Forest fire spread simulation method, device, electronic device and storage medium

Abstract

The present invention provides a method, device, electronic device and storage medium for simulating the spread of forest fires, which are applied to the field of computer technology. The method comprises: determining a fire point in a preset forest area, and determining the initial cell corresponding to the fire point in the grid data according to the location of the fire point and the grid data corresponding to the forest area; determining an initial set of burning cells according to the initial cells; using a cellular automaton model to simulate and calculate the fire state and fire intensity corresponding to the neighbor cells of the burning cells according to the simulated environmental parameters and simulated terrain parameters corresponding to the burning cells and the forest area, and putting the target neighbor cells of the burning cells into the initial set of burning cells to obtain a new set of initial set of burning cells; iteratively executing the fire simulation calculation operation according to the new set of initial set of burning cells until a preset iteration cutoff condition is met. The technical solution of the present invention can easily realize the efficient simulation of the spread of forest fires.

Description

Technical Field

[0001] The present invention relates to the field of computer technology, and in particular to a forest fire spread simulation method, device, electronic equipment and storage medium. Background Art

[0002] Forest fire is a common natural disaster. The spread of forest fire will not only cause serious harm to the ecological environment, but also cause some economic losses. By simulating and predicting the spread of forest fire, we can provide strong decision-making support for fire fighting and command. Therefore, the simulation and prediction of forest fire spread is particularly important.

[0003] At present, the simulation methods for forest fire spread can be roughly divided into empirical model simulation forest fire spread method, physical model simulation forest fire spread method and deep learning model simulation forest fire spread method. Taking the deep learning simulation forest fire spread method as an example, this method can first collect a large amount of historical fire data, environmental data and geographical data during the simulation, and then annotate the fire spread range of these data to obtain training data. After that, the training data can be used to train the deep learning model to obtain a trained deep learning model, and then the trained deep learning model can be used to predict / simulate the spread of forest fires.

[0004] However, the above technologies have the problem that training data is difficult to obtain and model training is time-consuming. Summary of the invention

[0005] The present invention provides a forest fire spread simulation method, device, electronic device and storage medium to solve the defects in the prior art that training data is difficult to obtain and model training is time-consuming, so as to achieve the purpose of easily and efficiently simulating the spread of forest fires.

[0006] The present invention provides a forest fire spread simulation method, comprising:

[0007] Determine a fire point in a preset forest area, and determine an initial cell corresponding to the fire point in the grid data according to the location of the fire point and the grid data corresponding to the forest area; the grid data includes a plurality of cells constituting the forest area and the location of each cell;

[0008] Determine an initial ignition cell set according to the initial cells; the initial ignition cell set includes at least one ignition cell whose ignition state is ignition;

[0009] Executing a fire simulation calculation operation, the fire simulation calculation operation comprising: for each burning cell in the initial burning cell set, according to the simulated environment parameters and simulated terrain parameters corresponding to the burning cell and the forest area, using a preset cellular automaton model to simulate and calculate the fire state and fire intensity corresponding to the neighboring cells of the burning cell, and putting the target neighboring cells of the burning cell into the initial burning cell set to obtain a new initial burning cell set; the above-mentioned fire state indicates whether the neighboring cell is on fire;

[0010] According to the new initial ignition cell set, the above ignition simulation calculation operation is iteratively performed until a preset iteration cutoff condition is met, and the new initial ignition cell set that meets the iteration cutoff condition is determined as the target ignition cell set.

[0011] According to a forest fire spread simulation method provided by the present invention, the grid data also includes a grid code corresponding to each cell, and the simulated environment parameters and simulated terrain parameters corresponding to the burning cell and the forest area are used to simulate and calculate the fire state and fire intensity corresponding to the neighboring cells of the burning cell using a preset cellular automaton model, including:

[0012] According to the grid code of the burning cell and the grid codes of each cell, at least one neighboring cell adjacent to the burning cell is determined in the grid data; wherein the grid codes of the adjacent multiple cells are associated with each other;

[0013] If any neighbor cell of the burning cell is not in the initial burning cell set, any neighbor cell is put into the potential burning cell set; the potential burning cell set includes at least one potential burning neighbor cell;

[0014] Acquire simulated environmental parameters and simulated terrain parameters corresponding to the forest area; the simulated environmental parameters include simulated temperature, simulated humidity and simulated wind direction corresponding to the forest area, and the simulated terrain parameters include simulated terrain data and simulated blocking zone data corresponding to the forest area;

[0015] For each neighbor cell in the potential fire cell, the cellular automaton model is used to simulate and calculate the fire state corresponding to the neighbor cell according to the simulated wind direction, simulated terrain data and simulated blocking zone data of the forest area. The cellular automaton model is also used to simulate and calculate the fire intensity corresponding to the neighbor cell according to the simulated temperature and simulated humidity of the forest area.

[0016] According to a method for simulating the spread of forest fire provided by the present invention, the blocking zone data includes a plurality of simulated blocking zone vectors for blocking fire, and the cellular automaton model is used to simulate and calculate the fire state corresponding to the neighboring cells according to the simulated wind direction, simulated terrain data and simulated blocking zone data of the forest area, including:

[0017] According to the position of the burning cell and the position of the neighboring cells, the azimuth from the burning cell to the neighboring cells is calculated, and according to the azimuth and the simulated wind direction of the forest area, the angle between the azimuth and the simulated wind direction is calculated, and according to the angle and the preset first fire probability calculation formula, the first probability of whether the neighboring cell is on fire is calculated;

[0018] According to the position of the burning cell and the position of the neighboring cells, the elevation of the burning cell and the elevation of the neighboring cells are obtained from the elevation data of the forest area, and the elevation difference between the elevation of the burning cell and the elevation of the neighboring cells is calculated, and according to the elevation difference and a preset second fire probability calculation formula, the second probability of whether the neighboring cell is on fire is calculated;

[0019] According to the position of the burning cell and the position of the neighboring cells, the connection line between the burning cell and the neighboring cells is determined, and it is judged whether the connection line and each simulated blocking zone vector intersect to obtain the blocking zone determination result;

[0020] According to the first probability, the second probability and the blocking band determination result, the fire state corresponding to the neighboring cell is determined.

[0021] According to a forest fire spread simulation method provided by the present invention, the above-mentioned cellular automaton model is used to simulate and calculate the fire intensity corresponding to the neighboring cells according to the simulated temperature and simulated humidity of the forest area, including:

[0022] According to the simulated temperature corresponding to the forest area, a simulated temperature influence factor corresponding to the simulated temperature is determined in a preset first ignition function; the first ignition function includes different temperature ranges and a temperature influence factor corresponding to each temperature range, and the temperature influence factor indicates the influence of the corresponding temperature range on the fire intensity;

[0023] According to the simulated humidity corresponding to the forest area, a simulated humidity influence factor corresponding to the simulated humidity is determined in a preset second ignition function; the second ignition function includes different humidity ranges and humidity influence factors corresponding to each humidity range, and the humidity influence factors represent the influence of the corresponding humidity range on the fire intensity;

[0024] According to the product of the simulated temperature influence factor and the simulated humidity influence factor, the fire intensity corresponding to the neighbor cell is obtained.

[0025] According to a forest fire spread simulation method provided by the present invention, the method further comprises:

[0026] Acquire vector data of forest areas, and merge the forest areas into one forest area according to the vector data;

[0027] According to the area of ​​the forest area, the grid level for dividing the forest area is determined, and the forest area is divided into hexagonal grids according to the grid level to obtain multiple hexagonal grids corresponding to the forest area;

[0028] According to the hexagonal vectors corresponding to each hexagonal grid, the position of the center point corresponding to each hexagonal grid is calculated;

[0029] The hexagonal grids corresponding to the center points are determined as cells, and the positions of the center points are determined as the positions corresponding to the cells.

[0030] According to a forest fire spread simulation method provided by the present invention, the grid data also includes the radius of the hexagonal grid corresponding to each cell, and the method further includes:

[0031] According to the radius of the hexagonal grid corresponding to the cell, the distance between the burning cell and one of its neighboring cells is calculated, and the distance is used as the iterative spreading distance;

[0032] According to an iterative spread distance and the preset single fire spread time, the unit fire spread speed of the forest area is simulated and calculated;

[0033] According to the type of forest area, the simulated windless spreading speed and the simulated wind force influence coefficient corresponding to the forest area are determined in a preset speed function; the speed function includes the windless spreading speed and the wind force influence coefficient corresponding to different types of forest areas, and the wind force influence coefficient corresponds to the wind force level;

[0034] According to the above-mentioned simulated windless spread speed and simulated wind influence coefficient, the simulated fire spread speed in the forest area is calculated, and according to the simulated fire spread speed and unit fire spread speed, the simulated multiple of the fire spread speed in the forest area is calculated.

[0035] According to a forest fire spread simulation method provided by the present invention, the above-mentioned iteration cutoff condition includes any one of the following:

[0036] The burning cell in the current iteration round has no neighbor cells;

[0037] The current number of iterations reaches the preset number of iterations.

[0038] The present invention also provides a forest fire spread simulation device, comprising the following modules:

[0039] An initial cell determination module is used to determine a fire point in a preset forest area, and determine an initial cell corresponding to the fire point in the grid data according to the location of the fire point and the grid data corresponding to the forest area; the above-mentioned grid data includes a plurality of cells constituting the forest area and the location of each cell;

[0040] An initial ignition cell set determination module is used to determine an initial ignition cell set according to the initial cells; the initial ignition cell set includes at least one ignition cell whose ignition state is ignition;

[0041] The fire simulation calculation module is used to perform the fire simulation calculation operation, which includes: for each burning cell in the initial burning cell set, according to the simulated environment parameters and simulated terrain parameters corresponding to the burning cell and the forest area, a preset cellular automaton model is used to simulate and calculate the fire state and fire intensity corresponding to the neighboring cells of the burning cell, and the target neighboring cells of the burning cell are placed into the initial burning cell set to obtain a new initial burning cell set; the above-mentioned fire state indicates whether the neighboring cell is on fire;

[0042] The iteration module is used to iteratively execute the above-mentioned ignition simulation calculation operation according to the new initial ignition cell set until a preset iteration cutoff condition is met, and determine the new initial ignition cell set that meets the iteration cutoff condition as the target ignition cell set.

[0043] The present invention also provides an electronic device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the computer program, the forest fire spread simulation method as described in any one of the above is implemented.

[0044] The present invention also provides a non-transitory computer-readable storage medium having a computer program stored thereon, and when the computer program is executed by a processor, the forest fire spread simulation method as described in any one of the above is implemented.

[0045] The present invention also provides a computer program product, comprising a computer program, wherein when the computer program is executed by a processor, the method for simulating the spread of forest fire as described in any one of the above is implemented.

[0046] The forest fire spread simulation method, device, electronic device and storage medium provided by the present invention determine the initial cell corresponding to the fire point in the grid data by the position of the fire point determined in the preset forest area and the grid data corresponding to the forest area, and determine the initial fire cell set including at least one fire cell with a fire state of being on fire according to the initial cell, and then perform the fire simulation calculation operation, the fire simulation calculation operation includes: for each fire cell in the initial fire cell, according to the fire cell, the simulated environmental parameters of the forest area and the simulated terrain parameters, using a preset cellular automaton model to simulate and calculate the fire state and fire intensity corresponding to the neighbor cells of the fire cell, and put the target neighbor cells of the fire cell into the initial fire cell to obtain a new initial fire cell set, iteratively perform the fire simulation calculation operation according to the new initial fire cell set until the preset iteration cutoff condition is met, and determine the new initial fire cell set that meets the iteration cutoff condition as the target fire cell set; wherein the grid data includes multiple cells constituting the forest area and the position of each cell. In this method, since only one fire point needs to be selected in the forest area, the cellular automaton model can be used to automatically iteratively simulate the fire status and fire intensity of the neighbors around the fire point to obtain the final set of fire cells. This does not require too much historical data and it is easier to select a fire point in the forest area. At the same time, the cellular automaton model has higher simulation efficiency and does not require model training, so it is easier to achieve efficient simulation of the spread of forest fires. BRIEF DESCRIPTION OF THE DRAWINGS

[0047] In order to more clearly illustrate the technical solutions in the present invention or the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0048] Figure 1 This is one of the flow charts of the forest fire spread simulation method provided by the present invention.

[0049] Figure 2 It is a schematic diagram of the data of the forest area collected in advance provided by the present invention.

[0050] Figure 3 This is a data schematic diagram of the forest area after the polygons are merged provided by the present invention.

[0051] Figure 4 It is a schematic diagram of the hexagonal grid for dividing the forest area provided by the present invention.

[0052] Figure 5It is a schematic diagram of extracting the center points of each hexagonal grid provided by the present invention.

[0053] Figure 6 This is the second flow chart of the forest fire spread simulation method provided by the present invention.

[0054] Figure 7 It is a schematic diagram of the overall process of the forest fire spread simulation method provided by the present invention.

[0055] Figure 8 It is a structural schematic diagram of the forest fire spread simulation device provided by the present invention.

[0056] Fig. 9 It is a structural schematic diagram of the electronic device provided by the present invention. DETAILED DESCRIPTION

[0057] In order to make the purpose, technical solution and advantages of the present invention clearer, the technical solution of the present invention will be clearly and completely described below in conjunction with the drawings of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0058] Forest fire is a common natural disaster. The spread of forest fire not only brings serious harm to the ecological environment, but also causes huge economic losses to the country and society. Forest fire is characterized by suddenness, randomness, rapid spread, destructiveness, and complexity. Once a fire occurs, if it is not intervened in time, it will cause immeasurable losses. The simulation and prediction of the spread of forest fire can provide strong decision-making support for fire fighting and command. The current methods of forest fire spread include simulation through empirical models, simulation through physical models, simulation through deep learning models, etc.

[0059] Among them, the empirical model simulation method can use the Rothermel model, combined with the GIS (Geographic Information System) visualization capability, to visualize the spread of forest fires. It plays an important role in forest fire spread simulation, especially when there is sufficient data and stable conditions, they can provide fast and intuitive prediction results. However, there are many shortcomings in empirical models: first, empirical models are usually based on observational and experimental data under specific conditions, so they may only be applicable to specific types of fires and environments; second, empirical models usually rely on some simplified assumptions for ease of calculation and application, however, these assumptions may not accurately reflect the actual fire spread process; third, empirical models are less adaptable to environmental changes. For example, climate change may cause changes in forest fire behavior, and empirical models based on past data may not accurately predict the behavior of future fires; finally, forest fire spread is a complex multi-factor interaction process, including wind force, wind direction, humidity, temperature, terrain and vegetation type, etc. Empirical models can usually only capture some of these factors, and it is difficult to fully describe these complex interactions. In summary, due to the defects and shortcomings of empirical models, in practical applications, it is usually necessary to combine other types of models (such as physical models, machine learning methods, etc.) to improve the accuracy and reliability of predictions.

[0060] Through the method of physical model simulation, for example, spatial data provided by GIS and RS (Remote Sensing) can be used to simulate the spread of forest fires on a large spatial and temporal scale and reflect the instantaneous state of forest fire behavior. Physical models have many advantages in simulating the spread of forest fires, but they also have some shortcomings. First, the computational complexity of physical models is relatively high. They usually require detailed simulation of the combustion process, heat transfer, airflow, etc. of the fire, involving a large number of differential equations and numerical calculations. Such computational complexity is relatively high and usually requires strong computing power and a long computing time. Secondly, model parameters are difficult to obtain. Physical models require a large number of parameters, such as fuel type, fuel moisture, terrain, wind force, wind direction, etc. The accurate measurement and acquisition of these parameters may be very difficult in practical applications. In addition, these parameters may vary greatly in different regions and conditions. Thirdly, physical models are sensitive to initial conditions. Small errors may lead to huge differences in simulation results. The accurate setting of initial conditions is difficult to achieve in practical applications, especially in the case of large-scale fires. Finally, the physical model has a large demand for data, which requires a large amount of high-precision data input, including detailed terrain data, vegetation types and distribution, meteorological data, etc. The acquisition and processing of these data is very cumbersome and costly. In short, the physical model has high accuracy and detailed description capabilities in forest fire spread simulation, but its high computational complexity, difficulty in parameter acquisition, and difficulty in model verification limit its widespread use in practical applications.

[0061] For example, the simulation method of deep learning models can use big data and deep learning technology to combine historical and real-time fire data for model training. The trained deep learning model can more accurately predict the spread of fires, so as to take countermeasures in advance and reduce the loss of human life and property caused by fires. However, deep learning models also have the following shortcomings in predicting the spread of forest fires: First, deep learning models require a large amount of high-quality training data to achieve good performance. For forest fire simulation, it may be very difficult to obtain enough historical fire data, environmental data (such as temperature, humidity, wind speed, etc.), and geographical data (such as terrain, vegetation distribution, etc.); second, the training time is long, and the training process of deep learning models usually requires a lot of computing resources and time, especially when dealing with large-scale data sets. High-performance computing devices such as GPUs (graphics processing units) or TPUs (tensor processing units) are required, which may increase the cost of research and application; finally, parameter tuning complex deep learning models have many hyperparameters (such as learning rate, batch size, number of network layers, etc.) that need to be tuned to achieve optimal performance. Finding the right combination of parameters can be very time-consuming and challenging.

[0062] Based on this, an embodiment of the present invention provides a forest fire spread simulation method, device, electronic device and storage medium, which can solve the above technical problems.

[0063] Combine the following Figure 1-Figure 7 A forest fire spread simulation method according to an embodiment of the present invention is described.

[0064] It should be noted that the execution subject of the embodiments of the present invention may be a forest fire spread simulation device, or may be an electronic device including the forest fire spread simulation device, or may be other devices or systems. No specific limitation is made here. The following embodiments will be described using an electronic device as an example of the execution subject.

[0065] Figure 1 FIG. 1 is one of the flow charts of the forest fire spread simulation method provided by the present invention, such as Figure 1 As shown, the method comprises the following steps:

[0066] S102, determining a fire point in a preset forest area, and determining an initial cell corresponding to the fire point in the grid data according to the location of the fire point and the grid data corresponding to the forest area; the above grid data includes multiple cells constituting the forest area and the location of each cell.

[0067] Before simulating the fire spread of a preset forest area, relevant data of the forest area for fire spread simulation can be collected in advance, such as DEM (Digital Elevation Model) data, combustible data, fire isolation zone data, etc. Figure 2 The diagram shows the pre-collected forest area data, where the black and white grid data is the preset DEM elevation data of the forest area, the multiple green polygons are the combustible data of the forest area (the combustible data is generally vector data, so it can also be called public welfare forest vector data), and the red line is the fire isolation zone vector data. The forest area here can be, for example, a grassland area, a bush area, a forest surface area, a tree crown area, etc.

[0068] After obtaining the relevant data of the forest area, since the vector data of the forest area may be composed of multiple polygons, in order to facilitate the subsequent gridding process, optionally, the vector data of the forest area can be obtained first, and the forest area can be merged into one forest area according to the vector data, that is, multiple polygons of the forest area can be merged into one large polygon, for example, see Figure 3 The data diagram of the forest area after the polygons are merged is shown, in which a green polygon area is the combustible data of the merged forest area. In order to facilitate the explanation of the spread of fire, the combustible data of the forest area will be explained as the forest area.

[0069] After merging multiple polygonal areas, as an optional embodiment, the grid level for dividing the forest area can be determined based on the area of ​​the forest area, and the forest area can be divided into hexagonal grids according to the grid level to obtain multiple hexagonal grids corresponding to the forest area; based on the hexagonal vectors corresponding to each hexagonal grid, the position of the center point corresponding to each hexagonal grid is calculated; the hexagonal grids corresponding to each center point are determined as cells, and the position of each center point is determined as the position corresponding to each cell.

[0070] Specifically, the area of ​​the forest area can be calculated through the vector data of the forest area, and then the grid level of the grid division can be determined according to the size of the area. For example, the smaller the area, the higher the grid level. For example, the grid level of the forest area in the present invention can be determined as level 10. If the forest area is smaller than the forest area in the present invention, for example, the 11th grid level can be selected, wherein the radius of each grid divided in the 11th grid level is 28.663897 meters, and the area of ​​each grid is 2149.643 square meters. In addition, it should be noted that when dividing the grid, it is also necessary to consider that the radius of each divided grid should be controlled within the range of 100 meters, that is, the radius of the divided grid should not be too large, and it is generally more appropriate to be in the range of tens of meters. In this way, the fire spread simulated by the divided grid is more in line with the actual fire spread and more valuable for reference.

[0071] After determining the grid level, Uber's h3 encoding technology can be used to divide the vector data of the merged forest area into hexagonal grids according to the 10-level grid level to obtain multiple hexagonal grids. Figure 4 The schematic diagram of the hexagonal grid for dividing the forest area is shown in FIG., which includes multiple hexagonal grids, each of which has a radius of 75.863783 meters and an area of ​​15047.502 square meters. Here, each hexagonal grid is a cell in the cellular automaton model. When the hexagonal grid is divided as described above, the hexagonal vector corresponding to each hexagonal grid (including the position, side length, direction and other information of the hexagonal vertices) can also be obtained. Then, the position of the center point of each hexagonal grid can be calculated through the hexagonal vector of each hexagonal grid, and the corresponding center point can be extracted from each hexagonal grid. The center point of each hexagonal grid is used to represent the corresponding hexagonal grid or cell. Here, please refer to Figure 5The schematic diagram of extracting the center point of each hexagonal grid is shown, that is, the position of the center point of each hexagonal grid is used as the position of the corresponding cell, which is convenient for the subsequent calculation of the cell state. The position here can be, for example, longitude and latitude. In addition, when the hexagonal grid is divided as above, the h3 grid code corresponding to each divided hexagonal grid can also be obtained (that is, the grid code of each cell is obtained), and then the hexagonal vector, center point position and grid code of each hexagonal grid can be used as the grid data of the forest area.

[0072] Here, the forest area is divided into hexagonal grids. Compared with the traditional rectangular grid, which has four neighbors (von Neumann neighborhood) or eight neighbors (Moore neighborhood), the hexagonal grid can provide an intermediate and more uniform adjacency relationship. The distances between all six neighbors are equal, thus avoiding the diagonal distance in the rectangular grid ( The difference between the length of the side (times the length of the side) and the horizontal or vertical distance leads to uneven or asymmetric adjacency, which may introduce directional deviation in the simulation of fire spread. For example, in the process of expansion or diffusion, the rectangular grid may tend to expand along the direction of the grid instead of expanding evenly in all directions. The hexagonal grid used in the present invention can better avoid such deviation due to its symmetry, thereby improving the accuracy of the final forest fire spread.

[0073] After dividing the forest area into hexagonal grids, a point can be randomly selected as the starting point in the fire area in the forest area (if there is no fire area in the initial simulation state, a point can be randomly selected as the starting point in the grid data of the forest area) to obtain the location of the starting point, recorded as s(x,y). Then a buffer zone b can be made with r as the radius, where r is the grid radius, that is, the radius of the hexagonal grid divided above, for example r=75.863783; then the buffer zone b can be used to overlay the center point layer (that is, the hexagonal grid data obtained after extracting the center point above) to obtain the set of center points C (c1, c2, c3, ..., c n ,); then we can traverse the set C and find the distance d between each center point in C and s respectively n , and find out the values ​​of each d n The minimum value min(d n ), the minimum value min(d n ) corresponds to the center point c n is the center point closest to s, and this center point can be recorded as the initial cell.

[0074] The distance between two points is generally calculated using the following spherical distance calculation formula:

[0075] .

[0076] Where d is the distance between the two points being calculated, and R is the radius of the earth, which is generally 6371000 meters. is the longitude and latitude of a point, is the longitude and latitude of another point.

[0077] In addition, after obtaining the divided hexagonal grids, each hexagonal grid represents a cell, and the structure object Stru{fire, strength, position, id} can be set for the cell. Among them, fire is the fire state (0-not on fire, 1 on fire), strength is the fire intensity (0-1), position is the position (x, y), id is the unique identifier of the cell, and the h3 grid code is used here. Then the structure object of the initial cell here can be recorded as {fire=1, strength=1, position=(x, y), id=grid code}, that is, the initial cell is the cell on fire (hereinafter referred to as the on fire cell).

[0078] Furthermore, the formats of the above-mentioned related data are explained, where the center point of the hexagonal grid is the center point of each hexagonal grid after the combustible polygon is divided into 10 levels of grids, which is a point vector. The fire isolation belt is a line vector, assuming that the spread of forest fire cannot cross the fire isolation belt. DEM data, elevation data, is raster type data.

[0079] S104, determining an initial ignition cell set according to the initial cell; the initial ignition cell set includes at least one ignition cell whose ignition state is ignition.

[0080] In this step, after the initial cell is determined, the structure object of the initial cell can be placed in the initial fire cell set. The initial fire cell set is an empty set at the beginning. As the fire spreads, fire cells with a fire state of being on fire are continuously generated, and the fire cells are placed in the empty set to obtain the initial fire cell set. The initial fire cell set here can be recorded as an initial FCC (FiredCellularCollection), where the fire cells included in the initial fire cell set can be recorded as FC (FiredCellular).

[0081] S106, executing a fire simulation calculation operation, the fire simulation calculation operation includes: for each burning cell in the initial burning cell set, according to the simulated environment parameters and simulated terrain parameters corresponding to the burning cell and the forest area, using a preset cellular automaton model to simulate and calculate the fire state and fire intensity corresponding to the neighboring cells of the burning cell, and putting the target neighboring cells of the burning cell into the initial burning cell set to obtain a new initial burning cell set; the above-mentioned fire state indicates whether the neighboring cell is on fire.

[0082] In this step, after obtaining the initial set of burning cells, a fire simulation calculation operation can be performed, and the fire simulation operation specifically includes: traversing the initial set of burning cells, and for each burning cell therein, calculating the fire state and fire intensity of its adjacent neighboring cells. Taking a burning cell as an example, the simulated environmental parameters and simulated terrain parameters of the forest area can be obtained, and then the structure object, simulated environmental parameters and simulated terrain parameters of the burning cell are input into the cellular automaton model, and the cellular automaton model is used to automatically simulate and calculate the fire state and fire intensity corresponding to the neighboring cells of the burning cell. Among them, the cellular automaton model here can pre-configure the fire state calculation method and the fire intensity calculation method, so that the fire state and fire intensity of the neighboring cells can be automatically simulated and calculated.

[0083] The above-mentioned simulated environmental parameters may include simulated temperature, simulated humidity, simulated wind direction, simulated wind force, etc. corresponding to the forest area. The simulated terrain parameters may include simulated terrain data, simulated blocking zone data, etc. corresponding to the forest area.

[0084] The following is an explanation of simulated temperature, simulated humidity, simulated wind direction, simulated wind force, etc.:

[0085] Simulated temperature: represents the current simulated temperature, which is an enumeration type and can be divided into three sections: less than or equal to 10℃, greater than 10℃ and less than or equal to 20℃, and greater than 20℃.

[0086] Simulated humidity: indicates the simulated humidity at the current moment. It is an enumeration type and can be divided into four sections: less than or equal to 30%, greater than 30% and less than or equal to 55%, greater than 55% and less than or equal to 75%, and greater than 75%.

[0087] Simulated wind force: enumeration type, which can be divided into: level 1, level 2, level 3, level 4, level 5, level 6, level 7, level 8 and above.

[0088] Simulated wind direction: enumeration type, 8 directions in total, including: east, southeast, south, southwest, west, northwest, north, and northeast.

[0089] It should be noted that the division intervals of the above parameters are all example intervals, which can be adjusted adaptively according to the actual simulation situation. In addition, the above simulated temperature, simulated humidity, simulated wind force and simulated wind direction are variable parameters, and the above variable parameters can be adjusted for each step of the iteration process. The center point data of the above hexagonal grid, DEM data (i.e. simulated terrain data), simulated blocking belt data (i.e. fire isolation belt data), etc. are fixed parameters and remain unchanged during the entire fire spread simulation process.

[0090] After the cellular automaton model is used to simulate and calculate the ignition state of the neighbor cells of each ignited cell in the initial ignited cell set, the neighbor cells whose ignition state is ignited can be obtained, recorded as target neighbor cells, and put into the initial ignited cell set to obtain a new initial ignited cell set corresponding to the current iteration round. In addition, it can be understood that the neighbor cells of each ignited cell can be obtained through the grid code of each cell in the grid data, that is, the grid code between each ignited cell and its adjacent neighbor cells is related.

[0091] S108, iteratively executing the above-mentioned ignition simulation calculation operation according to the new initial ignition cell set until a preset iteration cutoff condition is satisfied, and determining the new initial ignition cell set satisfying the iteration cutoff condition as the target ignition cell set.

[0092] In this step, after obtaining the new initial set of ignition cells corresponding to the current iteration round, the variable parameters (simulated temperature, simulated humidity, simulated wind force and simulated wind direction) may be changed or not, and then based on the new initial set of ignition cells of the first iteration round, the second iteration is continued, that is, the fire simulation calculation operation is performed for the second time to obtain the new initial set of ignition cells corresponding to the second iteration round. Then, based on the new initial set of ignition cells of the second iteration round, the third iteration is continued, that is, the fire simulation calculation operation is performed for the third time to obtain the new initial set of ignition cells corresponding to the third iteration round, and the fire simulation calculation operation is iteratively performed in this way to continuously simulate the spread of forest fires until the preset iteration cutoff condition is met at the current moment.

[0093] As an optional embodiment, the above-mentioned iteration cutoff condition includes any one of the following: the burning cell in the current iteration round has no neighbor cells (that is, the burning cell has reached the boundary of the forest area and the fire can no longer spread); the current number of iterations reaches the preset number of iterations (the size of the preset number of iterations here can be set according to actual conditions).

[0094] In the above-mentioned process of iterative simulation of fire spread, when the iteration cutoff condition is currently met, a new initial set of burning cells obtained in the current iteration round can be obtained, and the new initial set of burning cells is recorded as the set of burning cells after the final fire spread, which can be recorded as the target set of burning cells (i.e., the target FCC). It can be understood that the target set of burning cells includes multiple burning cells and their structure objects. The position in the structure object of each burning cell can be used to determine the fire spread range of the forest area, and the fire intensity of each burning cell can be obtained. Then, the fire spread range and fire intensity of the forest area can be accurately simulated through multi-dimensional and multi-component information such as the fire state, fire intensity, and the location of the fire point, so as to provide more accurate and effective reference indicators for forest fire rescue and disaster relief.

[0095] In this embodiment, the initial cell corresponding to the fire point in the grid data is determined by the position of the fire point determined in the preset forest area and the grid data corresponding to the forest area, and an initial fire cell set including at least one fire cell with a fire state of being on fire is determined according to the initial cell, and then a fire simulation calculation operation is performed, and the fire simulation calculation operation includes: for each fire cell in the initial fire cells, according to the fire cell and the simulated environmental parameters and simulated terrain parameters of the forest area, a preset cellular automaton model is used to simulate and calculate the fire state and fire intensity corresponding to the neighbor cells of the fire cell, and the target neighbor cells of the fire cell are placed in the initial fire cell to obtain a new initial fire cell set, and the fire simulation calculation operation is iteratively performed according to the new initial fire cell set until a preset iteration cutoff condition is met, and the new initial fire cell set that meets the iteration cutoff condition is determined as the target fire cell set; wherein, the grid data includes multiple cells constituting the forest area and the position of each cell. In this method, since only one fire point needs to be selected in the forest area, the cellular automaton model can be used to automatically iteratively simulate the fire status and fire intensity of the neighbors around the fire point to obtain the final set of fire cells. This does not require too much historical data and it is easier to select a fire point in the forest area. At the same time, the cellular automaton model has higher simulation efficiency and does not require model training, so it is easier to achieve efficient simulation of the spread of forest fires.

[0096] The following embodiment illustrates the process of simulating the fire state and fire intensity corresponding to the neighboring cells of the burning cell.

[0097] Figure 6 FIG. 2 is a flow chart of the forest fire spread simulation method provided by the present invention. Figure 6As shown, the grid data also includes the grid code corresponding to each cell. The above S106 "according to the simulated environment parameters and simulated terrain parameters corresponding to the burning cell and the forest area, using a preset cellular automaton model to simulate and calculate the fire state and fire intensity corresponding to the neighboring cells of the burning cell" may include the following steps:

[0098] S202, determining at least one neighboring cell adjacent to the fire cell in the grid data according to the grid code of the fire cell and the grid codes of each cell; wherein the grid codes of the adjacent cells are correlated.

[0099] In this step, as mentioned above, the grid codes of the divided multiple hexagonal grids (or cells) are correlated, and the neighbor cells adjacent to each cell in the forest area can be obtained through this correlation. Here, the number of neighbor cells of each cell can be one to six, that is, each cell includes at least one neighbor cell and at most six neighbor cells.

[0100] Specifically, after obtaining the initial set of on-fire cells, one or more neighbor cells adjacent to each on-fire cell can be found in each cell according to the grid code of each cell. The neighbor cells here can be recorded as NC (Neighbour Cellular).

[0101] S204: If any neighbor cell of the burning cell is not in the initial burning cell set, any neighbor cell is put into a potential burning cell set; the potential burning cell set includes at least one potential burning neighbor cell.

[0102] In this step, after obtaining the neighbor cells corresponding to each ignited cell in the initial ignited cell set, the grid code of each neighbor cell can be used to search in the initial ignited cell set to determine whether each neighbor cell is in the initial ignited cell set (i.e., whether it is in the initial FCC); if a neighbor cell is not in the initial ignited cell set, it means that the neighbor cell is not currently on fire, but because the adjacent cell next to it is a ignited cell, it may be on fire, so the neighbor cell can be placed in a temporary potential ignited cell set. The potential ignited cell set here can be recorded as PFCC (Potential Fired Cellular Collection), and the neighbor cells included in it are all potential ignited neighbor cells.

[0103] The potential fire cell set here can be an empty set at the beginning. As the fire spreads, neighbor cells that may potentially catch fire are continuously generated, and their corresponding structure objects are continuously put into the empty potential fire cell set to obtain a potential fire cell set including neighbor cells. It can be understood that the structure objects of the neighbor cells in the potential fire cell set here have fire=0 and strength=0.

[0104] If the neighbor cell is in the initial set of burning cells, it means that the neighbor cell is already a burning cell, and the fire simulation calculation operation can be performed according to the execution operation of the burning cell.

[0105] S206, obtaining simulated environmental parameters and simulated terrain parameters corresponding to the forest area; the above-mentioned simulated environmental parameters include simulated temperature, simulated humidity and simulated wind direction corresponding to the forest area, and the above-mentioned simulated terrain parameters include simulated terrain data and simulated blocking zone data corresponding to the forest area.

[0106] In this step, for the simulated environmental parameters of the forest area, the simulated environmental parameters of the forest area can be obtained according to the data setting method in the above S106, or the actual environmental parameters of the forest area can be obtained by collecting through relevant data acquisition equipment (such as temperature and humidity sensors, etc.) when the forest area is actually on fire, and used as the simulated environmental parameters for fire spread simulation.

[0107] For the simulated terrain parameters of the forest area, after the forest area is selected, the actual terrain parameters of the forest area can be obtained through the relevant geographic data platform as the simulated terrain parameters for fire spread simulation, or the corresponding terrain parameters can be simulated by itself according to the terrain of the forest area. There is no specific limitation here.

[0108] S208, for each neighbor cell in the potential fire cell, according to the simulated wind direction, simulated terrain data and simulated blocking zone data of the forest area, a cellular automaton model is used to simulate and calculate the fire state corresponding to the neighbor cell, and according to the simulated temperature and simulated humidity of the forest area, a cellular automaton model is used to simulate and calculate the fire intensity corresponding to the neighbor cell.

[0109] In this step, after obtaining the potential fire cell set as mentioned above, the potential fire cell set can be traversed, and for each neighbor cell in the potential fire cell set, the fire state corresponding to the neighbor cell can be simulated and calculated using the cellular automaton model through the simulated wind direction, simulated terrain data and simulated blocking zone data of the forest area, and the fire intensity corresponding to the neighbor cell can be simulated and calculated using the cellular automaton model using the simulated temperature and simulated humidity.

[0110] It should be noted that the cellular automaton model here can pre-configure a calculation method for calculating the fire state of the cell through data such as wind direction, terrain data, and blocking zone data. At the same time, it can configure a calculation method for calculating the fire intensity of the cell through data such as temperature and humidity. In this way, the simulated wind direction, simulated terrain data, and simulated blocking zone data of the forest area can be input into the cellular automaton model, and the fire state of the neighboring cell can be automatically simulated and calculated through the cellular automaton model. At the same time, the simulated temperature and simulated humidity of the forest area can be input into the cellular automaton model, and the fire intensity of the neighboring cell can be automatically simulated and calculated through the cellular automaton model.

[0111] It can be understood that when the simulation calculates that the fire state of the neighbor cell is not on fire, the subsequent simulation calculation steps of the fire intensity of the neighbor cell can be omitted, and the fire intensity strength in the structure object of the neighbor cell can be directly recorded as 0, which can save simulation calculation amount and improve calculation efficiency.

[0112] Furthermore, it can be understood that this step is to calculate the corresponding fire state and fire intensity of each neighbor cell in the potential fire cell, which is actually a process of iteratively performing fire simulation calculation operations on each fire cell in the initial fire cell set, that is, first taking out a fire cell from the initial fire cell set, determining its corresponding potential fire neighbor cells for the fire cell, and calculating the fire state and fire intensity of each potential fire neighbor cell of the fire cell, and then taking out a new fire cell from the initial fire cell set to continue the simulation calculation, until the initial fire cell set is empty, and one iteration is completed.

[0113] In this embodiment, the grid code of each burning cell can be used to determine its neighbor cells and the potential burning neighbor cells can be determined based on this. Then, for each potential burning neighbor cell, the temperature, humidity, wind direction, terrain data of the forest area and the blocking zone data are used to simulate and calculate its fire state and fire intensity. In this way, the fire state and fire intensity of the neighbor cells of the burning cell can be accurately simulated, and the calculation efficiency of the simulation can be improved.

[0114] The following example illustrates a possible implementation method for simulating and calculating the fire status corresponding to neighbor cells.

[0115] In some embodiments, the blocking zone data includes multiple simulated blocking zone vectors that block fires. The step of "using a cellular automaton model to simulate and calculate the fire status corresponding to neighboring cells according to the simulated wind direction, simulated terrain data, and simulated blocking zone data of the forest area" in S208 may include the following steps:

[0116] Step A1, according to the position of the burning cell and the position of the neighboring cells, calculate the azimuth from the burning cell to the neighboring cells, and according to the azimuth and the simulated wind direction of the forest area, calculate the angle between the azimuth and the simulated wind direction, and according to the angle and the preset first fire probability calculation formula, calculate the first probability of whether the neighboring cell is on fire.

[0117] Among them, the center point of each hexagonal grid extracted during gridding in the above embodiment represents each hexagonal grid or cell, so the position of the cell can be obtained when the center point of the grid is extracted, that is, the position of the center point of the grid is the position of the cell.

[0118] This step mainly calculates the probability that the neighboring cells of the burning cell may catch fire after being affected by the wind direction. The burning cell is recorded as FC, and its neighboring cell is recorded as NC. Suppose the coordinates of FC are (x1, y1), and the coordinates of NC are (x2, y2). Calculate the azimuth from FC to NC azimuth , the steps are as follows.

[0119] 1. Calculate the difference d between the horizontal coordinates of FC and NC x The difference between the vertical coordinate and y , respectively: d x =x2-x1,d y =y2-y1;

[0120] Taking FC as the origin, if d x >0 and d y >0, the azimuth is in the first quadrant, and the azimuth is calculated as follows:

[0121] ;

[0122] If d x >0 and d y <0, the azimuth is in the second quadrant. The azimuth calculation formula is as follows:

[0123] ;in, for π The capital letters of , about 3.14;

[0124] If d x <0 and d y >0, the azimuth is in the third quadrant, and the azimuth is calculated as follows:

[0125] ;

[0126] If d x <0 and d y <0, the azimuth is in the fourth quadrant. The azimuth calculation formula is as follows:

[0127] ;

[0128] If d x = 0 and d y >0, the azimuth points to the due north direction. The azimuth calculation formula is as follows:

[0129] ;

[0130] If d x = 0 and d y <0, the azimuth points to the south direction. The azimuth calculation formula is as follows:

[0131] ;

[0132] If d x >0 and d y =0, the azimuth points to the east direction. The calculation formula of the azimuth is as follows:

[0133] ;

[0134] If d x <0 and d y =0, the azimuth points to the due west. The azimuth calculation formula is as follows:

[0135] ;

[0136] Then convert the azimuth angle from radians to degrees to facilitate probability calculation. The conversion formula is as follows:

[0137] ; azimuth' is the converted azimuth.

[0138] According to the wind direction rule, the probability of fire spreading along the wind direction is generally higher. Assume there are 8 wind directions, and the corresponding angles are: east - 90, southeast - 135, south - 180, southwest - 225, west - 270, north - 0, northeast - 45. Let the angle variable of the wind direction be windDirection ;

[0139] The calculation formula for the first probability of NC ignition, which is determined by wind direction, is as follows:

[0140] ;

[0141] in, windDirection_probabilit y represents the calculated first probability of whether the neighboring cell NC is on fire (related to the wind direction windDirection Related), angle Indicates the azimuth and wind direction after the above conversion ( windDirection), the calculation formula is as follows:

[0142] if ,but ;

[0143] if ,but .

[0144] Step A2, according to the position of the burning cell and the position of the neighboring cells, obtain the elevation of the burning cell and the elevation of the neighboring cells from the elevation data of the forest area, calculate the elevation difference between the elevation of the burning cell and the elevation of the neighboring cells, and calculate the second probability of whether the neighboring cell is on fire according to the elevation difference and a preset second fire probability calculation formula.

[0145] Among them, this step mainly calculates the probability that the neighboring cells of the burning cell may catch fire after being affected by the terrain. After obtaining the position of the burning cell and the position of the neighboring cells, the DEM elevation data of the forest area includes the elevation data of each position / point in the forest area. Then, the elevation of the burning cell position can be found from the position of the burning cell and the position of the neighboring cells (denoted as FC_elevation ), and find the elevation of the neighbor cell position (denoted as NC_elevation ). Then, based on the elevation of the burning cell and the elevation of the neighboring cells, the probability that the neighboring cells may catch fire due to the influence of the terrain can be calculated.

[0146] Assume that the blocking height is DecisionHeight =20, the blocking height means that the fire spreads to higher places, but when the height difference exceeds the blocking height, such as encountering a cliff, the fire probability (that is, the second probability of neighbor cells catching fire due to the influence of terrain) is reduced.

[0147] Here we can first calculate FC_elevation and NC_elevation Then, according to the following second fire probability calculation formula, the value of the second probability terrain_probability of the terrain affecting the fire of neighboring cells is determined. The second fire probability calculation formula is as follows:

[0148] Assume that the initial value of terrain_probability is terrain_probability=1;

[0149] if ,but

[0150] terrain_probability=0.75;

[0151] if ,or

[0152] ,

[0153] terrain_probability=0.5.

[0154] Step A3, according to the position of the burning cell and the position of the neighboring cells, determine the connection line between the burning cell and the neighboring cells, and judge whether the connection line and each simulated blocking zone vector intersect to obtain the blocking zone determination result.

[0155] In this step, the main purpose is to calculate whether the fire spread will be affected by the blocking zone between the burning cell and the neighboring cells. After obtaining the position of the burning cell and the position of the neighboring cell, the two positions can be connected to obtain the connecting line between the two and the vector corresponding to the connecting line. At the same time, when collecting relevant data of the forest area, multiple blocking zone vector data of the forest area can also be obtained (recorded as simulated blocking zone vectors). Then, it can be calculated whether the vector corresponding to the connecting line between the burning cell and the neighboring cell and each simulated blocking zone vector intersect to obtain the blocking zone determination result. The specific intersection determination method is not described in detail here.

[0156] Assume that the obtained blocking zone determination result is recorded as isBlock, which is a Boolean variable with an initial value of false. It can be updated after the intersection determination. If the vector corresponding to the line connecting the ignited cell and the neighboring cell intersects with any simulated blocking zone vector, isBlock=true; if the vector corresponding to the line connecting the ignited cell and the neighboring cell does not intersect with any simulated blocking zone vector, isBlock=false. Here, false means that the fire is not blocked, and true means that the fire is blocked.

[0157] Step A4, determining the fire state corresponding to the neighboring cell according to the first probability, the second probability and the blocking band determination result.

[0158] In this step, the first probability of the neighbor cells being affected by the wind direction is obtained above. windDirection_ probability and the second probability of fire affected by terrain terrain_probability After that, the total probability of fire of neighbor cells can be calculated by the first probability and the second probability fire_probability , the formula is as follows:

[0159] .

[0160] After that, the fire status of the neighboring cells can be determined using the following formula: isFire :

[0161] ;

[0162] in, Represents the logical NOT operation, Represents a logical "AND" operation, probability_function represents the probability judgment function, and its judgment formula is as follows:

[0163] ; random is a random number between 0 and 1 generated by the program. The formula means: if the total probability of fire of neighbor cells is greater than the product of the random number and 100, then the function probability_function The result is true if the function is set, otherwise it is false.

[0164] The meaning of the judgment formula of the fire state of the above neighbor cells is: perform a logical AND operation on the result after the blocking judgment result is negated and the result of the probability judgment function. If both are true, the final fire state of the neighbor cell is true, that is, it is on fire; if at least one of the result after the blocking judgment result is negated and the result of the probability judgment function is false, then the final fire state of the neighbor cell is false, that is, it is not on fire.

[0165] Through the above method, the fire status of neighbor cells can be calculated accurately and quickly.

[0166] In this embodiment, the probability of fire of neighbor cells affected by wind direction, the probability of fire affected by terrain, and the probability of fire affected by blocking zone are calculated, and the fire status of neighbor cells is comprehensively determined based on these fire probabilities. In this way, a variety of rich factors are used to simulate the fire status of neighbor cells, and the fire status determination result finally obtained is more accurate and more in line with the actual situation.

[0167] The following example illustrates a possible implementation method for simulating and calculating the fire intensity corresponding to neighbor cells.

[0168] In some embodiments, the above-mentioned S208 of “using a cellular automaton model to simulate and calculate the fire intensity corresponding to the neighboring cells according to the simulated temperature and simulated humidity of the forest area” may include the following steps:

[0169] Step B1, according to the simulated temperature corresponding to the forest area, determine the simulated temperature influence factor corresponding to the simulated temperature in the preset first ignition function; the above-mentioned first ignition function includes different temperature ranges and temperature influence factors corresponding to each temperature range, and the above-mentioned temperature influence factors represent the influence of the corresponding temperature range on the fire intensity.

[0170] Among them, high temperature can preheat combustibles, and preheated combustibles are easier to burn and the fire is bigger. Therefore, this step mainly calculates the influence of temperature on the ignition of neighbor cells. Different temperatures have different influences on the fire. Therefore, the temperature can be divided into different temperature ranges, such as three temperature ranges: less than or equal to 10℃, greater than 10℃ and less than or equal to 20℃, and greater than 20℃.

[0171] Here, it is assumed that the influence of temperature on the ignition of neighboring cells can be recorded as the temperature influence factor temperature_factor, where the temperature_factor corresponding to different temperature ranges is also different. It is assumed that the first ignition function can be pre-set, which includes each temperature range and its corresponding temperature_factor. The first ignition function is expressed as follows:

[0172] When the temperature is less than or equal to 10℃, temperature_factor=0.5;

[0173] When the temperature is greater than 10℃ and less than or equal to 20℃, temperature_factor=0.8;

[0174] When the temperature is greater than 20°C, temperature_factor=1.

[0175] It should be noted that the temperature ranges divided in the first ignition function and the temperature_factors corresponding to the different temperature ranges can be adaptively adjusted according to different types of forests, as long as the temperature_factor increases with the temperature, that is, the temperature range is proportional to the temperature_factor.

[0176] Specifically, after obtaining the simulated temperature of the current iteration round, the simulated temperature is brought into the first ignition function for matching, so as to obtain a temperature influence factor matching the simulated temperature, which is recorded as the simulated temperature influence factor.

[0177] Step B2, according to the simulated humidity corresponding to the forest area, determine the simulated humidity influence factor corresponding to the simulated humidity in the preset second ignition function; the above-mentioned second ignition function includes different humidity ranges and humidity influence factors corresponding to each humidity range, and the above-mentioned humidity influence factors represent the influence of the corresponding humidity range on the fire intensity.

[0178] Among them, when the humidity is high, the water content of the combustible is high, it is not easy to ignite and the combustion intensity is weak. Therefore, this step mainly calculates the impact of humidity on the ignition of neighbor cells. Different humidity has different effects on the fire. Therefore, the humidity can be divided into different humidity ranges here, such as four humidity ranges, namely: less than or equal to 30%, greater than 30% and less than or equal to 55%, greater than 55% and less than or equal to 75%, and greater than 75%.

[0179] Here, it is assumed that the influence of humidity on the ignition of neighbor cells can be recorded as the humidity influence factor moisture_factor, where the moisture_factor corresponding to different humidity ranges is also different. It is assumed that a second ignition function can be pre-set, which includes each humidity range and its corresponding moisture_factor. The second ignition function is expressed as follows:

[0180] When the humidity is less than or equal to 30%, moisture_factor=1;

[0181] When the humidity is greater than 30% and less than or equal to 55%, moisture_factor=0.8;

[0182] When the humidity is greater than 55% and less than or equal to 75%, moisture_factor=0.65;

[0183] When the humidity is greater than 75%, moisture_factor=0.5.

[0184] It should be noted that the humidity ranges divided in the second ignition function and the moisture_factor sizes corresponding to different humidity ranges can be adaptively adjusted according to different types of forests, as long as the moisture_factor size decreases with increasing humidity, that is, the size of the humidity range is inversely proportional to the size of the moisture_factor.

[0185] Specifically, after obtaining the simulated humidity of the current iteration round, the simulated humidity is brought into the second ignition function for matching, so as to obtain a humidity influence factor matching the simulated humidity, which is recorded as the simulated humidity influence factor.

[0186] Step B3, obtaining the fire intensity corresponding to the neighboring cell according to the product of the simulated temperature influence factor and the simulated humidity influence factor.

[0187] In this step, after obtaining the simulated temperature influence factor temperature_factor and the simulated humidity influence factor moisture_factor of the current iteration round, the two can be multiplied to obtain the fire intensity of the neighboring cell or the fire intensity strength of the current iteration round. The formula is as follows:

[0188] .

[0189] Furthermore, the fire intensity can be reflected by the transparency of the cells during the visualization simulation process, that is, the more transparent the cells are, the smaller the fire intensity is.

[0190] In this embodiment, the simulated temperature and humidity are used to determine the simulated temperature influence factor and the simulated humidity influence factor in the corresponding ignition function, and the fire intensity is determined jointly by the temperature and humidity influence factors. In this way, a variety of rich factors are used to simulate the fire intensity of neighbor cells, and the fire intensity finally obtained is more accurate and more in line with the actual situation.

[0191] The following example illustrates a possible implementation method for calculating the fire spread speed during the fire spread simulation process.

[0192] In some embodiments, the grid data further includes the radius of the hexagonal grid corresponding to each cell, and the method further includes the following steps:

[0193] Step C1, according to the radius of the hexagonal grid corresponding to the cell, calculate the distance between the burning cell and one of its adjacent neighbor cells, and use the distance as an iterative spreading distance.

[0194] In this step, taking the fire spread as a cell unit as an example, the distance between two cells can be calculated. Here, taking the burning cell as an example, it can be recorded as the distance between the burning cell and one of its neighboring cells. Assuming that the calculated distance is recorded as d, the calculation formula of d is as follows:

[0195] ;

[0196] Among them, r is the radius of the hexagonal grid corresponding to the cell. After the above 10-level grid division, r=75.863783 meters can be obtained.

[0197] After d is calculated using the above formula, the calculated distance can be used as a unit spreading distance or an iterative spreading distance.

[0198] Step C2, based on an iterative spread distance and a preset single fire spread time, simulate and calculate the unit fire spread speed in the forest area.

[0199] In this step, the iterative process of simulating the fire in the above embodiment is executed by the timer. The timer is executed once every 1000 mm (i.e., 1 second) by default, and its frequency f=1000 milliseconds (i.e., the single fire spread time). The unit fire spread speed simulated by the electronic equipment of the computer is v It can be calculated using the following formula:

[0200] ;

[0201] in, v The unit is meters per second (multiply by 1000 to convert milliseconds to seconds).

[0202] Step C3, according to the type of the forest area, determine the simulated windless spread speed and the simulated wind influence coefficient corresponding to the forest area in the preset speed function; the above speed function includes the windless spread speed and wind influence coefficient corresponding to different types of forest areas, and the above wind influence coefficient corresponds to the wind level.

[0203] Among them, the fire spread speed is generally related to or determined by the wind force. The empirical formula for the relationship between the fire spread speed and wind force can be expressed as follows:

[0204] ;

[0205] Among them, R is the actual speed of fire spread, R0 is the speed of fire spread under windless conditions, e is the natural logarithm value, which is 2.71828, k is the wind influence coefficient (positive number), which reflects the influence of wind on the speed of fire spread, and w is the wind force level.

[0206] Different forest types (such as grassland types, shrub types, forest surface types, canopy types, etc.) have different fire spread speeds R0 and wind influence coefficients k under windless conditions. The following is a pre-established speed function, which includes windless spread speeds and wind influence coefficients corresponding to different types of forest areas, as follows:

[0207] Grassland fire: R0 is 0.1~0.5 m / min, k is between 0.1 and 0.2;

[0208] Brush fire: R0 is 0.2~1.0 m / min, k is between 0.15 and 0.25;

[0209] Forest surface fire: R0 is 0.05~0.3 m / min, k is between 0.1 and 0.3;

[0210] Crown fire: R0 is 0.5~1.5 m / min, and k is between 0.1 and 0.3.

[0211] Assuming that the fire spread of the crown type is simulated in this embodiment, the simulated windless spread speed R0 and the simulated wind force influence coefficient k corresponding to the crown fire can be obtained in the above speed function. For example, R0 can be 1 meter / minute, and k can be 0.2. The corresponding wind force level can be obtained by simulating the wind force influence coefficient k, and the actual spread speed of the fire corresponding to the crown fire can be obtained by combining the simulated windless spread speed R0.

[0212] Step C4, calculates the simulated fire spread speed in the forest area according to the simulated windless spread speed and the simulated wind influence coefficient, and calculates the simulated multiple of the fire spread speed in the forest area according to the simulated fire spread speed and the unit fire spread speed.

[0213] In this step, the simulated windless spread speed and the simulated wind influence coefficient of the currently simulated forest area can be used to calculate the actual fire spread speed R (i.e., the simulated fire spread speed) of the type of forest area, and then the unit fire spread speed v , the actual fire spread rate / speed multiple n is simulated and calculated by the following formula, as follows:

[0214] ;

[0215] The reason for multiplying by 60 is that the unit of R is meter / minute. v The unit of is meter / second, so the units of the two need to be consistent, that is, both are converted into meter / minute.

[0216] In this step, the simulated multiples of the fire spread speed in the forest area are calculated. In this way, in the relevant program for simulating the fire spread, the speed of the fire spread in the simulation program can be changed by adjusting the frequency f of the timer execution, so that the fire spread speed in different types of forest areas can be simulated.

[0217] In this embodiment, the multiple of the fire spread is determined by calculating the unit fire spread speed and simulating the actual fire spread speed of the corresponding type of forest area. This makes it easy to obtain different multiples by adjusting the frequency f of the timer execution to achieve the purpose of simulating the fire spread speed of different types of forest areas.

[0218] In order to explain the technical solution of the present invention in more detail, a detailed embodiment is given below for illustration. Figure 7 The overall process diagram of the forest fire spread simulation method shown in FIG. 1 may include the following steps:

[0219] First, select the ignition point and set the initial state for the cellular automaton model (or called cellular automaton model), that is, obtain the initial cell, and obtain the initial ignition cell set FCC including at least one ignition cell.

[0220] Second, according to the simulated environment parameters, simulated terrain parameters, and the burning cell, calculate the state of its neighboring cells and perform an iteration, which specifically includes:

[0221] 1. Take out a fire cell FC from the initial FCC, calculate the neighbor cells of each FC, each cell has 1 to 6 neighbor cells NC; and put the NCs that are not in the initial FCC into the potential fire cell set PFCC;

[0222] 2. According to the ignition function of wind direction affecting ignition, the ignition function of terrain affecting ignition, and the blocking determination results of the blocking zone, comprehensively determine whether the cell in the PFCC is the next ignition cell;

[0223] 3. Traverse the potential fire cell PFCC set of the current FC's neighbors, execute step 2, and determine the fire status of the NC cells in the PFCC;

[0224] 4. Traverse the initial FCC, take out each ignition cell FC, and execute step 3;

[0225] 5. Place the ignition cell into the FCC to complete one iteration.

[0226] 3. Simulation of fire spread speed.

[0227] 1. For the second step above, perform the next iteration, calculate the neighbor cells NC of the burning cell FC, and simulate the fire spread speed by controlling the iteration frequency (that is, the frequency of execution of the above timer).

[0228] From the above description, it can be seen that the forest fire spread simulation method based on the hexagonal multi-component cellular automaton model proposed in the present invention can perform comprehensive forest fire range prediction, dynamic fire intensity simulation, and real-time spread rate calculation. It has high-efficiency computing performance and adaptability to multiple factors, which can significantly improve the accuracy and efficiency of forest fire prediction and emergency response. At the same time, the model can also play an important role in forest fire prevention, emergency command and fire warning. The specific technical effects of the technical solution of the present invention are as follows:

[0229] 1. The original purpose of cellular automaton theory is to deduce the states of other cells through the states and rules of cells, which is suitable for simulating fluid movement. The spread of forest fire can be likened to a special kind of fluid movement, so cellular automaton is very suitable for fire spread simulation.

[0230] 2. Multi-state simulation capability: Cellular automata can define multiple states, which can not only simulate the range of fire spread, but also add fire intensity states. Fire intensity and spread range are both important parameters in forest fire rescue, which can help decision makers understand the fire situation more comprehensively.

[0231] 3. Real-time data requirements: The cellular automaton model does not require a large amount of historical data, but only real-time data at the time. Based on these real-time data, the model can infer the state of neighboring cells according to the set rules. This makes cellular automata very suitable for real-time prediction and emergency response.

[0232] 4. Time dimension simulation: In addition to simulation in spatial scope, cellular automata can also simulate the speed of fire spread, that is, the time dimension. By adjusting the iteration frequency of the simulation program, the model can dynamically reflect the fire spread process and provide predictions on the development of fire.

[0233] 5. High computational efficiency. Compared with complex physical models and deep learning models, the cellular automaton model has higher computational efficiency. It does not require solving complex differential equations or a large amount of data training. It can perform large-scale simulations in a shorter time, which is very suitable for real-time prediction and emergency response.

[0234] 6. Simple and intuitive. Cellular automata simulate complex global behaviors by defining simple local rules. This method is intuitive and easy to understand. It does not require complex mathematical formulas or advanced statistical knowledge. Users can easily build and adjust the model.

[0235] 7. Flexibility and scalability. The cellular automaton model has high flexibility and can adjust the cell state and transition rules as needed to adapt to different simulation requirements. The model has good scalability and can be easily expanded into a multi-state, multi-level model to adapt to different types of fire behaviors.

[0236] 8. Suitable for parallel computing. Since the calculation process of the cellular automaton model is based on the update of local rules and neighbor states, it is very suitable for parallel computing. Multi-core processors or distributed computing resources can be used to improve computing efficiency and conduct large-scale fire spread simulations.

[0237] The forest fire spread simulation device provided by the present invention is described below. The forest fire spread simulation device described below and the forest fire spread simulation method described above can be referenced to each other.

[0238] Figure 8 is a schematic diagram of the structure of the forest fire spread simulation device provided by the present invention, see Figure 8 As shown, the device may include:

[0239] The initial cell determination module 310 is used to determine the fire point in the preset forest area, and determine the initial cell corresponding to the fire point in the grid data according to the location of the fire point and the grid data corresponding to the forest area; the above-mentioned grid data includes multiple cells constituting the forest area and the location of each cell;

[0240] The initial ignition cell set determination module 320 is used to determine the initial ignition cell set according to the initial cells; the initial ignition cell set includes at least one ignition cell whose ignition state is ignition;

[0241] The fire simulation calculation module 330 is used to perform a fire simulation calculation operation, which includes: for each burning cell in the initial burning cell set, according to the simulated environment parameters and simulated terrain parameters corresponding to the burning cell and the forest area, using a preset cellular automaton model to simulate and calculate the fire state and fire intensity corresponding to the neighboring cells of the burning cell, and putting the target neighboring cells of the burning cell into the initial burning cell set to obtain a new initial burning cell set; the above-mentioned fire state indicates whether the neighboring cell is on fire;

[0242] The iteration module 340 is used to iteratively perform the above-mentioned fire simulation calculation operation according to the new initial fire cell set until a preset iteration cutoff condition is met, and determine the new initial fire cell set that meets the iteration cutoff condition as the target fire cell set.

[0243] Optionally, the above iteration cutoff condition includes any one of the following: the ignited cell in the current iteration round has no neighboring cells; the current iteration number reaches a preset iteration number.

[0244] In some embodiments, the grid data further includes a grid code corresponding to each cell, and the fire simulation calculation module 330 may include:

[0245] A neighbor cell determination unit, used to determine at least one neighbor cell adjacent to the fire cell in the grid data according to the grid code of the fire cell and the grid codes of each cell; wherein the grid codes of the adjacent multiple cells are associated with each other;

[0246] A potential fire cell set determination unit, used for placing any neighbor cell of the fire cell into the potential fire cell set if the neighbor cell is not in the initial fire cell set; the potential fire cell set includes at least one potential fire neighbor cell;

[0247] A parameter acquisition unit, used to acquire simulated environmental parameters and simulated terrain parameters corresponding to the forest area; the simulated environmental parameters include simulated temperature, simulated humidity and simulated wind direction corresponding to the forest area, and the simulated terrain parameters include simulated terrain data and simulated blocking zone data corresponding to the forest area;

[0248] The fire simulation calculation unit is used to simulate and calculate the fire state corresponding to each neighbor cell in the potential fire cell using a cellular automaton model according to the simulated wind direction, simulated terrain data and simulated blocking zone data of the forest area, and to simulate and calculate the fire intensity corresponding to the neighbor cell using a cellular automaton model according to the simulated temperature and simulated humidity of the forest area.

[0249] In some embodiments, the blocking band data includes multiple simulated blocking band vectors that block fires. The fire simulation calculation unit is specifically used to calculate the azimuth from the burning cell to the neighboring cell according to the position of the burning cell and the position of the neighboring cell, and calculate the angle between the azimuth and the simulated wind direction according to the azimuth and the simulated wind direction of the forest area, and calculate the first probability of whether the neighboring cell is on fire according to the angle and a preset first fire probability calculation formula; obtain the elevation of the burning cell and the elevation of the neighboring cell from the elevation data of the forest area according to the position of the burning cell and the position of the neighboring cell, calculate the elevation difference between the elevation of the burning cell and the elevation of the neighboring cell, and calculate the second probability of whether the neighboring cell is on fire according to the elevation difference and a preset second fire probability calculation formula; determine the connecting line between the burning cell and the neighboring cell according to the position of the burning cell and the position of the neighboring cell, and determine whether the connecting line intersects with each simulated blocking band vector to obtain a blocking band determination result; determine the fire state corresponding to the neighboring cell according to the first probability, the second probability and the blocking band determination result.

[0250] In some embodiments, the above-mentioned fire simulation calculation unit is specifically used to determine, in a preset first fire function, a simulated temperature influence factor corresponding to the simulated temperature corresponding to the forest area; the above-mentioned first fire function includes different temperature ranges and temperature influence factors corresponding to each temperature range, and the above-mentioned temperature influence factors indicate the influence of the corresponding temperature range on the fire intensity; according to the simulated humidity corresponding to the forest area, a simulated humidity influence factor corresponding to the simulated humidity is determined in a preset second fire function; the above-mentioned second fire function includes different humidity ranges and humidity influence factors corresponding to each humidity range, and the above-mentioned humidity influence factors indicate the influence of the corresponding humidity range on the fire intensity; and according to the product of the simulated temperature influence factor and the simulated humidity influence factor, the fire intensity corresponding to the neighbor cell is obtained.

[0251] In some embodiments, the above-mentioned device may also include a grid division module, which is used to obtain vector data of forest areas and merge forest areas into one forest area based on the vector data; determine the grid level for dividing the forest area based on the area of ​​the forest area, and perform hexagonal grid division on the forest area according to the grid level to obtain multiple hexagonal grids corresponding to the forest area; calculate the position of the center point corresponding to each hexagonal grid based on the hexagonal vector corresponding to each hexagonal grid; determine the hexagonal grids corresponding to each center point as individual cells, and determine the position of each center point as the position corresponding to each cell.

[0252] In some embodiments, the above-mentioned grid data also includes the radius of the hexagonal grid corresponding to each cell, and the above-mentioned device may also include a spread speed simulation module, which is used to calculate the distance between the burning cell and one of its adjacent neighbor cells according to the radius of the hexagonal grid corresponding to the cell, and use the distance as an iterative spread distance; simulate and calculate the unit fire spread speed of the forest area according to the iterative spread distance and the preset single fire spread time; according to the type corresponding to the forest area, determine the simulated windless spread speed and simulated wind influence coefficient corresponding to the forest area in the preset speed function; the above-mentioned speed function includes windless spread speeds and wind influence coefficients corresponding to different types of forest areas, and the above-mentioned wind influence coefficient corresponds to the wind force level; according to the above-mentioned simulated windless spread speed and simulated wind influence coefficient, calculate the simulated fire spread speed of the forest area, and according to the simulated fire spread speed and the unit fire spread speed, calculate the simulated multiple of the fire spread speed of the forest area.

[0253] It should be noted here that the above-mentioned device provided in the embodiment of the present invention can implement all the method steps implemented in the above-mentioned method embodiment, and can achieve the same technical effect. The parts and beneficial effects that are the same as the method embodiment in this embodiment will not be described in detail here.

[0254] Fig. 9 An example of a physical structure diagram of an electronic device is shown in FIG. Fig. 9As shown, the electronic device may include: a processor 410, a communications interface 420, a memory 430 and a communications bus 440, wherein the processor 410, the communications interface 420 and the memory 430 communicate with each other through the communications bus 440. The processor 410 may call the logic instructions in the memory 430 to execute a forest fire spread simulation method, the method comprising: determining a fire point in a preset forest area, and determining an initial cell corresponding to the fire point in the grid data according to the location of the fire point and the grid data corresponding to the forest area; the grid data comprises a plurality of cells constituting the forest area and the location of each cell; determining an initial on-fire cell set according to the initial cells; the initial on-fire cell set comprises at least one on-fire cell whose on-fire state is on-fire; executing a fire simulation calculation operation, the fire simulation calculation operation comprising: for the initial on-fire cell For each burning cell in the set, according to the simulated environment parameters and simulated terrain parameters corresponding to the burning cell and the forest area, a preset cellular automaton model is used to simulate and calculate the fire state and fire intensity corresponding to the neighboring cells of the burning cell, and the target neighboring cells of the burning cell are put into the initial burning cell set to obtain a new initial burning cell set; the above-mentioned fire state indicates whether the neighboring cells are on fire; according to the new initial burning cell set, the above-mentioned fire simulation calculation operation is iteratively performed until the preset iteration cutoff condition is met, and the new initial burning cell set that meets the iteration cutoff condition is determined as the target burning cell set.

[0255] In addition, the logic instructions in the above-mentioned memory 430 can be implemented in the form of a software functional unit and can be stored in a computer-readable storage medium when it is sold or used as an independent product. Based on this understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art or the part of the technical solution, can be embodied in the form of a software product, and the computer software product is stored in a storage medium, including a number of instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present invention. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), disk or optical disk, etc. Various media that can store program codes.

[0256] On the other hand, the present invention also provides a computer program product, which includes a computer program, which can be stored on a non-transitory computer-readable storage medium. When the computer program is executed by a processor, the computer can execute the forest fire spread simulation method provided by the above-mentioned methods, and the method includes: determining a fire point in a preset forest area, and determining an initial cell corresponding to the fire point in the grid data based on the location of the fire point and the grid data corresponding to the forest area; the above-mentioned grid data includes multiple cells constituting the forest area and the location of each cell; determining an initial burning cell set based on the initial cells; the above-mentioned initial burning cell set includes at least one burning cell with a burning state of being on fire ; Execute a fire simulation calculation operation, the above-mentioned fire simulation calculation operation includes: for each burning cell in the initial burning cell set, according to the simulated environment parameters and simulated terrain parameters corresponding to the burning cell and the forest area, use a preset cellular automaton model to simulate and calculate the fire state and fire intensity corresponding to the neighbor cells of the burning cell, and put the target neighbor cells of the burning cell into the initial burning cell set to obtain a new initial burning cell set; the above-mentioned fire state indicates whether the neighbor cells are on fire; according to the new initial burning cell set, iteratively execute the above-mentioned fire simulation calculation operation until the preset iteration cutoff condition is met, and determine the new initial burning cell set that meets the iteration cutoff condition as the target burning cell set.

[0257] On the other hand, the present invention also provides a non-transitory computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, is implemented to execute the forest fire spread simulation method provided by the above-mentioned methods, the method comprising: determining a fire point in a preset forest area, and determining an initial cell corresponding to the fire point in the grid data according to the location of the fire point and the grid data corresponding to the forest area; the above-mentioned grid data includes a plurality of cells constituting the forest area and the location of each cell; determining an initial on-fire cell set according to the initial cells; the above-mentioned initial on-fire cell set includes at least one on-fire cell whose on-fire state is on-fire; performing a fire simulation calculation operation, the above-mentioned on-fire cell set comprising at least one on-fire cell whose on-fire state is on-fire; The fire simulation calculation operation includes: for each burning cell in the initial burning cell set, according to the simulated environment parameters and simulated terrain parameters corresponding to the burning cell and the forest area, a preset cellular automaton model is used to simulate and calculate the fire state and fire intensity corresponding to the neighbor cells of the burning cell, and the target neighbor cells of the burning cell are placed into the initial burning cell set to obtain a new initial burning cell set; the above-mentioned fire state indicates whether the neighbor cells are on fire; according to the new initial burning cell set, the above-mentioned fire simulation calculation operation is iteratively executed until the preset iteration cutoff condition is met, and the new initial burning cell set that meets the iteration cutoff condition is determined as the target burning cell set.

[0258] The device embodiments described above are merely illustrative, wherein the units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they may be located in one place, or they may be distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the scheme of this embodiment. Ordinary technicians in this field can understand and implement it without paying creative labor.

[0259] Through the description of the above implementation methods, those skilled in the art can clearly understand that each implementation method can be implemented by means of software plus a necessary general hardware platform, and of course, can also be implemented by hardware. Based on this understanding, the above technical solution is essentially or the part that contributes to the prior art can be embodied in the form of a software product, and the computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, a disk, an optical disk, etc., including a number of instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in each embodiment or some parts of the embodiments.

[0260] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for simulating the spread of forest fire, characterized in that: include: Determine a fire point in a preset forest area, and determine an initial cell corresponding to the fire point in the grid data according to the location of the fire point and the grid data corresponding to the forest area; the grid data includes a plurality of cells constituting the forest area and the location of each cell; Determine an initial ignition cell set according to the initial cells; The initial set of ignited cells includes at least one ignited cell whose ignition state is ignited; Executing a fire simulation calculation operation, the fire simulation calculation operation comprising: for each of the on-fire cells in the initial on-fire cell set, according to the on-fire cell, the simulated environment parameters and the simulated terrain parameters corresponding to the forest area, using a preset cellular automaton model to simulate and calculate the fire state and fire intensity corresponding to the neighbor cells of the on-fire cell, and putting the target neighbor cells of the on-fire cell into the initial on-fire cell set to obtain a new initial on-fire cell set; the fire state indicates whether the neighbor cell is on fire; Iteratively executing the fire simulation calculation operation according to the new initial fire cell set until a preset iteration cutoff condition is satisfied, and determining the new initial fire cell set satisfying the iteration cutoff condition as the target fire cell set; The grid data also includes a grid code corresponding to each of the cells. The simulation environment parameters and the simulation terrain parameters corresponding to the burning cells and the forest area are used to simulate and calculate the fire state and fire intensity corresponding to the neighboring cells of the burning cells using a preset cellular automaton model, including: According to the grid code of the burning cell and the grid codes of each of the cells, at least one neighboring cell adjacent to the burning cell is determined in the grid data; wherein the grid codes of the adjacent multiple cells are associated with each other; If any neighbor cell of the on-fire cell is not in the initial on-fire cell set, then any neighbor cell is placed in a potential on-fire cell set; the potential on-fire cell set includes at least one potential on-fire neighbor cell; Acquire simulated environmental parameters and simulated terrain parameters corresponding to the forest area; the simulated environmental parameters include simulated temperature, simulated humidity and simulated wind direction corresponding to the forest area, and the simulated terrain parameters include simulated terrain data and simulated blocking zone data corresponding to the forest area; For each of the neighbor cells in the potential fire cell set, the cellular automaton model is used to simulate and calculate the fire state corresponding to the neighbor cell according to the simulated wind direction, simulated terrain data and simulated blocking zone data of the forest area, and the cellular automaton model is used to simulate and calculate the fire intensity corresponding to the neighbor cell according to the simulated temperature and simulated humidity of the forest area.

2. The forest fire spread simulation method according to claim 1, characterized in that: The blocking zone data includes a plurality of simulated blocking zone vectors for blocking fire, and the cellular automaton model is used to simulate and calculate the fire state corresponding to the neighboring cells according to the simulated wind direction, simulated terrain data and simulated blocking zone data of the forest area, including: Calculating the azimuth from the burning cell to the neighboring cell according to the position of the burning cell and the position of the neighboring cell, and calculating the angle between the azimuth and the simulated wind direction according to the azimuth and the simulated wind direction of the forest area, and calculating the first probability of whether the neighboring cell is on fire according to the angle and a preset first fire probability calculation formula; According to the position of the burning cell and the position of the neighboring cells, the elevation of the burning cell and the elevation of the neighboring cells are obtained from the elevation data of the forest area, and the elevation difference between the elevation of the burning cell and the elevation of the neighboring cells is calculated, and according to the elevation difference and a preset second fire probability calculation formula, a second probability of whether the neighboring cell is on fire is calculated; According to the position of the ignited cell and the position of the neighboring cells, a connection line between the ignited cell and the neighboring cells is determined, and it is judged whether the connection line and each of the simulated blocking band vectors intersect to obtain a blocking band determination result; The fire state corresponding to the neighbor cell is determined according to the first probability, the second probability and the blocking band determination result.

3. The forest fire spread simulation method according to claim 1, characterized in that: The method of using the cellular automaton model to simulate and calculate the fire intensity corresponding to the neighboring cells according to the simulated temperature and simulated humidity of the forest area includes: According to the simulated temperature corresponding to the forest area, a simulated temperature influence factor corresponding to the simulated temperature is determined in a preset first ignition function; the first ignition function includes different temperature ranges and a temperature influence factor corresponding to each temperature range, and the temperature influence factor indicates the influence of the corresponding temperature range on the fire intensity; According to the simulated humidity corresponding to the forest area, determining a simulated humidity influence factor corresponding to the simulated humidity in a preset second ignition function; the second ignition function includes different humidity ranges and humidity influence factors corresponding to each humidity range, and the humidity influence factor represents the influence of the corresponding humidity range on the fire intensity; The fire intensity corresponding to the neighbor cell is obtained according to the product of the simulated temperature influence factor and the simulated humidity influence factor.

4. The method for simulating the spread of forest fire according to any one of claims 1 to 3, characterized in that: The method further comprises: Acquire vector data of the forest area, and merge the forest areas into one forest area according to the vector data; Determine a grid level for dividing the forest area according to the area of ​​the forest area, and perform hexagonal grid division processing on the forest area according to the grid level to obtain a plurality of hexagonal grids corresponding to the forest area; Calculating the position of the center point corresponding to each of the hexagonal grids according to the hexagonal vectors corresponding to each of the hexagonal grids; The hexagonal grids corresponding to the central points are determined as cells, and the positions of the central points are determined as the positions corresponding to the cells.

5. The method for simulating the spread of forest fire according to any one of claims 1 to 3, characterized in that: The grid data also includes the radius of the hexagonal grid corresponding to each cell, and the method further includes: According to the radius of the hexagonal grid corresponding to the cell, the distance between the burning cell and one of its adjacent neighbor cells is calculated, and the distance is used as an iterative spreading distance; According to the one-time iterative spread distance and the preset single fire spread time, the unit fire spread speed of the forest area is simulated and calculated; According to the type of the forest area, the simulated windless spreading speed and the simulated wind influence coefficient corresponding to the forest area are determined in a preset speed function; the speed function includes windless spreading speeds and wind influence coefficients corresponding to different types of forest areas, and the wind influence coefficient corresponds to the wind force level; The simulated fire spread speed in the forest area is calculated based on the simulated windless spread speed and the simulated wind influence coefficient, and the simulated multiple of the fire spread speed in the forest area is calculated based on the simulated fire spread speed and the unit fire spread speed.

6. The method for simulating the spread of forest fire according to any one of claims 1 to 3, characterized in that: The iteration cutoff condition includes any one of the following: The burning cell in the current iteration round has no neighbor cells; The current number of iterations reaches the preset number of iterations.

7. A forest fire spread simulation device, characterized in that: include: An initial cell determination module is used to determine a fire point in a preset forest area, and determine an initial cell corresponding to the fire point in the grid data according to the location of the fire point and the grid data corresponding to the forest area; the grid data includes a plurality of cells constituting the forest area and the location of each cell; An initial ignition cell set determination module, used to determine an initial ignition cell set according to the initial cell; the initial ignition cell set includes at least one ignition cell whose ignition state is ignition; A fire simulation calculation module is used to perform a fire simulation calculation operation, wherein the fire simulation calculation operation includes: for each of the on-fire cells in the initial on-fire cell set, according to the simulated environment parameters and simulated terrain parameters corresponding to the on-fire cell and the forest area, a preset cellular automaton model is used to simulate and calculate the fire state and fire intensity corresponding to the neighbor cells of the on-fire cell, and the target neighbor cells of the on-fire cell are placed into the initial on-fire cell set to obtain a new initial on-fire cell set; the fire state indicates whether the neighbor cell is on fire; An iteration module, configured to iteratively execute the fire simulation calculation operation according to the new initial fire cell set until a preset iteration cutoff condition is satisfied, and determine the new initial fire cell set satisfying the iteration cutoff condition as the target fire cell set; The grid data also includes a grid code corresponding to each of the cells, and the fire simulation calculation module includes: A neighbor cell determination unit, configured to determine at least one neighbor cell adjacent to the fire cell in the grid data according to the grid code of the fire cell and the grid codes of each of the cells; wherein the grid codes of the adjacent multiple cells are associated with each other; A potential fire cell set determination unit, configured to place any neighbor cell of the fire cell into a potential fire cell set if the neighbor cell is not in the initial fire cell set; the potential fire cell set includes at least one potential fire neighbor cell; A parameter acquisition unit, used to acquire simulated environmental parameters and simulated terrain parameters corresponding to the forest area; the simulated environmental parameters include simulated temperature, simulated humidity and simulated wind direction corresponding to the forest area, and the simulated terrain parameters include simulated terrain data and simulated blocking zone data corresponding to the forest area; The fire simulation calculation unit is used to simulate and calculate the fire state corresponding to each neighbor cell in the potential fire cell set using the cellular automaton model according to the simulated wind direction, simulated terrain data and simulated blocking zone data of the forest area, and to simulate and calculate the fire intensity corresponding to the neighbor cell using the cellular automaton model according to the simulated temperature and simulated humidity of the forest area.

8. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein: When the processor executes the computer program, the forest fire spread simulation method according to any one of claims 1 to 6 is implemented.

9. A non-transitory computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the forest fire spread simulation method according to any one of claims 1 to 6 is implemented.

Images