Forest fire danger grade determination method and device and electronic equipment

Abstract

The invention discloses a forest fire danger grade determination method and device and electronic equipment, and relates to the technical field of forest fire early warning, and the method comprises the steps: inputting the meteorological data, vegetation data and topographic data of a target forest into a target model, and obtaining N fire spreading factors and a target label corresponding to the target forest, the fire behavior spreading factor is used for representing the fire behavior spreading speed after a fire occurs in the target forest, and the target label is used for representing the region type of the region where the target forest is located; and determining a target probability and a target grade corresponding to the target forest according to the N fire spreading factors and the target label, the target probability being used for representing the probability of the fire of the target forest, and the target grade being used for representing the scale of the fire of the target forest. According to the method and the device, the technical problem of low accuracy of predicting the probability and the grade of forest fire danger in the prior art is solved.

Description

Technical Field

[0001] This application relates to the field of forest fire early warning technology, and more specifically, to a method, apparatus, and electronic device for determining forest fire risk levels. Background Technology

[0002] With the increasing impact of global climate change and human activities, the frequency and intensity of forest fires are constantly rising, posing a serious threat to the ecological environment and human society. Forest fires not only destroy natural ecosystems but can also lead to a series of environmental problems such as soil erosion, biodiversity loss, and air quality deterioration. Furthermore, forest fires cause enormous losses to people's lives and property. Therefore, effective assessment and management of forest fire risks are crucial.

[0003] Traditional forest fire risk assessment methods typically rely on empirical judgment and qualitative analysis. These methods often focus on large-scale fire events. However, fire risk factors and spread characteristics often differ significantly across different regions and forest types. Existing technologies lack the ability to measure different local factors corresponding to different regions and forest types when predicting forest fire risk, resulting in low accuracy in predicting the probability and severity of forest fires.

[0004] There is currently no effective solution to the above problems. Summary of the Invention

[0005] This application provides a method, apparatus, and electronic device for determining forest fire risk levels, in order to at least solve the technical problem of low accuracy in predicting the probability and level of forest fire occurrence in the prior art.

[0006] According to one aspect of this application, a method for determining forest fire risk levels is provided, comprising: inputting meteorological data, vegetation data, and topographic data of a target forest into a target model to obtain N fire spread factors and a target label corresponding to the target forest, wherein the target model is a neural network model trained based on historical forest data and historical fire data of L forests, where L and N are both positive integers, the fire spread factors are used to characterize the speed of fire spread after a fire occurs in the target forest, and the target label is used to characterize the regional type of the area where the target forest is located; determining the target probability and target level corresponding to the target forest based on the N fire spread factors and the target label, wherein the target probability is used to characterize the probability of a fire occurring in the target forest, and the target level is used to characterize the scale of a fire occurring in the target forest.

[0007] Optionally, before inputting the meteorological, vegetation, and topographic data of the target forest into the target model, the method for determining the forest fire risk level includes: acquiring meteorological, vegetation, and topographic information of the target forest, wherein the meteorological information includes at least the temperature, precipitation, wind speed, and humidity of the target forest, the vegetation information includes at least the tree and biological information of the target forest, and the topographic information includes at least the regional information of the area where the target forest is located; normalizing and cleaning the meteorological, vegetation, and topographic information to obtain meteorological, vegetation, and topographic data, wherein the normalization process is used to unify the data format corresponding to the meteorological, vegetation, and topographic information, and the data cleaning process is used to process the missing information in the meteorological, vegetation, and topographic information according to the interpolation method.

[0008] Optionally, the N fire spread factors include at least meteorological factors, vegetation factors, and topographic factors. The meteorological factors include at least the daily maximum temperature, consecutive days without precipitation, daily maximum wind speed, and daily minimum relative humidity of the target forest. The vegetation factors include at least the forest type, forest canopy density, average tree height, biomass index, and temperature-vegetation-drought index of the target forest. The topographic factors include at least the slope, aspect, and altitude of the target forest.

[0009] Optionally, the target probability and target level of the target forest are determined based on N fire spread factors and target labels, including: using the sum of coefficients corresponding to the target forest's daily maximum temperature, consecutive days without precipitation, daily maximum wind speed, and daily minimum relative humidity as the first factor coefficient corresponding to the meteorological factor; using the sum of coefficients corresponding to the target forest's forest type, forest canopy density, average tree height, biomass index, and temperature-vegetation-drought index as the second factor coefficient corresponding to the vegetation factor; using the sum of coefficients corresponding to the target forest's slope, aspect, and altitude as the third factor coefficient corresponding to the topographic factor; determining the target probability based on the first factor coefficient, second factor coefficient, and third factor coefficient; and determining the target level based on the first factor coefficient, second factor coefficient, third factor coefficient, and target label.

[0010] Optionally, determining the target probability based on the first factor coefficient, the second factor coefficient, and the third factor coefficient includes: determining the preset weights corresponding to the first factor coefficient, the second factor coefficient, and the third factor coefficient based on the target model; and performing a weighted summation of the first factor coefficient, the second factor coefficient, and the third factor coefficient based on the preset weights to obtain the target probability.

[0011] Optionally, the target level is determined based on the first factor coefficient, the second factor coefficient, the third factor coefficient, and the target label, including: determining the fire risk probability index corresponding to the target forest based on the first factor coefficient, the second factor coefficient, and the third factor coefficient, wherein the fire risk probability index and the target probability are positively correlated; and determining the target level based on the fire risk probability index and the target label.

[0012] Optionally, the target model is trained through the following steps: A training set and a test set are generated based on historical forest data and historical fire data from L forests. The training set includes historical forest data and historical fire data from P forests, and the test set includes historical forest data and historical fire data from Q forests, where P and Q are both positive integers, and L equals the sum of P and Q. The initial neural network is iteratively trained using the historical forest data and historical fire data from the P forests in the training set to obtain an initial model. The model parameters in the initial model are iteratively updated based on the historical forest data and historical fire data from the Q forests in the test set. The model obtained from the last update is used as the target model.

[0013] According to another aspect of this application, a device for determining forest fire risk levels is also provided, comprising: an input unit for inputting meteorological data, vegetation data, and topographic data of a target forest into a target model to obtain N fire spread factors and a target label corresponding to the target forest, wherein the target model is a neural network model trained based on historical forest data and historical fire data of L forests, where L and N are both positive integers, the fire spread factors are used to characterize the speed of fire spread after a fire occurs in the target forest, and the target label is used to characterize the regional type of the area where the target forest is located; and a determination unit for determining the target probability and target level corresponding to the target forest based on the N fire spread factors and the target label, wherein the target probability is used to characterize the probability of a fire occurring in the target forest, and the target level is used to characterize the scale of a fire occurring in the target forest.

[0014] According to another aspect of this application, a computer program product is also provided, which stores a computer program, wherein, when the computer program is running, it controls the computer program product to execute any of the above-mentioned methods for determining the forest fire risk level.

[0015] According to another aspect of this application, an electronic device is also provided, wherein the electronic device includes one or more processors and a memory for storing one or more programs, wherein when the one or more programs are executed by the one or more processors, the one or more processors implement the method for determining the forest fire risk level described above.

[0016] In this application, meteorological, vegetation, and topographic data of the target forest are first collected. These data are then input into a target model. The target model is used to analyze and obtain N fire spread factors and target labels corresponding to the target forest. The target model is a neural network model trained based on historical forest data and historical fire data of L forests. The fire spread factors represent the speed at which a fire spreads after it occurs in the target forest, and the target labels represent the regional type of the area where the target forest is located. Then, based on the N fire spread factors and target labels, this application determines the target probability and target level corresponding to the target forest. The target probability is used to characterize the probability of a fire occurring in the target forest, and the target level is used to characterize the scale of a fire occurring in the target forest.

[0017] As can be seen from the above, this application analyzes the meteorological data, vegetation data, and topographic data of the target forest through a target model, thereby obtaining N fire spread factors and target labels corresponding to the target forest. This application uses target labels to represent the regional characteristics (e.g., regional type) of different areas where the target forest is located. Then, this application uses the fire spread rate of the target forest (i.e., fire spread factor) and the regional characteristics of different areas where the target forest is located as evaluation indicators for the probability and level of fire risk in the forest to be evaluated, thereby achieving the goal of improving the accuracy of predicting the probability and level of forest fire risk.

[0018] Therefore, this application achieves the goal of determining the probability and level of forest fire risk based on the fire spread factor and regional type of the target forest by comprehensively analyzing the meteorological data, vegetation data and topographic data of the target forest through the target model. This improves the accuracy of predicting the probability and level of forest fire risk and solves the technical problem of low accuracy in predicting the probability and level of forest fire risk in the prior art. Attached Figure Description

[0019] The accompanying drawings, which are included to provide a further understanding of this application and form part of this application, illustrate exemplary embodiments and are used to explain this application, but do not constitute an undue limitation of this application. In the drawings:

[0020] Figure 1 This is a flowchart of an optional method for determining forest fire risk level according to an embodiment of this application;

[0021] Figure 2 This is a flowchart of an optional forest fire risk assessment method according to an embodiment of this application;

[0022] Figure 3This is a comparison chart of the predicted results and actual results for different regions according to an embodiment of this application;

[0023] Figure 4 This is a schematic diagram of an optional forest fire risk level determination device according to an embodiment of this application;

[0024] Figure 5 This is a schematic diagram of an optional electronic device according to an embodiment of this application. Detailed Implementation

[0025] To enable those skilled in the art to better understand the present application, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present application, and not all embodiments. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative effort should fall within the scope of protection of the present application.

[0026] It should be noted that the terms "first," "second," etc., in the specification, claims, and accompanying drawings of this application are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of this application described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0027] It should also be noted that all relevant information (including login information of target users and relevant information about system resources) and data (including but not limited to data used for display and analysis) involved in this application are information and data authorized by the user or fully authorized by all parties. For example, if there is an interface between this system and relevant users or organizations, before obtaining relevant information, it is necessary to send an acquisition request to the aforementioned users or organizations through the interface, and obtain the relevant information only after receiving consent from the aforementioned users or organizations.

[0028] Furthermore, the collection, storage, use, processing, transmission, provision, disclosure, and application of relevant information and data involved in this application all comply with the relevant laws, regulations, and standards of the relevant regions, and necessary confidentiality measures have been taken. This application does not violate public order and good morals. In addition, this application provides a corresponding operation entry point for users to choose to agree to or refuse authorization. If the user chooses to refuse authorization, the corresponding expert decision-making process will be initiated.

[0029] According to an embodiment of this application, an embodiment of a method for determining forest fire risk level is provided. It should be noted that the steps shown in the flowchart in the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions. Also, although a logical order is shown in the flowchart, in some cases, the steps shown or described may be executed in a different order than that shown here.

[0030] This application provides a system for determining forest fire risk levels (hereinafter referred to as the determination system) for implementing the method for determining forest fire risk levels in this application. Figure 1 This is a flowchart of an optional method for determining forest fire risk levels according to an embodiment of this application, such as... Figure 1 As shown, the method includes the following steps:

[0031] Step S101: Input the meteorological data, vegetation data and topographic data of the target forest into the target model to obtain N fire spread factors and target labels corresponding to the target forest.

[0032] In step S101, the target model is a neural network model trained based on historical forest data and historical fire data of L forests, where L and N are both positive integers. The fire spread factor is used to characterize the speed of fire spread after a fire occurs in the target forest, and the target label is used to characterize the regional type of the area where the target forest is located.

[0033] Optionally, the target forest refers to a specific forest area or ecosystem for which a fire risk assessment is being conducted; meteorological data is an important environmental factor affecting the spread of fire, and the meteorological data should include at least the daily maximum temperature, consecutive days without precipitation, daily maximum wind speed, and daily minimum relative humidity; vegetation data directly affects the speed and manner of fire spread, and the vegetation data should include at least the type of target forest, forest canopy density, tree height, and biomass, and vegetation factors directly affect the speed and manner of fire spread; the topography's undulations and orientation affect the path and speed of fire propagation, and the topography data should include at least digital elevation (i.e., altitude), slope, and aspect.

[0034] In step S101, the target model is a neural network model trained using deep learning technology and based on a large amount of historical forest and fire data. Its purpose is to capture and learn the patterns and laws of forest fire occurrence, especially the dynamic characteristics related to fire spread. When the input data (i.e., pre-processed meteorological, vegetation, and topographic data) enters the target model, the model automatically analyzes and processes the data based on its learned knowledge, ultimately outputting N fire spread factors and a target label related to the target forest. Here, the "fire spread factor" is a quantitative description of the characteristics of forest fire spread, while the "target label" is used to identify the geographical location or ecological type of the forest, thus facilitating a more accurate subsequent assessment of the specific fire risk scenarios faced by forests in different regions or types.

[0035] Step S102: Determine the target probability and target level of the target forest based on N fire spread factors and target labels.

[0036] In step S102, the target probability is used to characterize the probability of a fire occurring in the target forest, and the target level is used to characterize the scale of a fire occurring in the target forest.

[0037] Optionally, the system can determine the scale and severity of a forest fire based on the target probability. The classification of target levels helps forest management departments and emergency response teams to develop more specific prevention and control measures and resource allocation plans.

[0038] In step S102, the target probability is used to predict the likelihood of a fire occurring in the target forest at a certain time in the future by comprehensively analyzing the fire spread factors and target labels mentioned above. It transforms the complex forest fire risk assessment into an intuitive probability value, making it easier for users and relevant personnel to understand and apply. The target level is divided into five levels from low to high: low fire risk, relatively low fire risk, relatively high fire risk, high fire risk, and extremely high fire risk. This classification method can provide customized fire prevention guidance for different forest areas, ensuring the effective allocation of resources and timely response.

[0039] As can be seen from the above, this application analyzes the meteorological data, vegetation data, and topographic data of the target forest through a target model, thereby obtaining N fire spread factors and target labels corresponding to the target forest. This application uses target labels to represent the regional characteristics (e.g., regional type) of different areas where the target forest is located. Then, this application uses the fire spread rate of the target forest (i.e., fire spread factor) and the regional characteristics of different areas where the target forest is located as evaluation indicators for the probability and level of fire risk in the forest to be evaluated, thereby achieving the goal of improving the accuracy of predicting the probability and level of forest fire risk.

[0040] Therefore, this application achieves the goal of determining the probability and level of forest fire risk based on the fire spread factor and regional type of the target forest by comprehensively analyzing the meteorological data, vegetation data and topographic data of the target forest through the target model. This improves the accuracy of predicting the probability and level of forest fire risk and solves the technical problem of low accuracy in predicting the probability and level of forest fire risk in the prior art.

[0041] In one optional embodiment, before inputting the meteorological data, vegetation data, and topographic data of the target forest into the target model, the determination system first acquires the meteorological information, vegetation information, and topographic information of the target forest. The meteorological information includes at least the temperature, precipitation, wind speed, and humidity of the target forest; the vegetation information includes at least the tree information and biological information of the target forest; and the topographic information includes at least the regional information of the area where the target forest is located. Then, the determination system performs normalization processing and data cleaning on the meteorological information, vegetation information, and topographic information to obtain meteorological data, vegetation data, and topographic data. The normalization processing is used to unify the data format corresponding to the meteorological information, vegetation information, and topographic information, and the data cleaning is used to process the missing information in the meteorological information, vegetation information, and topographic information according to the interpolation method.

[0042] Optionally, the determination system improves the reliability of data subsequently input into the target model by normalizing and cleaning meteorological, vegetation, and topographic information. Normalization involves unifying meteorological, vegetation, and topographic information to the same numerical range or data format, thereby eliminating biases caused by differences in measurement units or scales, enabling the target model to correctly compare and analyze the data. During data cleaning, the determination system focuses on using interpolation to handle missing values ​​in the data. For example, if wind speed data for a certain time point is missing, the determination system can use wind speed data from the corresponding time points before and after that point to perform interpolation estimation, filling in the gaps and ensuring the integrity of the processed meteorological data.

[0043] In summary, the technical solutions of the above embodiments create a high-quality data input environment for the target model. First, by comprehensively and meticulously collecting meteorological, vegetation, and topographic information, the system ensures that the target model can obtain multi-dimensional data reflecting forest fire risk in a timely manner, thus laying a solid foundation for improving prediction accuracy. Second, the normalization and data cleaning steps, through standardizing data formats and filling missing values, further improve the quality of input data. This not only enables the model to better understand and interpret the data but also avoids prediction bias caused by data inconsistencies or missing values, ensuring the stability and reliability of the model. The technical solutions of the above embodiments enable the target model to conduct a more refined and reliable assessment of fire risk in specific forest areas based on rich and accurate input data, thereby providing scientific decision support for forest management departments and ultimately preventing and controlling forest fires.

[0044] In one optional embodiment, the N fire spread factors include at least meteorological factors, vegetation factors, and topographic factors. The meteorological factors include at least the daily maximum temperature, consecutive days without precipitation, daily maximum wind speed, and daily minimum relative humidity of the target forest. The vegetation factors include at least the forest type, forest canopy density, average tree height, biomass index, and temperature-vegetation-drought index of the target forest. The topographic factors include at least the slope, aspect, and altitude of the target forest.

[0045] Optionally, meteorological factors, vegetation factors, and topographic factors are shown in Table 1 below.

[0046] Table 1

[0047]

[0048] Optionally, the value range and coefficient interpretation of the N fire spread factors are as follows:

[0049] Optionally, the values ​​of the forest drought degree influence factor (TVDI) and forest canopy closure influence factor (CD) in the original data collected by the system are determined to be between 0 and 1. After multiplying the values ​​of the forest drought degree influence factor (TVDI) and forest canopy closure influence factor (CD) by a coefficient of 0.3, the results are directly applied to formula (10).

[0050] Optionally, the value of tree height (in meters) can be multiplied by a coefficient of 0.0025 (if the result is greater than 0.1, take 0.1 directly) to obtain the value of tree height level, which can then be directly applied to formula (10).

[0051] Optionally, the value of biomass (unit converted to t / ha) is multiplied by a coefficient of 0.0025 (0.2 is used when the result is greater than 0.2) to obtain the value of biomass level, which is then directly applied to formula (10).

[0052]

[0053] In the above formula (1), Tx is the daily maximum temperature value in °C, and Txlevel is the maximum temperature factor value. When the value of Tx is less than 10, Txlevel is 0.06; when the value of Tx is greater than 10 and less than or equal to 20, Txlevel is 0.09; when the value of Tx is greater than 20 and less than or equal to 25, Txlevel is 0.11; when the value of Tx is greater than 25 and less than or equal to 30, Txlevel is 0.13; and when the value of Tx is greater than 30, Txlevel is 0.15.

[0054]

[0055] In the above formula (2), RRR represents consecutive days without precipitation, in days, and RRRlevel is the influence factor of consecutive days without precipitation. When the RRR value is less than or equal to 2, the RRRlevel value is 0.16; when the RRR value is greater than 2 and less than or equal to 5, the RRRlevel value is 0.2; when the RRR value is greater than 5 and less than or equal to 7, the RRRlevel value is 0.24; when the RRR value is greater than 7 and less than or equal to 9, the RRRlevel value is 0.28; when the RRR value is greater than 9 and less than or equal to 12, the RRRlevel value is 0.32; when the RRR value is greater than 12 and less than or equal to 16, the RRRlevel value is 0.48; when the RRR value is greater than 16 and less than or equal to 20, the RRRlevel value is 0.6; and when the RRR value is greater than 20, the RRRlevel value is 0.72.

[0056]

[0057] In the above formula (3), FF3 is the maximum wind speed factor, with the unit being m / s, and FF3level is the daily maximum wind speed influence factor; when the value of FF3 is less than or equal to 3.5, the value of FF3level is 0.08; when the value of FF3 is greater than 3.5 and less than or equal to 6, the value of FF3level is 0.13; when the value of FF3 is greater than 6 and less than or equal to 9, the value of FF3level is 0.19; when the value of FF3 is greater than 9 and less than or equal to 13, the value of FF3level is 0.23; when the value of FF3 is greater than 13 and less than or equal to 17, the value of FF3level is 0.27; when the value of FF3 is greater than 17, the value of FF3level is 0.31.

[0058]

[0059] In the above formula (4), U is the minimum daily relative humidity, in %; Ulevel is the influence factor of the minimum daily relative humidity. When U is less than or equal to 30, Ulevel is 0.54; when U is greater than 30 and less than or equal to 40, Ulevel is 0.45; when U is greater than 40 and less than or equal to 50, Ulevel is 0.36; when U is greater than 50 and less than or equal to 60, Ulevel is 0.27; when U is greater than 60 and less than or equal to 70, Ulevel is 0.18; when U is greater than 70 and less than or equal to 80, Ulevel is 0.09; and when U is greater than 80, Ulevel is 0.

[0060]

[0061] In the above formula (5), forest_type is the forest type, forest_typelevel is the forest type influence factor, and the value range of the forest type influence factor is 0 to 1. When forest_type is assigned a value of 5, forest_typelevel takes a value of 1; when forest_type is assigned a value of 3, forest_typelevel takes a value of 0.5; when forest_type is assigned a value of 1, forest_typelevel takes a value of 0.7; when forest_type is assigned a value of 2 or 4, forest_typelevel takes a value of 0.3.

[0062]

[0063] In the above formula (6), DEM is the altitude in meters (m), and DEMlevel is the altitude influence factor. When the DEM value is less than or equal to 200, the DEMlevel value is 0.35; when the DEM value is greater than 200 and less than or equal to 600, the DEMlevel value is 0.26; when the DEM value is greater than 600 and less than or equal to 1500, the DEMlevel value is 0.15; when the DEM value is greater than 1500 and less than or equal to 3000, the DEMlevel value is 0.09; and when the DEM value is greater than 3000, the DEMlevel value is 0.03.

[0064]

[0065] In the above formula (7), slope is the slope, in percentage, and slopelevel is the slope influence factor. When the slope value is less than or equal to 10, the slopelevel value is 0.12; when the slope value is greater than 10 and less than or equal to 20, the slopelevel value is 0.33; when the slope value is greater than 20 and less than or equal to 35, the slopelevel value is 0.28; when the slope value is greater than 35 and less than or equal to 45, the slopelevel value is 0.2; and when the slope value is greater than 45, the slopelevel value is 0.04.

[0066]

[0067] In the above formula (8), aspect is the slope aspect in degrees, aspectlevel is the slope aspect influence factor. When the aspect value is greater than 158 and less than or equal to 247, aspectlevel is 0.32; when the aspect value is greater than 113 and less than or equal to 158, aspectlevel is 0.26; when the aspect value is greater than 247 and less than or equal to 337, aspectlevel is 0.14; when the aspect value is greater than 337 and less than or equal to 113, aspectlevel is 0.08.

[0068] Optionally, in the above embodiments, in order to accurately capture meteorological factors, the determination system not only monitors the daily maximum temperature, number of consecutive days without precipitation, daily maximum wind speed, and daily minimum relative humidity of the target forest in real time, but also uses professional meteorological instruments and satellite remote sensing technology to ensure the real-time nature and accuracy of the acquired meteorological data; for vegetation factors, the determination system uses ground surveys and high-resolution satellite imagery to accurately record the forest type, forest canopy density, average tree height, and biomass of the target forest, and at the same time, comprehensively assesses the drought degree of the vegetation by calculating the Temperature Vegetation Drought Index (TVDI); in terms of quantifying topographic factors, the determination system obtains the forest's elevation information through a digital elevation model (DEM) and uses topographic maps and GIS software to measure slope and aspect.

[0069] In summary, the technical solutions of the above embodiments, by refining meteorological factors, vegetation factors, and topographic factors, can more accurately capture various environmental factors affecting forest fire risk. The refined factors not only include environmental data crucial for fire risk assessment, such as temperature, precipitation, wind speed, and humidity, but also encompass complex information within the forest ecosystem, such as tree species, tree height, vegetation density, and drought level, as well as topographic slope, aspect, and altitude. Through diversified factor analysis, this application enables the assessment results to comprehensively reflect the fire risk faced by the target forest, thereby providing more detailed and accurate fire risk warnings for forest management departments, environmental protection agencies, and emergency response teams.

[0070] In one optional embodiment, firstly, the system determines the sum of coefficients corresponding to the daily maximum temperature, consecutive days without precipitation, daily maximum wind speed, and daily minimum relative humidity of the target forest as the first factor coefficient corresponding to the meteorological factor; the sum of coefficients corresponding to the forest type, forest canopy density, average tree height, biomass index, and temperature-vegetation-drought index of the target forest as the second factor coefficient corresponding to the vegetation factor; and the sum of coefficients corresponding to the slope, aspect, and altitude of the target forest as the third factor coefficient corresponding to the topographic factor. Then, the system determines the target probability based on the first factor coefficient, the second factor coefficient, and the third factor coefficient. Finally, the system determines the target level based on the first factor coefficient, the second factor coefficient, the third factor coefficient, and the target label.

[0071] Optionally, the formula for calculating the first factor coefficient is shown in the following formula (9).

[0072] A=Txlevel+RRRlevel+FF3level+Ulevel (9)

[0073] Optionally, the system converts key indicators of the target forest, such as the daily maximum temperature, number of consecutive days without precipitation, daily maximum wind speed, and daily minimum relative humidity, into corresponding quantitative coefficients to obtain temperature factor, no-precipitation day impact factor, maximum wind speed factor, and minimum relative humidity factor. This reflects the degree of influence of meteorological conditions on the probability of forest fires. The system adds the above coefficients to obtain the first factor coefficient, which essentially comprehensively considers the direct effect of short-term meteorological conditions on fire risk level, providing a meteorological baseline for subsequent fire risk prediction.

[0074] Optionally, the formula for calculating the second factor coefficient is shown in the following formula (10).

[0075] B=forest_typelevel+CD+Treeheightlevel+biomasslevel+TVDI(10)

[0076] Optionally, the system converts key indicators of the target forest, such as forest type, forest canopy density, average tree height, biomass index, and temperature-vegetation-drought index, into corresponding quantitative coefficients, resulting in forest type factor, forest canopy density factor, average tree height factor, biomass index factor, and temperature-vegetation-drought index factor. The second factor coefficient is the sum of these vegetation factor coefficients. The second factor coefficient measures the potential impact of biodiversity and flammability within the forest on fire spread, providing rich ecological background information for fire risk assessment.

[0077] Optionally, the formula for calculating the third factor coefficient is shown in the following formula (11).

[0078] C=DEMlevel+slopelevel+aspectlevel (11)

[0079] Optionally, the system determines the key indicators of slope, aspect and altitude of the target forest area and converts them into corresponding quantitative coefficients to obtain slope influence factor, aspect influence factor and altitude influence factor. The third factor coefficient is the sum of these topographic factor coefficients. The third factor coefficient means that the influence of topography on the direction and speed of fire spread is included in the fire risk prediction, which enhances the geographical adaptability and accuracy of the model.

[0080] Optionally, the technical solution of the above embodiments quantifies and converts the key factors (meteorology, vegetation, and topography) in the fire spread model into first, second, and third factor coefficients, which not only enhances the scientificity and practicality of forest fire risk assessment, but also realizes refined management of fire risk. In the above embodiments, the calculation of target probability is based on the weighted summation of factor coefficients, ensuring that the assessment results can comprehensively reflect the possibility of fire occurrence in the environment where the target forest is located; while the classification of target levels provides a graded risk assessment based on factor coefficients and the location characteristics of the target forest, making it easier for forest managers to take corresponding levels of fire prevention measures.

[0081] In one optional embodiment, firstly, the system determines the preset weights corresponding to the first factor coefficient, the second factor coefficient, and the third factor coefficient based on the target model. Then, the system performs a weighted summation of the first factor coefficient, the second factor coefficient, and the third factor coefficient based on the preset weights to obtain the target probability.

[0082] Optionally, the formula for calculating the target probability is shown in the following formula (12).

[0083] FRID=a*A+b*B+c*C (12)

[0084] In the above formula (12), a is the preset weight corresponding to the first factor coefficient, b is the preset weight corresponding to the second factor coefficient, c is the preset weight corresponding to the third factor coefficient, and FRID is the target probability.

[0085] Optionally, in the above embodiments, the determining system determines the preset weights corresponding to the first factor coefficient (meteorological factor), the second factor coefficient (vegetation factor), and the third factor coefficient (topography factor) based on the target model. The preset weights reflect the degree of contribution of each factor to the final fire risk prediction result during model training. For example, by analyzing historical forest data, the target model determines that the meteorological factor plays a dominant role in fire risk assessment. Therefore, the preset weight 'a' of the first factor coefficient is greater than the preset weight 'b' of the second factor coefficient, and the preset weight 'a' of the first factor coefficient is greater than the preset weight 'c' of the third factor coefficient.

[0086] In one optional embodiment, the system determines the fire risk probability index corresponding to the target forest based on the first factor coefficient, the second factor coefficient, and the third factor coefficient, wherein the fire risk probability index and the target probability are positively correlated. Then, the target level is determined based on the fire risk probability index and the target label.

[0087] Optionally, the formula for calculating the fire risk probability index is shown in the following formula (13).

[0088] (0.3*A / 4+0.6*B / 3+0.09*C / 3+0.01)*200 (13)

[0089] Optionally, in the calculation of the above factors, since the conditions for taking the values ​​of all factors cannot simultaneously trigger the extreme case of taking the maximum value (for example, the minimum relative humidity will not be in the minimum state when the tree height is above 40m, i.e., the area where the Treeheightlevel reaches the maximum value), the value of the fire risk probability index in formula (13) is less than 100. After obtaining the fire risk probability index, the system determines the target level by querying from the following table 2 based on the fire risk probability index and the target label.

[0090] Table 2

[0091]

[0092] Optionally, in Table 2 above, Zone 1, Zone 2, Zone 3, Zone 4, and Zone 5 are target labels divided according to different terrain features.

[0093] In one optional embodiment, the target model is trained through the following steps: First, a training set and a test set are generated based on historical forest data and historical fire data of L forests. The training set includes historical forest data and historical fire data of P forests, and the test set includes historical forest data and historical fire data of Q forests. P and Q are both positive integers, and L is equal to the sum of P and Q. Then, the initial neural network is iteratively trained using the historical forest data and historical fire data of the P forests in the training set to obtain an initial model. Then, the model parameters in the initial model are iteratively updated based on the historical forest data and historical fire data of the Q forests in the test set. The model obtained from the last update is used as the target model.

[0094] Optionally, the system improves the accuracy and practicality of forest fire risk assessment by analyzing historical forest data and historical fire data, and by combining iterative training and testing of neural networks. During model training, the separation of training and testing sets ensures that the model can learn general fire risk assessment rules, rather than just the characteristics of specific datasets, thereby enhancing the generalization ability of the target model. Furthermore, the target model in this application, after iterative training and parameter optimization, can predict fire risk levels based on real-time forest environmental data, thus providing decision support for forest management departments.

[0095] In one alternative embodiment, Figure 2 This is a flowchart of an optional forest fire risk assessment method according to an embodiment of this application, such as... Figure 2 As shown, the method includes the following steps:

[0096] (1) Data input: Collect data on combustible types, moisture content, topographic features, and real-time weather conditions for a specific forest area. The data collection stage in this step is crucial. The data collected in this step involves the collection of key parameters in the fire spread model of a specific forest area, such as tree height, biomass, combustible types, combustible moisture content, temperature, humidity, wind speed, and wind direction, to ensure that the subsequent target model can conduct fire risk assessment based on reliable real-time information.

[0097] In step (1) above, the specific method of collecting raw data is exemplified as follows:

[0098] Optionally, the forest canopy density influencing factor and the forest drought degree factor need to be calculated. NDVI (Normalized Difference Vegetation Index) and land surface temperature need to be calculated. It is determined that the system can calculate NDVI and land surface temperature based on downloaded MODIS (Moderate Resolution Imaging Spectroradiometer), NDVI products and land surface temperature products, or downloaded multi-view images of the target forest.

[0099] Optionally, the meteorological data corresponding to the daily maximum temperature influencing factor is updated along with the meteorological bureau data, and the system is able to directly analyze the meteorological bureau data to obtain daily average data, monthly average data, and difference acquisition surface data.

[0100] Optionally, the daily minimum relative humidity impact factor is updated with data from the China Meteorological Administration, and the difference is obtained from surface data. The impact factor for consecutive days without precipitation is determined by precipitation amount to indicate whether it rained that day.

[0101] Alternatively, forest type data can be determined using high-resolution data at the meter or sub-meter level.

[0102] Optionally, vector data of the area where the target forest is located can be obtained from data published by a geographic information center.

[0103] Optionally, tree height and biomass data can be determined using GEDI (Global Ecosystem Dynamics Investigation Lidar) datasets, which include at least ground height, vegetation height, canopy structure vegetation information, and biomass density information.

[0104] (2) Data preprocessing: The collected raw data is normalized and missing values ​​are processed. That is, data from different sources and types are converted into a unified format, and missing values ​​in the data are processed by interpolation or extrapolation, so as to provide a set of input parameters with consistent data format and high reliability for the fire spread model.

[0105] (3) Factor calibration: Input the preprocessed data into the fire spread model to obtain meteorological factors, vegetation factors and topographic factors. Among them, meteorological factors include the daily maximum temperature of the forest, consecutive days without precipitation, daily maximum wind speed and daily minimum relative humidity. Vegetation factors include forest type, forest canopy density, average tree height, biomass index and temperature-vegetation drought index. Topographic factors include forest slope, aspect and altitude.

[0106] (4) Fire risk prediction model construction: Based on the factors and factor proportions determined by the fire spread model, a fire risk prediction model is constructed.

[0107] (5) Fire risk level classification: Combining the fire risk prediction model and the fire risk meteorological level classification, the fire risk level is classified according to the prediction results of different regions.

[0108] (6) Modeling and verification: After obtaining the fire risk prediction probability value and fire risk level of a certain forest at a certain historical moment, this application verifies the predicted fire risk prediction probability value and fire risk level of the forest at the corresponding historical moment based on real historical fire risk data. If the prediction accuracy probability of the model is lower than the preset threshold, the model parameters are adjusted to obtain a reliable prediction model.

[0109] Optionally, Figure 3 This is a comparison chart of the predicted results and actual results for different optional regions according to an embodiment of this application, such as... Figure 3 As shown, this application classifies the fire risk levels of different regions based on the predicted results. Low and medium risk areas are less prone to fires. The probability of higher, higher, and extremely high risk fires occurring in each region type is statistically analyzed. Specifically, the probability of fire occurrence for cultivated land is 77.09%, for forest land 5.22%, for grassland 4.15%, and for other regions 13.54%. In verifying the prediction results based on historical fire risk data, the probability of fire occurrence for cultivated land is 48.09%, for forest land 17.57%, for grassland 1.62%, and for other regions 31.87%. Subsequently, this application ranks the predicted fire risk probabilities as follows: cultivated land > other > forest land > grassland.

[0110] Optionally, taking Province Z as an example, Province Z includes Zone Z1, Zone Z2, Zone Z3 and Zone Z4. Among them, the prediction accuracy of Zone Z2 and Zone Z3 is higher than that of Zone Z1 and Zone Z4. According to the statistics of the four zones, the accuracy of the fire risk prediction results of Province Z is 70.21%. The prediction accuracy of each zone in Province Z is shown in Table 3 below.

[0111] Table 3

[0112] area Sample size Predict the number of correct samples accuracy Z1 area 11 6 54.55％ Z2 area 2 2 100％ Z3 area 42 33 78.57％ Z4 area 237 164 69.20％ Z Province 292 205 70.21％

[0113] As can be seen from the above, this application analyzes the meteorological data, vegetation data, and topographic data of the target forest through a target model, thereby obtaining N fire spread factors and target labels corresponding to the target forest. This application uses target labels to represent the regional characteristics (e.g., regional type) of different areas where the target forest is located. Then, this application uses the fire spread rate of the target forest (i.e., fire spread factor) and the regional characteristics of different areas where the target forest is located as evaluation indicators for the probability and level of fire risk in the forest to be evaluated, thereby achieving the goal of improving the accuracy of predicting the probability and level of forest fire risk.

[0114] Therefore, this application achieves the goal of determining the probability and level of forest fire risk based on the fire spread factor and regional type of the target forest by comprehensively analyzing the meteorological data, vegetation data and topographic data of the target forest through the target model. This improves the accuracy of predicting the probability and level of forest fire risk and solves the technical problem of low accuracy in predicting the probability and level of forest fire risk in the prior art.

[0115] According to another aspect of the embodiments of this application, a device for determining forest fire risk levels is also provided. Figure 4 This is a schematic diagram of an optional forest fire risk level determination device according to an embodiment of this application, such as... Figure 4 As shown, the device for determining the forest fire risk level includes an input unit 401 and a determination unit 402.

[0116] Optionally, the input unit is used to input meteorological data, vegetation data, and topographic data of the target forest into the target model to obtain N fire spread factors and target labels corresponding to the target forest. The target model is a neural network model trained based on historical forest data and historical fire data of L forests, where L and N are both positive integers. The fire spread factors are used to characterize the speed of fire spread after a fire occurs in the target forest, and the target labels are used to characterize the regional type of the area where the target forest is located. The determination unit is used to determine the target probability and target level corresponding to the target forest based on the N fire spread factors and target labels. The target probability is used to characterize the probability of a fire occurring in the target forest, and the target level is used to characterize the scale of a fire occurring in the target forest.

[0117] In an optional embodiment, the forest fire risk level determination device further includes an acquisition unit and a preprocessing unit.

[0118] Optionally, the acquisition unit is used to acquire meteorological information, vegetation information, and topographic information of the target forest. The meteorological information includes at least the temperature, precipitation, wind speed, and humidity of the target forest. The vegetation information includes at least the tree information and biological information of the target forest. The topographic information includes at least the regional information of the area where the target forest is located. The preprocessing unit is used to perform normalization processing and data cleaning on the meteorological information, vegetation information, and topographic information to obtain meteorological data, vegetation data, and topographic data. The normalization processing is used to unify the data format corresponding to the meteorological information, vegetation information, and topographic information. The data cleaning is used to process the missing information in the meteorological information, vegetation information, and topographic information according to the interpolation method.

[0119] In one optional embodiment, the N fire spread factors in the forest fire risk level determination device include at least meteorological factors, vegetation factors, and topographic factors. The meteorological factors include at least the daily maximum temperature, consecutive days without precipitation, daily maximum wind speed, and daily minimum relative humidity of the target forest. The vegetation factors include at least the forest type, forest canopy density, average tree height, biomass index, and temperature-vegetation drought index of the target forest. The topographic factors include at least the slope, aspect, and altitude of the target forest.

[0120] In one optional embodiment, the determining unit includes: a first determining subunit, a second determining subunit, a third determining subunit, a fourth determining subunit, and a fifth determining subunit.

[0121] Optionally, the first determining subunit is used to sum the coefficients corresponding to the daily maximum temperature, consecutive days without precipitation, daily maximum wind speed, and daily minimum relative humidity of the target forest as the first factor coefficient corresponding to the meteorological factor; the second determining subunit is used to sum the coefficients corresponding to the forest type, forest canopy density, average tree height, biomass index, and temperature-vegetation-drought index of the target forest as the second factor coefficient corresponding to the vegetation factor; the third determining subunit is used to sum the coefficients corresponding to the slope, aspect, and altitude of the target forest as the third factor coefficient corresponding to the topographic factor; the fourth determining subunit is used to determine the target probability based on the first factor coefficient, the second factor coefficient, and the third factor coefficient; and the fifth determining subunit is used to determine the target level based on the first factor coefficient, the second factor coefficient, the third factor coefficient, and the target label.

[0122] In one optional embodiment, the fourth determining subunit includes: a first determining module and a second determining module.

[0123] Optionally, the first determining module is used to determine the preset weights corresponding to the first factor coefficient, the second factor coefficient, and the third factor coefficient according to the target model; the second determining module is used to perform a weighted summation of the first factor coefficient, the second factor coefficient, and the third factor coefficient according to the preset weights to obtain the target probability.

[0124] In one optional embodiment, the fifth determining subunit includes a third determining module and a fourth determining module.

[0125] Optionally, the third determining module is used to determine the fire risk probability index corresponding to the target forest based on the first factor coefficient, the second factor coefficient, and the third factor coefficient, wherein the fire risk probability index and the target probability are positively correlated; the fourth determining module is used to determine the target level based on the fire risk probability index and the target label.

[0126] In one optional embodiment, the forest fire risk level determination device further includes: a generation unit, an iterative training unit, an iterative update unit, and a target model determination unit.

[0127] Optionally, the generation unit is used to generate training and testing sets based on historical forest data and historical fire data of L forests, wherein the training set includes historical forest data and historical fire data of P forests, and the testing set includes historical forest data and historical fire data of Q forests, where P and Q are both positive integers, and L is equal to the sum of P and Q; the iterative training unit is used to iteratively train the initial neural network using the historical forest data and historical fire data of the P forests in the training set to obtain an initial model; the iterative update unit is used to iteratively update the model parameters in the initial model based on the historical forest data and historical fire data of the Q forests in the testing set; and the target model determination unit is used to take the model obtained from the last update as the target model.

[0128] As can be seen from the above, this application analyzes the meteorological data, vegetation data, and topographic data of the target forest through a target model, thereby obtaining N fire spread factors and target labels corresponding to the target forest. This application uses target labels to represent the regional characteristics (e.g., regional type) of different areas where the target forest is located. Then, this application uses the fire spread rate of the target forest (i.e., fire spread factor) and the regional characteristics of different areas where the target forest is located as evaluation indicators for the probability and level of fire risk in the forest to be evaluated, thereby achieving the goal of improving the accuracy of predicting the probability and level of forest fire risk.

[0129] Therefore, this application achieves the goal of determining the probability and level of forest fire risk based on the fire spread factor and regional type of the target forest by comprehensively analyzing the meteorological data, vegetation data and topographic data of the target forest through the target model. This improves the accuracy of predicting the probability and level of forest fire risk and solves the technical problem of low accuracy in predicting the probability and level of forest fire risk in the prior art.

[0130] According to another aspect of the embodiments of this application, a computer program product is also provided, the computer program product including a stored computer program, wherein, when the computer program is running, the computer program product is controlled to execute the method for determining the forest fire risk level described above.

[0131] According to another aspect of the embodiments of this application, an electronic device is also provided, including: a processor; and a memory for storing executable instructions of the processor; wherein the processor is configured to perform the forest fire risk level determination method of any one of the above by executing the executable instructions.

[0132] Optionally, Figure 5 This is a schematic diagram of an optional electronic device according to an embodiment of this application, such as... Figure 5As shown, this application provides an electronic device, which includes a processor, a memory, and a program stored in the memory and executable on the processor. When the processor executes the program, it implements the method for determining the forest fire risk level as described above.

[0133] The above-described embodiments or examples disclosed in this application are not exhaustive, but merely illustrative of some embodiments or examples, and are not intended to limit the scope of protection of this application. Unless otherwise specified, each step in a particular embodiment or example can be implemented as an independent embodiment, and the steps can be arbitrarily combined. For example, a solution after removing some steps in a particular embodiment or example can also be implemented as an independent embodiment, and the order of the steps in a particular embodiment or example can be arbitrarily interchanged. Furthermore, optional methods or examples in a particular embodiment or example can be arbitrarily combined; moreover, various embodiments or examples can be arbitrarily combined. For example, some or all steps of different embodiments or examples can be arbitrarily combined, and a particular embodiment or example can be arbitrarily combined with optional methods or examples of other embodiments or examples.

[0134] In the above embodiments of this application, the descriptions of each embodiment have different focuses. For parts not described in detail in a certain embodiment, please refer to the relevant descriptions of other embodiments.

[0135] This application is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of this application. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart... Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0136] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0137] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0138] In a typical configuration, a computing device includes one or more processors (CPUs), input / output interfaces, network interfaces, and memory. Memory may include non-persistent memory in computer-readable media, such as random access memory (RAM) and / or non-volatile memory, such as read-only memory (ROM) or flash RAM. Memory is an example of computer-readable media.

[0139] Computer-readable media includes both permanent and non-permanent, removable and non-removable media that can store information using any method or technology. Information can be computer-readable instructions, data structures, modules of programs, or other data. Examples of computer storage media include, but are not limited to, phase-change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technologies, CD-ROM, digital versatile optical disc (DVD) or other optical storage, magnetic tape, magnetic disk storage or other magnetic storage devices, or any other non-transferable medium that can be used to store information accessible by a computing device. As defined herein, computer-readable media does not include transient computer-readable media, such as modulated data signals and carrier waves.

[0140] It should also be noted that the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus. Unless otherwise specified, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes that element.

[0141] Those skilled in the art will understand that embodiments of this application can be provided as methods, systems, or computer program products. Therefore, this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this application can take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0142] The above are merely embodiments of this application and are not intended to limit the scope of this application. Various modifications and variations can be made to this application by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the scope of the claims of this application.

Claims

1. A method for determining forest fire risk levels, characterized in that, include: Meteorological data, vegetation data, and topographic data of the target forest are input into the target model to obtain N fire spread factors and target labels corresponding to the target forest. The target model is a neural network model trained based on historical forest data and historical fire data of L forests, where L and N are both positive integers. The fire spread factors are used to characterize the speed of fire spread after a fire occurs in the target forest, and the target labels are used to characterize the regional type of the area where the target forest is located. The target probability and target level of the target forest are determined based on the N fire spread factors and the target label, wherein the target probability is used to characterize the probability of the target forest fire and the target level is used to characterize the scale of the target forest fire.

2. The method for determining forest fire risk level according to claim 1, characterized in that, Before inputting meteorological, vegetation, and topographic data of the target forest into the target model, the method for determining the forest fire risk level includes: Acquire meteorological information, vegetation information and topographic information of the target forest, wherein the meteorological information includes at least the temperature, precipitation, wind speed and humidity of the target forest, the vegetation information includes at least the tree information and biological information of the target forest, and the topographic information includes at least the regional information of the area where the target forest is located. The meteorological information, vegetation information, and terrain information are normalized and cleaned to obtain the meteorological data, vegetation data, and terrain data. The normalization process is used to unify the data formats corresponding to the meteorological information, vegetation information, and terrain information, and the data cleaning process is used to process the missing information in the meteorological information, vegetation information, and terrain information according to the interpolation method.

3. The method for determining forest fire risk level according to claim 1, characterized in that, The N fire spread factors include at least meteorological factors, vegetation factors, and topographic factors. The meteorological factors include at least the daily maximum temperature, consecutive days without precipitation, daily maximum wind speed, and daily minimum relative humidity of the target forest. The vegetation factors include at least the forest type, forest canopy density, average tree height, biomass index, and temperature-vegetation drought index of the target forest. The topographic factors include at least the slope, aspect, and altitude of the target forest.

4. The method for determining forest fire risk level according to claim 3, characterized in that, Determining the target probability and target level of the target forest based on the N fire spread factors and the target label includes: The sum of the coefficients corresponding to the daily maximum temperature, consecutive days without precipitation, daily maximum wind speed, and daily minimum relative humidity of the target forest is taken as the first factor coefficient corresponding to the meteorological factor. The sum of the coefficients corresponding to the forest type, forest canopy density, average tree height, biomass index, and temperature-vegetation-drought index of the target forest is taken as the second factor coefficient corresponding to the vegetation factor. The sum of the coefficients corresponding to the slope, aspect, and altitude of the target forest is taken as the third factor coefficient corresponding to the terrain factor. The target probability is determined based on the first factor coefficient, the second factor coefficient, and the third factor coefficient. The target level is determined based on the first factor coefficient, the second factor coefficient, the third factor coefficient, and the target label.

5. The method for determining forest fire risk level according to claim 4, characterized in that, Determining the target probability based on the first factor coefficient, the second factor coefficient, and the third factor coefficient includes: Based on the target model, determine the preset weights corresponding to the first factor coefficient, the second factor coefficient, and the third factor coefficient; The target probability is obtained by weighting and summing the first factor coefficient, the second factor coefficient, and the third factor coefficient according to the preset weights.

6. The method for determining forest fire risk level according to claim 4, characterized in that, Determining the target level based on the first factor coefficient, the second factor coefficient, the third factor coefficient, and the target label includes: The fire risk probability index corresponding to the target forest is determined based on the first factor coefficient, the second factor coefficient, and the third factor coefficient, wherein the fire risk probability index and the target probability are positively correlated. The target level is determined based on the fire risk probability index and the target label.

7. The method for determining forest fire risk level according to claim 1, characterized in that, The target model is trained through the following steps: Training and testing sets are generated based on the historical forest data and historical fire data of the L forests. The training set includes historical forest data and historical fire data of P forests, and the testing set includes historical forest data and historical fire data of Q forests. P and Q are both positive integers, and L is equal to the sum of P and Q. The initial neural network is iteratively trained using historical forest data and historical fire data from P forests in the training set to obtain the initial model; The model parameters in the initial model are iteratively updated based on the historical forest data and historical fire data of the Q forests in the test set. The model obtained from the last update is used as the target model.

8. A device for determining forest fire risk levels, characterized in that, include: The input unit is used to input meteorological data, vegetation data and topographic data of the target forest into the target model to obtain N fire spread factors and target labels corresponding to the target forest. The target model is a neural network model trained based on historical forest data and historical fire data of L forests, where L and N are both positive integers. The fire spread factors are used to characterize the speed of fire spread after a fire occurs in the target forest, and the target labels are used to characterize the regional type of the area where the target forest is located. The determining unit is used to determine the target probability and target level corresponding to the target forest based on the N fire spread factors and the target label, wherein the target probability is used to characterize the probability of the target forest fire and the target level is used to characterize the scale of the target forest fire.

9. A computer program product, characterized in that, The computer program product includes a computer program, wherein, when the computer program is run, it controls the computer program product to execute the method for determining the forest fire risk level as described in any one of claims 1 to 7.

10. An electronic device, characterized in that, It includes one or more processors and a memory, the memory being used to store one or more programs, wherein when the one or more programs are executed by the one or more processors, the one or more processors cause the one or more processors to implement the method for determining the forest fire risk level as described in any one of claims 1 to 7.

Images