity correction term, a complete calculation link from conventional meteorological observation data (ambient temperature and relative humidity) to the atmospheric aridity correction term can be completed. This calculation link allows the embodiments of this application to directly utilize conventional data provided by meteorological stations or weather forecasts without the need for additional measurement equipment, facilitating its application in actual forest fire management operations.

[0121] According to an embodiment of this application, the construction of the fuel moisture content correction term includes: obtaining the fuel moisture content under the current environmental conditions; obtaining a pre-set critical fuel moisture content; subtracting the ratio of the fuel moisture content to the critical fuel moisture content from a first value to obtain a second ratio; and performing a power operation on the second ratio using the power exponent of the fuel moisture content as the exponent to obtain the fuel moisture content correction term.

[0122] According to embodiments of this application, fuel moisture content refers to the percentage of water mass in the fuel relative to the dry fuel mass, and is one of the most critical fuel parameters affecting fire spread rate. When fuel contains a large amount of moisture, additional heat is required during flame propagation to evaporate this moisture, thereby slowing down the heating, pyrolysis, and ignition processes of the fuel and reducing the fire spread rate. As fuel moisture content increases, the fire spread rate exhibits a non-linear decreasing trend.

[0123] According to embodiments of this application, the critical fuel moisture content refers to the fuel moisture content at which the fire spread rate theoretically drops to zero. When the fuel moisture content reaches or exceeds this critical value, the flame cannot propagate continuously, and the fire spread rate is zero. The value of the critical fuel moisture content is related to factors such as fuel type and environmental conditions. In one specific embodiment, based on experimental observations of pine needle fuel, the critical fuel moisture content is set at 35%. This value indicates that when the moisture content of pine needle fuel reaches 35%, even under favorable conditions, the flame cannot propagate continuously in the fuel bed.

[0124] According to an embodiment of this application, the fuel moisture content correction term is expressed by the following expression:

[0125]

[0126] in, This is a correction term for fuel moisture content. Fuel moisture content, expressed as a percentage. The critical fuel moisture content, This represents the power exponent of fuel moisture content. When the fuel moisture content is zero, the value in parentheses is 1, and the fuel moisture content correction term is 1, indicating that completely dry fuel has no inhibitory effect on the fire spread rate. As the fuel moisture content gradually increases and approaches the critical fuel moisture content, the value in parentheses approaches 0, and the fuel moisture content correction term approaches 0, indicating that the fire spread rate approaches zero. This expression correctly satisfies the physical limit condition.

[0127] According to embodiments of this application, a power function is used to introduce the fuel moisture content power exponent as a parameter to be fitted. This power exponent reflects the sensitivity of the fire spread rate to changes in fuel moisture content. When the power exponent is close to 1, the fire spread rate decreases approximately linearly with increasing fuel moisture content; when the power exponent is greater than 1, the fire spread rate decreases more slowly in the low moisture content range and more rapidly in the high moisture content range; when the power exponent is less than 1, the opposite is true. In one specific embodiment, the fuel moisture content power exponent obtained by fitting 48 sets of pine needle combustion experimental data is 0.9581, which is very close to 1, indicating that under these experimental conditions, the fire spread rate decreases approximately linearly with increasing fuel moisture content.

[0128] According to embodiments of this application, fuel moisture content is obtained through laboratory measurement and field estimation. Under laboratory conditions, fuel moisture content is calculated by weighing fuel samples after drying them to constant weight in an 80°C constant-temperature drying oven. In practical field applications, fuel moisture content can be measured on-site using a handheld moisture meter; alternatively, estimation models based on time-delay-equilibrium moisture content theory can be used, exemplarily including the Fine Combustion Humidity Code model in the Canadian Fire Weather Index system, which uses daily or hourly ambient temperature, relative humidity, wind speed, and precipitation as meteorological input parameters; or the Nelson model, which is based on the physical processes of moisture and heat transfer and is suitable for real-time estimation using high-frequency observation data from automated weather stations.

[0129] According to an embodiment of this application, the construction of the volume density correction term includes: obtaining the fuel bed surface density and fuel bed thickness under the current environmental conditions; dividing the fuel bed surface density by the fuel bed thickness to obtain the fuel volume density; obtaining a pre-set reference volume density; dividing the fuel volume density by the reference volume density to obtain a third ratio; and using the volume density power exponent as the exponent, performing a power operation on the third ratio to obtain the volume density correction term.

[0130] According to embodiments of this application, fuel bulk density is an important parameter characterizing the compactness of the fuel bed, defined as the ratio of fuel bed surface density to fuel bed thickness. Fuel bed surface density refers to the mass of fuel per unit area of ​​the fuel bed, typically expressed in grams per square centimeter; fuel bed thickness refers to the vertical height of the fuel bed, typically expressed in centimeters. Fuel bulk density reflects the degree of packing of fuel particles in the fuel bed; a higher bulk density indicates a denser fuel packing and a greater amount of combustible material per unit volume.

[0131] According to embodiments of this application, the effect of fuel bulk density on fire spread rate exhibits a non-monotonic characteristic. On the one hand, higher bulk density means that more combustible material is contained per unit volume, which can release more heat and is conducive to flame propagation; on the other hand, excessively high bulk density will reduce the porosity of the fuel bed, reduce the supply of oxygen to the combustion zone, and increase the thermal resistance to heat transfer into the fuel bed, which may inhibit flame propagation. Therefore, the power exponent of the bulk density correction term is usually negative, indicating that within a certain range, increasing the bulk density will actually lead to a decrease in the fire spread rate.

[0132] According to an embodiment of this application, the volume density correction term is expressed by the following expression:

[0133]

[0134] in For volume density correction term, This refers to the fuel's bulk density, expressed in grams per cubic centimeter. As the reference volume density, This is the power exponent of the bulk density. Fuel bulk density. , which is the ratio of surface density ω to thickness δ.

[0135] According to an embodiment of this application, the reference bulk density is the bulk density value corresponding to a preset reference fuel condition, and is taken as 0.020 grams per cubic centimeter. When the fuel bulk density equals the reference bulk density, the bulk density correction term is 1, indicating that the density of the current fuel bed is the same as the reference condition. When the fuel bulk density is greater than the reference bulk density, since the bulk density power exponent is usually negative, the bulk density correction term is less than 1, indicating that the fire spread rate is lower than the reference condition. In one specific embodiment, the bulk density power exponent obtained by fitting 48 sets of pine needle combustion experimental data is -0.3695, confirming the trend that an increase in bulk density leads to a decrease in the fire spread rate.

[0136] According to embodiments of this application, the fuel bed surface density and fuel bed thickness are obtained through experimental setup and field measurement. Under laboratory conditions, the fuel bed surface density is determined by weighing the fuel mass laid on a fuel bed of known area, and the fuel bed thickness is measured using a thickness gauge placed at the edge of the fuel bed. In practical field applications, the fuel bed surface density can be measured using the quadrat harvesting method, i.e., collecting and weighing combustible material per unit area of ​​the ground surface; the fuel bed thickness can be obtained by averaging multiple measurements.

[0137] Figure 3 The diagram illustrates a comparison between the theoretical limit and experimental fit of the edge effect coefficient according to an embodiment of this application.

[0138] like Figure 3 As shown, this embodiment performs parameter fitting and theoretical limit analysis on the edge effect coefficient in the edge effect correction term. Figure 3 The horizontal axis represents the fuel bed width-to-thickness ratio Λ, which is the ratio of the fuel bed width W to the fuel bed thickness δ, and is dimensionless; the vertical axis represents the normalized fire spread rate, which is the ratio of the fire spread rate of a finite-width fuel bed to the reference fire spread rate of an infinite-width fuel bed (or when the width-to-thickness ratio is sufficiently large), and is dimensionless.

[0139] According to an embodiment of this application, the construction of the edge effect correction term includes: obtaining the fuel bed width and fuel bed thickness under the current environmental conditions; dividing the fuel bed width by the fuel bed thickness to obtain the fuel bed width-to-thickness ratio; obtaining the edge effect coefficient to be fitted; and dividing the fuel bed width-to-thickness ratio by the sum of the fuel bed width-to-thickness ratio and the edge effect coefficient to obtain the edge effect correction term.

[0140] According to embodiments of this application, the edge effect refers to the phenomenon in finite-width fuel bed combustion experiments where heat loss from the flame to the lateral edges leads to a decrease in the fire spread rate. In an infinitely wide fuel bed (i.e., a real forest fire scenario), flame propagation can be approximated as a two-dimensional process, with heat primarily transferred forward and downward, and no lateral heat loss. However, in a finite-width fuel bed at a laboratory scale, in addition to transferring heat forward and downward, the flame also loses heat to the lateral edges of the fuel bed, resulting in a reduction in the effective heat used to preheat the fuel in front, and a corresponding decrease in the fire spread rate. The narrower the fuel bed, the greater the proportion of edge heat loss to total heat loss, and the more significant the decrease in fire spread rate.

[0141] According to embodiments of this application, the fuel bed width-to-thickness ratio is defined as the ratio of the fuel bed width to the fuel bed thickness, denoted as . Where W is the fuel bed width and δ is the fuel bed thickness. The width-to-thickness ratio is a dimensionless parameter characterizing the intensity of the edge effect. When the width-to-thickness ratio is large, the fuel bed is relatively wide, and the influence of edge heat loss is relatively small; when the width-to-thickness ratio is small, the fuel bed is relatively narrow, and the influence of edge heat loss is significantly enhanced.

[0142] According to embodiments of this application, the edge effect correction term is expressed in the form of a rational function:

[0143]

[0144] in, This is the edge effect correction term, where Λ is the fuel bed width-to-thickness ratio. Let be the edge effect coefficient, and be the dimensionless parameter to be fitted. This rational function form has the following advantages: First, the expression is concise, containing only one parameter to be fitted; second, it correctly satisfies the physical limit condition; and third, it can fit experimental observation data well.

[0145] like Figure 3 As shown in the figure, the dashed line represents the theoretical limit curve, that is, when the width-to-thickness ratio of the fuel bed approaches infinity, i.e., when the width of the fuel bed is much greater than its thickness. ,but →1 indicates no edge heat loss, and the fire spread rate is not affected by edge effects; when the width-to-thickness ratio of the fuel bed approaches zero, i.e., when the fuel bed is extremely narrow, Λ→0, then →0 indicates complete edge heat loss, with the fire spread rate approaching zero.

[0146] According to embodiments of this application, the edge effect coefficient Empirical parameters combining various heat transfer mechanisms are related to factors such as the boundary convection heat transfer coefficient and the thermal characteristics of the fuel bed. The specific value of this coefficient cannot be obtained through theoretical derivation and needs to be determined by parameter fitting using combustion experimental data. For example... Figure 3As shown, each dot corresponds to a set of experimental results under a specific width-to-thickness ratio condition. The curve represents the edge effect correction term curve obtained based on nonlinear least squares fitting. The edge effect coefficient obtained based on 48 sets of pine needle combustion experimental data is 4.1486. ​​Substituting the fitted value into the above expression, the specific form of the edge effect correction term is obtained as follows: .

[0147] from Figure 3 It can be seen that the fitted curve and the experimental data points have a good agreement. In the region with a small width-to-thickness ratio (Λ < 20), the experimental data points are densely distributed near the fitted curve, indicating that the model can accurately capture the significant attenuation effect of edge heat loss on the fire spread rate under narrow fuel bed conditions. As the width-to-thickness ratio increases, the experimental data points gradually approach 1, indicating that when the width-to-thickness ratio is large enough, the influence of the edge effect gradually weakens, and the fire spread rate approaches the baseline value of an infinitely wide fuel bed.

[0148] Figure 3 The experimental fitting results verify the effectiveness of the edge effect correction term constructed in this invention. First, the rational function form correctly satisfies the physical limit condition; second, the single-parameter ( The fitting form of (=4.1486) is concise and efficient; furthermore, the fitting curve and the experimental data maintain good agreement throughout the entire range of the width-to-thickness ratio, proving that the expression can accurately describe the influence of the edge effect on the fire spread rate.

[0149] According to the embodiments of this application, the prediction model can be applied to actual forest fire scenarios in the following way: determine whether the width of the fuel bed in the target area is much greater than the thickness of the fuel bed; when the determination result is yes, set the value of the edge effect correction term to one; simplify the fire spread rate prediction model to the product of the baseline fire spread rate multiplied by the atmospheric dryness correction term, the fuel moisture content correction term and the bulk density correction term, and no longer include the edge effect correction term.

[0150] According to embodiments of this application, in actual forest fire scenarios, the width of the forest surface fuel bed is usually much greater than its thickness. For example, in continuous forest surface fires, the flame front width can reach tens or even hundreds of meters, while the thickness of the surface fuel layer is usually only a few centimeters to tens of centimeters, resulting in a fuel bed width-to-thickness ratio that can reach hundreds or even thousands. In this case, according to the mathematical properties of the edge effect correction term, when the fuel bed width-to-thickness ratio approaches infinity, the edge effect correction term approaches 1.

[0151] According to an embodiment of this application, the simplified prediction model is represented by the following expression:

[0152]

[0153] Where R is the predicted fire spread rate. As a reference fire spread rate, This is a correction term for atmospheric dryness. This is a correction term for fuel moisture content. This is a volume density correction term. This simplified model no longer requires inputting the fuel bed width parameter, significantly reducing the difficulty of data acquisition in practical applications.

[0154] According to embodiments of this application, the simplified model embodies one of the core values ​​of this application: extrapolating the prediction model established under laboratory conditions with finite-width fuel beds to real-world forest fire scenarios with infinite width. Through the quantitative description of the edge effect correction term, a clear mathematical connection is established between the laboratory scale and the field scale, solving the technical challenge of applying existing models across scales. In practical applications, when the ratio of the fuel bed width to thickness in the target area is greater than a certain threshold (e.g., greater than 100), the simplified model can be used for prediction, at which point the difference between the edge effect correction term value and 1 can be ignored.

[0155] According to the embodiments of this application, the parameters to be fitted can be fitted in the following manner: multiple sets of experimental data obtained from combustion experiments are acquired; the sum of the squared residuals between the predicted fire spread rate calculated by the prediction model for each set of experimental data and the corresponding measured fire spread rate is accumulated to obtain the objective function; the objective function is minimized using the nonlinear least squares method to obtain the optimal estimate of the parameters to be fitted.

[0156] According to an embodiment of this application, the parameter to be fitted includes the baseline fire spread rate. saturated vapor pressure difference power exponent Fuel moisture content power index Body density power index and edge effect coefficient These parameters constitute the parameter vector. .

[0157] In some embodiments, substituting the correction factors into the model, the complete model expression is:

[0158]

[0159] For practical forest fire applications, when W / δ→∞ →1, the model simplifies to:

[0160]

[0161] Model parameters The objective function is determined using the nonlinear least squares method. The objective function is defined as the sum of squared residuals:

[0162]

[0163] Where n=48 is the sample size. Let be the measured value of the fire spread rate in the i-th experiment. This is the predicted fire spread rate calculated based on the current parameter vector θ. The goal of parameter fitting is to find a set of parameter vectors θ that minimizes the objective function S(θ).

[0164] A trust-region reflection algorithm is employed to solve constrained nonlinear least squares problems. This algorithm works by applying the trust-region reflection algorithm at the current iteration point. Constructing a local quadratic approximation model to determine the search direction:

[0165] min||J×Δθ+r||², satisfying ||Δθ||≤Δ_max

[0166] Where J is the Jacobian matrix and r is the residual vector. Let θ be the radius of the trust region. Iteratively update the parameter vector until the objective function S(θ) is minimized, thus obtaining the optimal parameter estimate.

[0167] According to an embodiment of this application, the model parameters obtained by fitting using the nonlinear least squares method are specifically: reference fire spread rate. =0.4901 cm / s, saturated water vapor pressure difference power exponent Fuel moisture content power index Body density power index and edge effect coefficient =4.1486.

[0168] According to the embodiments of this application, the model fitting accuracy was evaluated using the following indicators: the standard error of the baseline fire spread rate was ±0.0108 cm / s, the model fit R² = 0.6991, the root mean square error (RMSE) was 0.0464 cm / s, the mean absolute percentage error (MAPE) was 7.09%, and the maximum relative error was 29.7%. These indicators show that the model's predicted values ​​agree well with the experimental values, and the prediction accuracy meets the requirements for engineering applications. Furthermore, 5-fold cross-validation was used to evaluate the model's generalization ability, with an average goodness of fit of 0.535 ± 0.293, indicating that the model has a certain generalization ability and can be applied to predicting fire spread rates under different experimental conditions.

[0169] According to the embodiments of this application, although the fitting parameters in the above specific embodiments are obtained based on experimental data of Pinus massoniana, those skilled in the art will understand that the product structure (baseline fire spread rate multiplied by multiple correction terms) of the fire spread rate prediction model constructed by this invention has universality. When it is necessary to apply the technical solution of this application to a second combustible material type (e.g., but not limited to broadleaf tree deciduous leaves, shrub stems and branches, herbaceous plants, or mixed coniferous and broadleaf fuels), there is no need to change the model structure; only the following migration steps need to be performed: First, determine the physicochemical properties of the second combustible material type, including but not limited to the base bulk density and critical fuel moisture content; second, use the environmentally controlled combustion experimental platform to conduct a limited number of orthogonal combustion experiments for the second combustible material type to obtain the measured value of the fire spread rate; finally, use the nonlinear least squares method described in operation S230 of this application to refit the model parameter vector to obtain a fire spread rate prediction model specifically for the second combustible material type.

[0170] According to the embodiments of this application, it needs to be clarified that the prediction target of this model is mainly aimed at the quasi-steady-state propagation stage of surface fire spread. The product form of this model is derived based on the assumptions of energy conservation and quasi-steady state of the thermal boundary layer. For unsteady extreme fire behaviors dominated by extremely strong convective instability, topographically induced vortices, or large-scale turbulent structures (such as fire whirls, flying fires, and the fire head acceleration process of sudden changes in crown fire), this model can be used as the bottom-level velocity input for computational fluid dynamics simulations or high-order cellular automata models, but it may not be able to fully capture its dynamic acceleration characteristics when applied independently. In operational applications, those skilled in the art can judge based on the risk level and combine empirical acceleration factors to provide auxiliary early warning when the input parameters exceed conventional thresholds.

[0171] Figure 4 The diagram illustrates the results of six sets of independent verification experiments according to embodiments of this application.

[0172] like Figure 4 As shown, this embodiment uses 6 sets of independent experimental data to verify the prediction effect of the fire spread rate prediction model. Figure 3 The horizontal axis represents the measured fire spread rate in centimeters per second; the vertical axis represents the predicted fire spread rate in centimeters per second. Each dot in the figure represents the correspondence between the measured and predicted values ​​of a set of independent verification experiments. The dashed line represents the ideal prediction line, i.e., the measured value equals the predicted value (1:1 line).

[0173] According to the embodiments of this application, the data from the six independent verification experiments were not used in the model parameter fitting, thus objectively reflecting the model's predictive ability for unknown data. The verification experiments covered combinations of different ambient temperatures, fuel moisture content, fuel bed width, fuel bed thickness, and fuel load, maintaining consistency with the parameter range of the fitting experiments.

[0174] like Figure 4 As shown, the measured and predicted values ​​of the six verification experiments are all distributed near the ideal prediction line. Except for Experiment 6, where the relative error is slightly larger, the predicted and measured values ​​of the other five experiments are very close, with absolute relative errors all within 12%. Statistical analysis of the six independent verification experiment data yielded the following evaluation indicators: coefficient of determination R² = 0.6298, mean absolute percentage error (MAPE) = 13.90%, and mean relative error (MRE) of 7.94%. These indicators demonstrate that the fire spread rate prediction model of this invention has good predictive ability for independent verification experiment data, and the prediction accuracy meets the requirements for engineering applications.

[0175] According to embodiments of this application, the elasticity coefficient method (local sensitivity analysis) and Sobol (global sensitivity) analysis can be used to quantify the contribution of input parameter uncertainty to the accuracy of fire spread rate prediction. The 5-parameter model (VPD, , (W, δ) and 6-parameter models (T, RH, Uncertainty was quantified using (ω, δ, W). The range of input parameters for sensitivity analysis was defined based on the minimum and maximum values ​​of the experimental dataset. For the Sobol analysis, 2048 samples were generated using the Saltelli sampling method, and the 95% confidence interval of the total effect index was estimated using 100 bootstrap resampling iterations.

[0176] The embodiments of this application construct a model that couples saturated vapor pressure difference (VPD) and edge effect correction. The model fit is R²=0.6991, the mean absolute percentage error (MAPE) is 7.09%, and the independent validation is R²=0.6298. It can accurately predict the fire spread rate of pine needle fuel beds under extreme temperature and humidity conditions.

[0177] The embodiments of this application derive the expressions for each correction factor based on energy conservation and boundary layer heat transfer theory, including the saturated water vapor pressure difference index. =0.0789 reflects the influence of atmospheric dryness; fuel moisture content index =0.9581 is close to 1, indicating that the fire spread rate decreases approximately linearly with the water content, and each parameter has a clear physical meaning.

[0178] The theoretical expression for the edge effect correction factor is derived from the embodiments of this application. And based on 48 sets of experimental data, the following was obtained through fitting. =4.1486, which enables a quantitative description of the heat loss at the edge of a fuel bed with a finite width, and provides a theoretical basis for extrapolating laboratory results to actual forest fire scenarios.

[0179] The prediction model of this application embodiment is established based on systematic experimental data of ambient temperature of 0-50℃ and fuel moisture content of 0-10%, which can be applied to the prediction of forest fire spread rate under extreme climate scenarios, filling the gap in the prediction capability of existing models under extreme temperature and humidity conditions.

[0180] The prediction model in this embodiment has six input parameters, which are simple and quick to calculate, and can meet the rapid prediction needs of actual forest fire management, making it easy to deploy and apply in operational forest fire behavior prediction systems.

[0181] Based on the above-mentioned method for predicting fire spread rate in extreme environments, this application also provides a detection device for predicting fire spread rate in extreme environments. The following will be combined with... Figure 5 The device is described in detail.

[0182] Figure 5 A schematic block diagram of an extreme environment fire spread rate prediction device according to an embodiment of this application is shown.

[0183] like Figure 5 As shown, the extreme environment fire spread rate prediction device of this embodiment includes an acquisition module 510, a construction module 520, a utilization module 530, and an input module 540.

[0184] The acquisition module 510 is used to conduct combustion experiments under different environmental conditions using a pre-constructed combustion experiment platform, and to acquire measured values ​​of environmental parameters, fuel parameters, and fire spread rate under different environmental conditions. In one embodiment, the acquisition module 510 can be used to perform the operation S210 described above, which will not be repeated here.

[0185] The construction module 520 is used to construct a fire spread rate prediction model. The prediction model is represented as a baseline fire spread rate multiplied by multiple correction terms. These correction terms include an atmospheric dryness correction term based on saturated vapor pressure difference, a fuel moisture content correction term based on fuel moisture content, a bulk density correction term based on fuel bulk density, and an edge effect correction term based on fuel bed width-to-thickness ratio. In one embodiment, the construction module 520 can be used to perform the operation S220 described above, which will not be repeated here.

[0186] Module 530 is used to fit the baseline fire spread rate and the parameters to be fitted in each correction term of the prediction model using environmental parameters, fuel parameters, and measured fire spread rate values ​​obtained from combustion experiments, thereby obtaining the fitted prediction model. In one embodiment, module 530 can be used to perform the operation S230 described above, which will not be repeated here.

[0187] The input module 540 is used to input the environmental parameters and fuel parameters of the target area into the fitted prediction model to determine the predicted fire spread rate of the target area. In one embodiment, the input module 540 can be used to perform the operation S240 described above, which will not be repeated here.

[0188] According to embodiments of this application, any multiple modules among the acquisition module 510, construction module 520, utilization module 530, and input module 540 can be merged into one module, or any one of these modules can be split into multiple modules. Alternatively, at least part of the functionality of one or more of these modules can be combined with at least part of the functionality of other modules and implemented in one module. According to embodiments of this application, at least one of the acquisition module 510, construction module 520, utilization module 530, and input module 540 can be at least partially implemented as hardware circuitry, such as a field-programmable gate array (FPGA), a programmable logic array (PLA), a system-on-a-chip, a system-on-a-substrate, a system-on-package, an application-specific integrated circuit (ASIC), or implemented in hardware or firmware by any other reasonable means of integrating or packaging circuitry, or implemented in any one of software, hardware, and firmware methods, or in a suitable combination of any of these. Alternatively, at least one of the acquisition module 510, construction module 520, utilization module 530, and input module 540 can be at least partially implemented as a computer program module, which can perform corresponding functions when the computer program module is run.

[0189] Figure 6 A block diagram schematically illustrates an electronic device suitable for implementing a method for predicting the fire spread rate in extreme environments, according to an embodiment of this application.

[0190] like Figure 6 As shown, an electronic device according to an embodiment of this application includes a processor 601, which can perform various appropriate actions and processes according to a program stored in a read-only memory (ROM) 602 or a program loaded from a storage portion 608 into a random access memory (RAM) 603. The processor 601 may include, for example, a general-purpose microprocessor (e.g., a CPU), an instruction set processor and / or an associated chipset and / or a special-purpose microprocessor (e.g., an application-specific integrated circuit (ASIC)), etc. The processor 601 may also include onboard memory for caching purposes. The processor 601 may include a single processing unit or multiple processing units for performing different actions of the method flow according to an embodiment of this application.

[0191] RAM 603 stores various programs and data required for the operation of the electronic device. Processor 601, ROM 602, and RAM 603 are interconnected via bus 604. Processor 601 executes various operations of the method flow according to embodiments of this application by executing programs in ROM 602 and / or RAM 603. It should be noted that programs may also be stored in one or more memories other than ROM 602 and RAM 603. Processor 601 may also execute various operations of the method flow according to embodiments of this application by executing programs stored in one or more memories.

[0192] According to embodiments of this application, the electronic device may further include an input / output (I / O) interface 605, which is also connected to a bus 604. The electronic device may also include one or more of the following components connected to the input / output (I / O) interface 605: an input section 606 including a keyboard, mouse, etc.; an output section 607 including a cathode ray tube (CRT), liquid crystal display (LCD), etc., and a speaker, etc.; a storage section 608 including a hard disk, etc.; and a communication section 609 including a network interface card such as a LAN card, modem, etc. The communication section 609 performs communication processing via a network such as the Internet.

[0193] This application also provides a computer-readable storage medium, which may be included in the device / apparatus / system described in the above embodiments; or it may exist independently and not assembled into the device / apparatus / system. The computer-readable storage medium carries one or more programs, which, when executed, implement the method according to the embodiments of this application.

[0194] According to embodiments of this application, the computer-readable storage medium can be a non-volatile computer-readable storage medium, such as including but not limited to: portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination thereof. In this application, the computer-readable storage medium can be any tangible medium containing or storing a program that can be used by or in conjunction with an instruction execution system, apparatus, or device. For example, according to embodiments of this application, the computer-readable storage medium may include ROM 602 and / or RAM 603 and / or one or more memories other than ROM 602 and RAM 603 described above.

[0195] Embodiments of this application also include a computer program product comprising a computer program containing program code for performing the methods shown in the flowchart. When the computer program product is run on a computer system, the program code is used to cause the computer system to implement the methods provided in the embodiments of this application.

[0196] When the computer program is executed by the processor 601, it performs the functions defined in the system / apparatus of this application embodiment. According to the embodiments of this application, the systems, apparatuses, modules, units, etc., described above can be implemented by computer program modules.

[0197] In such an embodiment, when the computer program is executed by the processor 601, it performs the functions defined in the system of this application embodiment. According to the embodiments of this application, the systems, devices, apparatuses, modules, units, etc., described above can be implemented by computer program modules.

[0198] According to embodiments of this application, program code for executing the computer programs provided in the embodiments of this application can be written in any combination of one or more programming languages. Specifically, these computational programs can be implemented using high-level procedural and / or object-oriented programming languages, and / or assembly / machine languages. Programming languages ​​include, but are not limited to, languages ​​such as Java, C++, "C", or similar programming languages. The program code can be executed entirely on the user's computing device, partially on the user's device, partially on a remote computing device, or entirely on a remote computing device or server. In cases involving remote computing devices, the remote computing device can be connected to the user's computing device via any type of network, including a local area network (LAN) or a wide area network (WAN), or it can be connected to an external computing device (e.g., via the Internet using an Internet service provider).

[0199] The flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of systems, methods, and computer program products according to various embodiments of this application. In this regard, each block in a flowchart or block diagram may represent a module, segment, or portion of code containing one or more executable instructions for implementing a specified logical function. It should also be noted that in some alternative implementations, the functions indicated in the blocks may occur in a different order than those indicated in the drawings. For example, two consecutively indicated blocks may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved. It should also be noted that each block in a block diagram or flowchart, and combinations of blocks in a block diagram or flowchart, may be implemented using a dedicated hardware-based system that performs the specified function or operation, or using a combination of dedicated hardware and computer instructions.

[0200] Those skilled in the art will understand that the features described in the various embodiments of this application can be combined and / or combined in various ways, even if such combinations or combinations are not explicitly described in this application. In particular, the features described in the various embodiments of this application can be combined and / or combined in various ways without departing from the spirit and teachings of this application. All such combinations and / or combinations fall within the scope of this application.

Claims

1. A method for predicting fire spread rate in extreme environments, characterized in that, include: Using a pre-constructed combustion test platform, combustion experiments were conducted under different environmental conditions to obtain measured values ​​of environmental parameters, fuel parameters, and fire spread rate under these different environmental conditions. A fire spread rate prediction model is constructed, which is represented by a baseline fire spread rate multiplied by multiple correction terms. The multiple correction terms include an atmospheric dryness correction term based on saturated water vapor pressure difference, a fuel moisture content correction term based on fuel moisture content, a volume density correction term based on fuel volume density, and an edge effect correction term based on fuel bed width-to-thickness ratio. Using the environmental parameters, fuel parameters, and measured fire spread rate obtained from the combustion experiment, the baseline fire spread rate in the prediction model and the parameters to be fitted in each correction term are fitted to obtain the fitted prediction model. The environmental and fuel parameters of the target area are input into the fitted prediction model to determine the predicted fire spread rate of the target area. Also includes: Obtain the ambient temperature and relative humidity under the current environmental conditions; The saturated vapor pressure difference is calculated based on the ambient temperature and relative humidity, and the saturated vapor pressure difference is used to characterize the coupling effect of ambient temperature and relative humidity on the fire spread rate. The first ratio is obtained by dividing the saturated water vapor pressure difference by the reference saturated water vapor pressure difference under the reference conditions. Using the saturated water vapor pressure difference power exponent as the exponent, the first ratio is exponentially calculated to obtain the atmospheric dryness correction term; The saturated water vapor pressure difference power exponent is a parameter to be fitted, used to reflect the sensitivity of fire spread rate to atmospheric dryness.

2. The method according to claim 1, characterized in that, Also includes: The saturated vapor pressure was calculated using the Magnus-Tetens formula with the ambient temperature as input. Multiply the saturated water vapor pressure by the ratio of relative humidity to 100 to obtain the actual water vapor pressure; The saturated vapor pressure difference is obtained by subtracting the actual vapor pressure from the saturated vapor pressure.

3. The method according to claim 1, characterized in that, Also includes: Obtain the fuel water content under current environmental conditions; Obtain a pre-set critical fuel moisture content, which represents the fuel moisture content threshold at which the fire spread rate theoretically drops to zero; Subtract the ratio of the fuel moisture content to the critical fuel moisture content from the first ratio to obtain the second ratio. Using the power exponent of fuel moisture content as the exponent, the second ratio is exponentially calculated to obtain the fuel moisture content correction term. The power exponent of fuel moisture content is a parameter to be fitted, used to reflect the sensitivity of fire spread rate to changes in fuel moisture content.

4. The method according to claim 1, characterized in that, Also includes: Obtain the fuel bed surface density and fuel bed thickness under current environmental conditions; Dividing the fuel bed surface density by the fuel bed thickness yields the fuel volume density, which is used to characterize the density of the fuel bed. Divide the fuel volume density by the preset reference volume density to obtain the third ratio; Using the volume density power exponent as the exponent, the third ratio is exponentially calculated to obtain the volume density correction term; The volume density power exponent is a parameter to be fitted, used to reflect the sensitivity of the fire spread rate to changes in the density of the fuel bed.

5. The method according to claim 1, characterized in that, Also includes: Obtain the fuel bed width and fuel bed thickness under the current environmental conditions; Divide the width of the fuel bed by the thickness of the fuel bed to obtain the width-to-thickness ratio of the fuel bed; Obtain the edge effect coefficients to be fitted, where the edge effect coefficients are dimensionless constants; Divide the fuel bed width-to-thickness ratio by the sum of the fuel bed width-to-thickness ratio and the edge effect coefficient to obtain the edge effect correction term; The edge effect correction term satisfies the following physical limits: when the fuel bed width-to-thickness ratio approaches infinity, the edge effect correction term approaches one, indicating no edge heat loss; when the fuel bed width-to-thickness ratio approaches zero, the edge effect correction term approaches zero, indicating complete edge heat loss.

6. The method according to claim 1, characterized in that, The method of fitting the baseline fire spread rate and the parameters to be fitted in each correction term of the prediction model using the environmental parameters, fuel parameters, and measured fire spread rate values ​​obtained from the combustion experiment to obtain the fitted prediction model includes: Multiple sets of experimental data were obtained from the combustion experiment. Each set of experimental data included environmental parameters, fuel parameters, and the corresponding measured value of fire spread rate. The objective function is obtained by summing the squared residuals between the predicted fire spread rate calculated by the prediction model for each set of experimental data and the corresponding measured fire spread rate. The objective function is minimized using the nonlinear least squares method to obtain the fitted prediction model; Also includes: Obtain the ambient temperature and relative humidity under the current environmental conditions; The saturated vapor pressure difference is calculated based on the ambient temperature and relative humidity, and the saturated vapor pressure difference is used to characterize the coupling effect of ambient temperature and relative humidity on the fire spread rate. The first ratio is obtained by dividing the saturated water vapor pressure difference by the reference saturated water vapor pressure difference under the reference conditions. Using the saturated water vapor pressure difference power exponent as the exponent, the first ratio is exponentially calculated to obtain the atmospheric dryness correction term; The saturated water vapor pressure difference power exponent is a parameter to be fitted, used to reflect the sensitivity of fire spread rate to atmospheric dryness.

7. A device for predicting the fire spread rate in extreme environments, characterized in that, include: The acquisition module is used to conduct combustion experiments under different environmental conditions using a pre-constructed combustion experiment platform, and to acquire measured values ​​of environmental parameters, fuel parameters and fire spread rate under the different environmental conditions. The construction module is used to construct a fire spread rate prediction model. The prediction model is represented as a baseline fire spread rate multiplied by multiple correction terms. The multiple correction terms include an atmospheric dryness correction term based on saturated water vapor pressure difference, a fuel moisture content correction term based on fuel moisture content, a volume density correction term based on fuel volume density, and an edge effect correction term based on fuel bed width-to-thickness ratio. The module is used to fit the baseline fire spread rate and the parameters to be fitted in each correction term of the prediction model using the environmental parameters, fuel parameters and measured values ​​of fire spread rate obtained from the combustion experiment, so as to obtain the fitted prediction model. The input module is used to input the environmental parameters and fuel parameters of the target area into the fitted prediction model to determine the predicted value of the fire spread rate of the target area.

8. An electronic device, comprising: One or more processors; Memory, used to store one or more computer programs. The characteristic feature is that the one or more processors execute the one or more computer programs to implement the steps of the method according to any one of claims 1 to 6.

9. A computer-readable storage medium having a computer program or instructions stored thereon, characterized in that, When the computer program or instructions are executed by a processor, they implement the steps of the method according to any one of claims 1 to 6.

Images