A forest fire spreading real-time simulation method, fire extinguishing decision method and early warning system
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- ZHEJIANG UNIV
- Filing Date
- 2023-02-16
- Publication Date
- 2026-05-29
AI Technical Summary
Existing forest fire spread simulation models are insufficient in terms of accuracy and real-time performance, and forest fire decision-making methods fail to effectively consider future spread scenarios, leading to difficulties in forest fire rescue and loss assessment.
An improved forest fire spread model was adopted, which combined the shortest path algorithm and the boundary extension method. Considering factors such as terrain, vegetation, wind direction and wind speed, a grid network simulation was carried out, and the optimal fire extinguishing point and firebreak location were determined by directed graph and topological sorting.
It improves the accuracy and real-time performance of forest fire spread simulation, enabling rapid calculation of fire extent and ignition time, providing direct and indirect firefighting decisions, reducing redundant calculation costs, and helping rescue personnel develop effective firefighting and evacuation strategies.
Smart Images
Figure CN116305832B_ABST
Abstract
Description
Technical Field
[0001] This invention belongs to the field of computer simulation technology, specifically relating to a real-time simulation method for forest fire spread, a fire extinguishing decision-making method, and an early warning system. Background Technology
[0002] Forest fires are natural disasters characterized by their suddenness, destructiveness, and difficulty in handling and rescue. Their randomness and unpredictability cause significant harm and loss to forest ecosystems and the safety of people's lives and property, and may even damage forest structure and the forest environment. Simulating the spread of forest fires can help people understand the fire's trajectory, enabling them to conduct appropriate firefighting and rescue operations and evacuate residents, thereby reducing the losses caused by forest fires.
[0003] Existing wildfire spread models can be divided into empirical models and physical models. Empirical models are statistical models that analyze patterns based solely on historical fire data. Their results are heavily dependent on historical factors such as terrain and vegetation distribution, making them difficult to apply to new terrains and resulting in low robustness. Physical models, on the other hand, simulate the physical and chemical processes of wildfires. However, the complex physicochemical processes involved in combustion make the simulation and verification of these models extremely difficult, leading to low simulation accuracy and a lack of real-time capability.
[0004] A commonly used approach to simulate wildfire spread is the cellular automata model. Cellular automata are discrete-time, spatial, and state-based grid dynamics models where spatial interactions and temporal causality are localized. By abstracting a spatial grid network into individual cells, the evolutionary spread of wildfires is simulated. Existing cellular automata-based methods meet real-time requirements, but their accuracy is insufficient. Furthermore, current wildfire decision-making methods primarily use shortest path algorithms to determine the current wildfire spread situation, without considering future wildfire development. Summary of the Invention
[0005] In response to this, this application provides a method to improve the simulation accuracy of forest fire spread, proposes to assist in fire fighting based on the future forest fire spread situation, and constructs a forest fire spread simulation and early warning system based on these methods.
[0006] The specific technical solution adopted in this invention is as follows:
[0007] In a first aspect, the present invention provides a real-time simulation method for forest fire spread, comprising:
[0008] S1. Obtain the three types of distribution data of slope aspect, slope and combustible type in the forest fire spread simulation area, as well as real-time wind direction, real-time wind speed and ignition point coordinates. After rasterizing the three types of distribution data, establish the raster network corresponding to the forest fire spread simulation area.
[0009] S2. Locate the ignition point grid in the grid network based on the ignition point coordinates. Simulate the spread of wildfire by gradually expanding the grid network from the ignition point grid. Simulate the advancement of the fire area based on the shortest path algorithm and the boundary extension method, and obtain the ignition time of each grid in the grid network. The spread rate of each ignition grid to any neighboring grid is calculated by the wildfire spread model based on the slope aspect, slope, combustible material type, real-time wind direction and real-time wind speed of the grid.
[0010] S3. Visualize the fire time of each grid in the grid network to display the fire simulation map at different times during the simulation period.
[0011] As a preferred embodiment of the first aspect above, the method for simulating the progression of the fire area during the gradual forest fire spread simulation, based on the shortest path algorithm and the boundary extension method, is as follows:
[0012] S201. Before performing the first step of the forest fire spread simulation, set the ignition time of the ignition point grid to the specified initial time and add the ignition point grid to the priority queue that is initially empty.
[0013] S202. Take the grid with the earliest fire time from the current priority queue, delete it from the priority queue and add it to the result set, and use it as the extension point of the fire area boundary.
[0014] S203. Traverse each neighboring grid in the eight neighborhoods of the expansion point that does not exist in the result set. For each current neighboring grid, firstly, the forest fire spread model calculates the spread rate of the forest fire from the expansion point to the current neighboring grid based on the slope aspect, slope, combustible material type, real-time wind direction, and real-time wind speed corresponding to the current neighboring grid. Then, it calculates the time required for the forest fire to spread from the expansion point to the current neighboring grid based on the spatial distance from the expansion point to the current neighboring grid. Finally, it obtains the ignition time of the current neighboring grid based on the ignition time of the expansion point itself. If the ignition time of the current neighboring grid is greater than the end time of the simulation period, the current neighboring grid is not added to the priority queue. Otherwise, it is necessary to determine whether the current neighboring grid already exists in the priority queue. If it does not exist, the expansion point is used as the ignition grid of the current neighboring grid, and the current neighboring grid is added to the priority queue and the corresponding ignition time is recorded. If it exists, the ignition time of this grid in the priority queue is updated to the earlier time between the calculated value of the current step size and the original recorded value, and the recorded ignition grid is updated.
[0015] S204. If there are still grids in the priority queue, S202 and S203 are iterated and repeated for the next long simulation until there are no grids in the priority queue. The forest fire spread simulation is then completed. The final result records all the grids that caught fire and their corresponding ignition times during the entire simulation period.
[0016] As a preferred embodiment of the first aspect above, the forest fire spread model is an improved Wang Zhengfei model, wherein the formula for calculating the spread rate of forest fire from the burned grid cell at the spread point to neighboring grid cells is:
[0017]
[0018] Where R0 is the initial velocity of forest fire spread; Ks is the combustible coefficient determined by the combustible type of the neighboring grid; θ1 is the real-time wind direction; and θ2 is the slope aspect of the neighboring grid. α represents the slope of the neighboring grid, α represents the direction of spread from the burned grid to the neighboring grid, and v represents the real-time wind speed; the real-time wind direction, slope aspect, and direction of spread are all 0 degrees north and the angles are calculated clockwise.
[0019] Secondly, the present invention provides a fire extinguishing decision-making method, which first obtains a result set recording the ignition time of each grid in the grid network according to the real-time forest fire spread simulation method described in the first aspect, and then selects one of the two methods, direct fire extinguishing and indirect fire extinguishing, to make a fire extinguishing decision.
[0020] The direct fire suppression method is used to determine the optimal set of fire suppression points required to establish a direct fire suppression line in the simulated forest fire spread area. The determination method is shown in A1 to A4:
[0021] A1. Form a directed graph by arranging all the grids in the result set according to the relationship between the ignited grid and the igniting grid. After setting an initial cost value for each grid, calculate the in-degree of each grid based on the directed graph, and add all grids with an in-degree of 0 to the initially empty grid queue.
[0022] A2. Take a grid from the grid queue and delete it from the grid queue. Add the value of the currently taken grid to the corresponding ignition grid to update the value of the ignition grid. At the same time, decrement the in-degree of the ignition grid by 1. If the in-degree of the ignition grid is 0 after decrementing by 1, add the ignition grid to the grid queue.
[0023] A3. If there are still grids in the grid queue, S42 is iterated and repeated until there are no grids in the grid queue, and the iteration is completed to obtain the result set after the cost value is updated.
[0024] A4. For the result set after the value is updated, select a specified number of the best fire extinguishing points that meet the fire extinguishing time limit in descending order of value; the fire extinguishing time limit is that the firefighters can arrive at the fire extinguishing point before it is ignited and complete the fire extinguishing isolation zone setting in advance.
[0025] The indirect fire extinguishing method is used to determine the optimal location for setting up a firebreak to prevent the spread of forest fire in the simulated forest fire area at the current moment. The determination method is shown in B1 to B4:
[0026] B1. Get the current time;
[0027] B2. Determine the main direction of forest fire spread based on the latest fire simulation map at the end of the simulation.
[0028] B3. Simulate and set up a firebreak of a preset length along the main direction of forest fire spread determined in B2, and gradually lengthen the firebreak from far to near the fire line in the current fire simulation map. Determine the optimal position closest to the fire line and where the forest fire has not yet spread to the firebreak when the firebreak is completed. Simulate and build a firebreak at the optimal position and set the forest fire spread rate coefficient at the firebreak to 0. Then, start the forest fire spread simulation again from the current moment to obtain a new fire simulation map after the current moment.
[0029] B4. Based on the new fire scene simulation map obtained in B3, determine whether the current firebreaks meet the requirements for blocking the spread of wildfire. If they do, output all the current firebreaks and their locations. If they do not meet the requirements, re-execute B2 and B3 based on the new fire scene simulation map until the requirements for blocking the spread of wildfire are met, and then output all the current firebreaks and their locations.
[0030] As a preferred embodiment of the second aspect above, in A1, the initial cost value of each grid in the result set is set according to the importance of the location space of the grid. The cost value of the grid in the location space without important facilities is set to 1 by default, while the cost value of the grid in the location space with important facilities is increased.
[0031] As a preferred embodiment of the second aspect above, in A4, the fire extinguishing time is limited to satisfy the following formula:
[0032] t_arrive≤node_igniteTime-Δt
[0033] In the formula: t_arrive represents the time required for firefighters to reach a fire extinguishing point, node_igniteTime represents the ignition time when the fire extinguishing point is ignited, and Δt represents the reserved time for firefighters to set up the fire extinguishing isolation zone after they arrive at the fire extinguishing point.
[0034] As a preferred embodiment of the second aspect above, in A4, the time required for firefighters to reach a fire extinguishing point is calculated based on the shortest path planning algorithm combined with road network information.
[0035] Thirdly, the present invention provides a forest fire spread simulation and early warning system, which includes a resource management module, a spread simulation module, a monitoring and early warning module, and a visualization module;
[0036] The resource management module is used to manage the input parameters, user information resources, and simulation results required for forest fire spread simulation, and to provide corresponding resource permissions for each user.
[0037] The spread simulation module is used to simulate forest fire spread in a designated forest fire spread simulation area according to the aforementioned simulation method based on the input parameters received by the resource management module, and is responsible for providing forest fire spread simulation results to other modules;
[0038] The monitoring and assessment module is used to detect the fire situation in real time through monitoring equipment, output fire extinguishing decision results according to the aforementioned fire extinguishing decision method to assist in the dispatch of fire extinguishing resources, and conduct loss assessment and fire level classification based on the forest fire spread simulation results.
[0039] The visualization module is used to dynamically visualize the fire scene at different times during the simulation period based on the ignition time of each grid in the grid network.
[0040] As a preferred embodiment of the third aspect mentioned above, in the spread simulation module, the forest fire spread simulation area and the ignition point are selected through the GIS map visualization interface provided by the visualization module.
[0041] As a preferred embodiment of the third aspect mentioned above, it also includes a fire-fighting resource scheduling decision module, which is used to assist in formulating fire-fighting resource allocation and fire-fighting resource scheduling strategies. The resource allocation is based on the assessed fire level to prepare resources, and the fire-fighting resource scheduling strategy is to use the shortest path algorithm to schedule the required resources to the fire-fighting point in the way with the shortest distance and the fastest arrival time.
[0042] Compared with the prior art, the beneficial effects of the present invention are as follows:
[0043] 1) This invention provides an improved forest fire spread algorithm that takes into account topographic factors (height, aspect, slope), vegetation distribution, wind direction, and wind speed, further improving the accuracy of traditional cellular automata models. In addition, the algorithm adopts the shortest path algorithm and the idea of boundary extension to quickly calculate the earliest ignition time of each ignition point, reducing the large amount of repetitive calculation overhead of existing methods. At the same time, based on the ignition time series results, the evolution process of forest fire spread can be quickly plotted and visualized.
[0044] 2) This invention proposes methods for assisting in direct and indirect fire suppression decisions based on a forest fire spread model. For direct fire suppression, the cost value of each ignition point is calculated based on the sequence of ignition points obtained from the forest fire spread. Then, the shortest path algorithm selects ignition points with high cost value that can be reached and extinguished in a timely manner. For indirect fire suppression, the method involves continuously approaching the edge of the fire spread in the direction of rapid fire spread to search for suitable firebreaks. Furthermore, the cost value calculation of ignition points in direct fire suppression is completed through topological sorting, the main purpose of which is to calculate the total loss value of all subsequent ignition points ignited by a certain ignition point. The initial loss value of each ignition point varies depending on its vegetation type, vegetation area, presence of important facilities, and presence of residents. Furthermore, the method for assisting in the determination of firebreaks in indirect fire suppression mainly aims to find the location of firebreaks that can be quickly established while minimizing the area of fire spread, thereby reducing the economic losses caused by the spread.
[0045] 3) The forest fire spread and early warning system provided by this invention provides functions such as GIS resource management, forest fire spread simulation, real-time fire detection, fire assessment and early warning, and fire extinguishing auxiliary decision-making. It can help rescuers to better carry out fire extinguishing planning and mass evacuation work in many ways. Attached Figure Description
[0046] Figure 1 This is a schematic diagram showing the direction of forest fire spread in neighboring areas.
[0047] Figure 2 A schematic diagram of the wind coefficient calculation projection.
[0048] Figure 3 This is a schematic diagram of the projection for calculating the aspect coefficient.
[0049] Figure 4 This is a flowchart of the forest fire spread based on the shortest path and the outer edge of the boundary.
[0050] Figure 5 This is a dynamic rendering based on the time series results of a forest fire spread simulation.
[0051] Figure 6 A schematic diagram illustrating the spread of the forest fire.
[0052] Figure 7 A schematic diagram showing the ignition path of the fire.
[0053] Figure 8 A schematic diagram to aid decision-making regarding firebreaks.
[0054] Figure 9 This is a functional module diagram of a forest fire spread simulation and early warning platform.
[0055] Figure 10 This is a system architecture diagram for a forest fire spread simulation and early warning platform. Detailed Implementation
[0056] To make the above-mentioned objects, features, and advantages of the present invention more apparent and understandable, the specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings. Many specific details are set forth in the following description to provide a thorough understanding of the present invention. However, the present invention can be practiced in many other ways different from those described herein, and those skilled in the art can make similar modifications without departing from the spirit of the present invention. Therefore, the present invention is not limited to the specific embodiments disclosed below. Technical features in the various embodiments of the present invention can be combined accordingly without mutual conflict.
[0057] This invention first provides a method for real-time simulation of forest fire spread, the steps of which include:
[0058] S1. Obtain the distribution data of three types of forest fire spread simulation areas: slope aspect, slope gradient, and combustible material type, as well as real-time wind direction, real-time wind speed, and ignition point coordinates. After rasterizing the three types of distribution data, establish a raster network corresponding to the forest fire spread simulation area.
[0059] S2. Locate the ignition point grid in the grid network based on the ignition point coordinates. Simulate the spread of forest fire by gradually expanding the grid network from the ignition point grid. Simulate the advancement of the fire area based on the shortest path algorithm and the boundary extension method, and obtain the ignition time of each grid in the grid network. The spread rate of each ignition grid to any neighboring grid is calculated by the forest fire spread model based on the slope aspect, slope, combustible material type, real-time wind direction and real-time wind speed of the grid.
[0060] The following method, based on the shortest path algorithm and boundary extension method, is used to simulate the advancement of the fire area during the gradual forest fire spread simulation process:
[0061] S201. Before performing the first step of the forest fire spread simulation, set the ignition time of the ignition point grid to the specified initial time and add the ignition point grid to the priority queue that is initially empty.
[0062] S202. Take the grid with the earliest fire time from the current priority queue, delete it from the priority queue and add it to the result set, and use it as the extension point of the fire area boundary.
[0063] S203. Traverse each neighboring grid in the eight neighborhoods of the expansion point that does not exist in the result set. For each current neighboring grid, firstly, the forest fire spread model calculates the spread rate of the forest fire from the expansion point to the current neighboring grid based on the slope aspect, slope, combustible material type, real-time wind direction, and real-time wind speed corresponding to the current neighboring grid. Then, it calculates the time required for the forest fire to spread from the expansion point to the current neighboring grid based on the spatial distance from the expansion point to the current neighboring grid. Finally, it obtains the ignition time of the current neighboring grid based on the ignition time of the expansion point itself. If the ignition time of the current neighboring grid is greater than the end time of the simulation period, the current neighboring grid is not added to the priority queue. Otherwise, it is necessary to determine whether the current neighboring grid already exists in the priority queue. If it does not exist, the expansion point is used as the ignition grid of the current neighboring grid, and the current neighboring grid is added to the priority queue and the corresponding ignition time is recorded. If it exists, the ignition time of this grid in the priority queue is updated to the earlier time between the calculated value of the current step size and the original recorded value, and the recorded ignition grid is updated.
[0064] S204. If there are still grids in the priority queue, S202 and S203 are iterated and repeated for the next long simulation until there are no grids in the priority queue. The forest fire spread simulation is then completed. The final result records all the grids that caught fire and their corresponding ignition times during the entire simulation period.
[0065] S3. Visualize the fire time of each grid in the grid network to display the fire simulation map at different times during the simulation period.
[0066] The forest fire spread model used in this invention is an improved Wang Zhengfei model, wherein the formula for calculating the spread rate of forest fire from the burned grid cell at the spread point to neighboring grid cells is as follows:
[0067]
[0068] Where R0 is the initial velocity of forest fire spread; Ks is the combustible coefficient determined by the combustible type of the neighboring grid; θ1 is the real-time wind direction; and θ2 is the slope aspect of the neighboring grid. α represents the slope of the neighboring grid, α represents the direction of spread from the burned grid to the neighboring grid, and v represents the real-time wind speed; the real-time wind direction, slope aspect, and direction of spread are all 0 degrees north and the angles are calculated clockwise.
[0069] It should be noted that in this invention, each grid is used as a simulation unit, and the distance between grids can be calculated from their center points. However, during the spread of a forest fire, an already ignited grid can ignite surrounding grids. For ease of description, two grids with an ignition and being ignited relationship are referred to as the igniting grid and the ignited grid, respectively, with the ignition time of the igniting grid preceding that of the ignited grid.
[0070] In addition, the present invention also provides a fire extinguishing decision-making method, which first obtains a result set recording the ignition time of each grid in the grid network according to the aforementioned real-time simulation method of forest fire spread, and then selects one of the two methods, direct fire extinguishing and indirect fire extinguishing, to make a fire extinguishing decision.
[0071] A. The direct fire suppression method is used to determine the optimal set of fire suppression points required to establish a direct fire suppression line in the simulated forest fire spread area. The determination method is shown in A1 to A4:
[0072] A1. Form a directed graph by arranging all the grids in the result set according to the relationship between the ignited grid and the igniting grid. After setting an initial cost value for each grid, calculate the in-degree of each grid based on the directed graph, and add all grids with an in-degree of 0 to the initially empty grid queue.
[0073] The initial cost value of each grid in the above result set is set according to the importance of the location space of the grid. The cost value of the grid in the location space without important facilities is set to 1 by default, and the cost value of the grid in the location space with important facilities is increased.
[0074] A2. Take a grid from the grid queue and delete it from the grid queue. Add the value of the currently taken grid to the corresponding ignition grid to update the value of the ignition grid. At the same time, decrement the in-degree of the ignition grid by 1. If the in-degree of the ignition grid is 0 after decrementing by 1, add the ignition grid to the grid queue.
[0075] A3. If there are still grids in the grid queue, S42 is iterated and repeated until there are no grids in the grid queue, and the iteration is completed to obtain the result set after the value is updated.
[0076] A4. For the result set after the value is updated, select a specified number of the best fire extinguishing points that meet the fire extinguishing time limit in descending order of value; the fire extinguishing time limit is that the firefighters can arrive at the fire extinguishing point before it is ignited and complete the fire extinguishing isolation zone setting in advance.
[0077] The above-mentioned fire extinguishing time limit is subject to the following formula:
[0078] t_arrive≤node_igniteTime-Δt
[0079] In the formula: t_arrive represents the time required for a firefighter to reach a fire extinguishing point, node_igniteTime represents the ignition time at which the fire extinguishing point is ignited, and Δt represents the time allotted for setting up the firebreak after the firefighter arrives at the fire extinguishing point. Furthermore, the time required for a firefighter to reach a fire extinguishing point can be calculated using a shortest path planning algorithm combined with road network information.
[0080] B. The indirect fire extinguishing method is used to determine the optimal location for setting up a firebreak in the simulated forest fire area at the current moment to prevent the spread of forest fire. The determination method is shown in B1 to B4.
[0081] B1. Get the current time;
[0082] B2. Determine the main direction of forest fire spread based on the latest fire simulation map at the end of the simulation.
[0083] B3. Simulate and set up a firebreak of a preset length along the main direction of forest fire spread determined in B2, and gradually lengthen the firebreak from far to near the fire line in the current fire simulation map. Determine the optimal position closest to the fire line and where the forest fire has not yet spread to the firebreak when the firebreak is completed. Simulate and build a firebreak at the optimal position and set the forest fire spread rate coefficient at the firebreak to 0. Then, start the forest fire spread simulation again from the current moment to obtain a new fire simulation map after the current moment.
[0084] B4. Based on the new fire scene simulation map obtained in B3, determine whether the current firebreaks meet the requirements for blocking the spread of wildfire. If they do, output all the current firebreaks and their locations. If they do not meet the requirements, re-execute B2 and B3 based on the new fire scene simulation map until the requirements for blocking the spread of wildfire are met, and then output all the current firebreaks and their locations.
[0085] This invention also provides a forest fire spread simulation and early warning system, which includes a resource management module, a spread simulation module, a monitoring and early warning module, and a visualization module;
[0086] The resource management module is used to manage the input parameters, user information resources, and simulation results required for forest fire spread simulation, and to provide corresponding resource permissions for each user.
[0087] The spread simulation module is used to simulate forest fire spread in a designated forest fire spread simulation area according to the aforementioned simulation method based on the input parameters received by the resource management module, and is responsible for providing forest fire spread simulation results to other modules;
[0088] The monitoring and assessment module is used to detect the fire situation in real time through monitoring equipment, output fire extinguishing decision results according to the aforementioned fire extinguishing decision method to assist in the dispatch of fire extinguishing resources, and conduct loss assessment and fire level classification based on the forest fire spread simulation results.
[0089] The visualization module is used to dynamically visualize the fire situation based on the ignition time of each grid in the grid network, displaying simulated fire scenes at different times during the simulation period. In the aforementioned fire spread simulation module, the simulated fire spread area and ignition point are selected through the GIS map visualization interface provided by the visualization module.
[0090] The aforementioned forest fire spread simulation and early warning system may also include a fire-fighting resource scheduling decision module, which is used to assist in formulating fire-fighting resource allocation and fire-fighting resource scheduling strategies. The resource allocation is based on the assessed fire level to prepare resources, and the fire-fighting resource scheduling strategy is to use the shortest path algorithm to schedule the required resources to the fire-fighting point in the way with the shortest distance and fastest arrival time.
[0091] The following preferred embodiment will be used to further describe in detail the specific implementation examples of the real-time forest fire spread simulation method of the present invention.
[0092] Example
[0093] I. Real-time simulation of forest fire spread
[0094] This embodiment first optimizes the forest fire spread model algorithm. The specific forest fire spread simulation process, from input to simulation result output, includes the following steps:
[0095] (S1) Obtain input data and initialize the cell grid network.
[0096] The required basic input data and their format are shown in the table below:
[0097]
[0098] The terrain-related distribution data uses remote sensing image format as input, with the value of each pixel in the remote sensing image serving as the corresponding attribute value for the raster cell. Therefore, the boundary size of the raster data is consistent with the image size, and each raster includes a series of data. Here, S is defined as the raster data, S... i,j For a raster corresponding to a pixel, it can be represented as a vector:
[0099] S i,j ={sl,sld,h,t,it}
[0100] Where sl is the slope, sld is the aspect, h is the height, t is the type of combustible material, and it records the ignition time of this grid cell during the spread simulation. The slope, aspect, and type of combustible material stored in the above grid cells are used for subsequent calculations of the spread rate and ultimately to determine the ignition time it. Height information can be used to understand the terrain of the simulated area and assist in firefighting decisions.
[0101] The following section describes the calculation method for the direction and speed of forest fire spread in this embodiment.
[0102] First, there's the determination and representation of the direction of fire spread. Traditionally, forest fire spread is generally represented using the eight-neighborhood method, such as... Figure 1 As shown, the eight directions represent the expansion from the center point outwards to the north, northeast, east, southeast, south, southwest, west, and northwest. With the northward angle set at 0 degrees, and rotating clockwise, the radians in each of the eight directions are respectively...
[0103] In this embodiment, the spread rate is calculated using an improved model of Wang Zhengfei's forest fire spread model developed by Mao Xianmin et al. This improved model comprehensively considers the influence of terrain slope and wind direction on the spread rate. It modifies the speed model to include five scenarios: uphill, downhill, left-side level slope, right-side level slope, and wind direction. In reality, the slope aspect and wind direction are not perfectly parallel. Figure 1 The eight directions shown are those that will interact with... Figure 1 The eight spread directions in the model all have certain angles, making it difficult to directly calculate the slope influence factor and wind influence factor using the five-direction formula in the improved model. In this embodiment, the spread velocity formulas in the five directions are calculated by projecting the slope aspect and wind direction onto the spread direction, thus unifying the spread velocity formulas in the five directions of Wang Zhengfei's improved model. The principle and calculation method of the improved formula are given below.
[0104] In Wang Zhengfei's model, the formula for calculating the spread rate of forest fire from burned cells to neighboring cells is:
[0105]
[0106] Where R0 is the initial velocity (m / s) of the wildfire spread.
[0107] R0 = aT + bW + c(100 - RH) - d
[0108] W = int(v / 0.836) (2 / 3)
[0109] T represents temperature (°C); W represents wind speed; RH represents humidity (%); v represents wind speed (m / s).
[0110] Ks is the combustibility coefficient, which can be obtained from a table:
[0111] Combustible material type Ks Other (roads, and other non-combustible materials) 0 coniferous forest 0.4 Mixed coniferous and broad-leaved forest 0.7 broad-leaved forest 1.0 Grass 1.6
[0112] Kw is the wind coefficient, influenced by the angle between the wind direction and the spread direction. Spreading is faster with the wind and slower against the wind. The effects of wind direction and speed are related to the projection of the wind vector onto the spread direction. The wind direction projection is as follows: Figure 2 As shown, the formula for calculating the wind coefficient can be obtained by projecting the wind direction onto the opposite side of the spread:
[0113]
[0114] When the angle between the wind direction and the direction of spread is obtuse, the wind force will impede the speed in the direction of spread; when the angle is obtuse, the effect will be amplified.
[0115] Similarly, the aspect projection is as follows: Figure 3 As shown. The formula for calculating the slope coefficient is:
[0116]
[0117] Figure 3 The spread direction 1 is a right-hand horizontal slope. When calculating the slope aspect projection coefficient, because the included angle is a right angle... The angle between the direction of spread 2 and the slope aspect is acute, therefore... This is because a southward slope indicates that the southward spread is downhill, and the slope will slow down the spread.
[0118] Finally, the improved formula for calculating the spread rate obtained in this embodiment is:
[0119]
[0120] Where R0 is the initial velocity of forest fire spread; Ks is the combustible coefficient determined by the combustible type of the neighboring grid; θ1 is the real-time wind direction; and θ2 is the slope aspect of the neighboring grid. α represents the slope of the neighboring grid, α represents the direction of spread from the burned grid to the neighboring grid, and v represents the real-time wind speed; the real-time wind direction, slope aspect, and direction of spread are all 0 degrees north and the angles are calculated clockwise.
[0121] (S2) Simulation of Forest Fire Spread
[0122] After determining the input data and the calculation method for the spread direction, the forest fire spread simulation process can begin. In this embodiment, the forest fire spread simulation process based on the shortest path and boundary extension is as follows:
[0123] (1) Its ignition point is associated with the corresponding grid S starti,startj The ignition time it is set to 0, and S is set to 0.starti,startj Add it to the priority queue, which is initially empty, and the grid points in the priority queue are always sorted in ascending order of it;
[0124] (2) Take the first grid cell from the priority queue, that is, the grid cell with the smallest it in the priority queue, add it to the result set and delete it from the priority queue, and use it as the expansion point when extending the boundary of the fire area.
[0125] (3) For each neighboring grid node that does not exist in the result set in the neighborhood of the expansion point, the forest fire spread model first calculates the speed at which the forest fire spreads from the expansion point to the current neighboring grid node and the spatial distance from the expansion point to the node based on the slope aspect, slope, combustible material type, real-time wind direction, and real-time wind speed corresponding to the current neighboring grid. From this, the time required for the forest fire to spread from the expansion point to the node is obtained. Then, the ignition time it of the node is calculated by the ignition time it of the expansion point itself. Finally, the following judgment is made:
[0126] i. If it' is greater than the maximum time limit T, i.e. the end time of the simulation period, then abandon adding the node to the priority queue; otherwise, check whether the node already exists in the priority queue, and execute ii or iii according to the result.
[0127] ii. If node already exists in the priority queue and it' < it node Then update the ignition time of the corresponding mountain point in the queue. node For it', update the node's ignition grid to the current extension point;
[0128] iii. If it does not exist, add the node to the priority queue and update the grid's ignition time it. node =it', and simultaneously record the ignition grid of the node.
[0129] (4) If there are still grids in the priority queue, then jump back to step (2) to perform the next long simulation until there are no grids in the priority queue to complete the forest fire spread simulation. The final result records all the ignition grids and their corresponding ignition time and ignition grids during the entire simulation period.
[0130] The flowchart of the above forest fire spread algorithm is as follows: Figure 4 As shown.
[0131] (S3) Display the spread dynamics based on the time series of the result set.
[0132] In the above steps, the result set generated by the forest fire spread dynamic simulation process is itself a set containing all burned grids and their ignition time series. By reading this series and setting the time interval for displaying the dynamic simulation, grids within different time periods can be shown, thus reconstructing the spread dynamics. For example... Figure 5 As shown, this figure illustrates the dynamic forest fire spread process based on the forest fire spread simulation results set. The figure shows the forest fire spread process from t=10000s to t=40000s.
[0133] II. Assisted fire suppression decision-making based on forest fire spread simulation results
[0134] Firefighting strategies are generally divided into two types: direct firefighting and indirect firefighting. Direct firefighting refers to extinguishing the fire line directly using a combination of methods based on the fire situation, typically employing water or wind directly. This is suitable for fires of low intensity. Indirect firefighting is mainly used to deal with high-intensity fires, and the most common method currently is establishing firebreaks. This paper proposes methods for determining fire extinguishing points and firebreaks based on forest fire spread simulation trends, addressing both direct and indirect firefighting approaches.
[0135] First, this paper introduces a method that uses topological sorting combined with a forest fire spread model to determine optimal fire extinguishing points to assist direct fire suppression. The main objective is to prioritize fire extinguishing points that are likely to cause greater damage after ignition. (See also...) Figure 6 ( Figure 6 'a' represents the direction of spread. Figure 6 (b represents the backtracking direction). In the result set of the forest fire spread simulation, the coordinate information of the previous grid where each fire point was ignited is saved. Then, backtracking can be performed based on this information. The backtracking direction is opposite to the spread direction.
[0136] By outputting the backtracking path from multiple points, for reference Figure 7 The fire's spread path can be analyzed: it first spreads from the ignition point O to the branch point S, and then branches into three routes: OA, OB, and OC at S. An assumption here is that it's best to extinguish the fire before it reaches S; otherwise, after reaching S, the fire might spread rapidly in three directions (SA, SB, SC), increasing the difficulty of firefighting. A cost value is assigned to each grid point. Generally, the cost can be set to 1. For special geographical points with important facilities, the cost value can be increased, and the more important the facility, the higher the cost value can be. The cost value of the ignited grid is added to the neighboring grids along the backtracking path. Figure 7 For example, considering only the cost value P of the five points in the graph, the updated cost values of points S and O are calculated as follows:
[0137] Ps_new=Pa+Pb+Pc+Ps
[0138] Po_new=Po+Ps_new=Po+Ps+Pa+Pb+Pc
[0139] The cost value can be calculated using topological sorting, with a complexity of O(n), by calculating the cost value of each node from the result set.
[0140] (1) Form a directed graph from the relationship between the ignition point and the ignition point in the result set, and calculate the in-degree of each grid cell based on this graph;
[0141] (2) Set up a grid point queue Q = [all grid points with an in-degree of 0];
[0142] (3) Take a grid point node from Q and delete the grid point node from Q, and then perform the following operations on the taken grid point node:
[0143] a) Update the value of node_last by adding the value of node to the ignition grid node_last that ignited node. At the same time, decrement the in-degree of node_last by 1.
[0144] b) For node_last after step a), if the in-degree of node_last is 0, then add node_last to Q.
[0145] (4) If there are still grids in Q, repeat the above steps (3) iteratively until Q is empty and the iteration is completed, and the result set after the value is updated is obtained.
[0146] For a single ignition point O, its cost is the sum of the original costs of all points, which is also the maximum cost. Therefore, if the ignition point O is extinguished before it is ignited, the losses caused by the fire can be minimized. However, this is impossible, as the ignition time of point O is 0, and it is impossible for rescue personnel to arrive at the scene before that time for an unexpected forest fire.
[0147] Therefore, the optimal fire extinguishing point should be selected from the grid cells where the fire ignition time is later than the current time "now", that is, from the grid cells where "it > nowTime". Thus, the set of candidate optimal points is:
[0148] Candidate={s∈S|s_it>=now}
[0149] In the formula: S represents the result set after the value is updated.
[0150] A greedy algorithm is used to select the optimal fire extinguishing points from the above candidates. The principle is to filter out a specified number of the best fire extinguishing points that meet the fire extinguishing time limit, in descending order of cost. The fire extinguishing time limit is that firefighters can arrive at the fire extinguishing point before it ignites and complete the necessary firebreak setup in advance. The specific greedy algorithm is as follows:
[0151] (1) Set an empty set, i.e., Result = {}
[0152] (2) Select the grid point node with the highest cost and use the shortest path algorithm to calculate the shortest time t_arrive for the firefighters to reach that point based on the road network information.
[0153] (3) If t_arrive≤node_igniteTime-Δt (Δt is a reserved time for setting up an isolation zone around the node to prevent further spread, and node_igniteTime is the ignition time when the extinguishing point node is ignited), then add the node to the Result and delete the points in the Candidate that are ignited by the node.
[0154] (4) If Candidate is empty or size(result)>threshold, end the loop of the algorithm and output all the currently set isolation zones and their positions; otherwise, jump back to step (2) of this process.
[0155] like Figure 8 As shown, indirect fire suppression primarily involves establishing firebreaks. The results of wildfire spread simulations can help examine the feasibility of setting up firebreaks at a particular location. The main idea is to locate the firebreaks and ensure they are completed before the wildfire spreads to them; the process is roughly as follows:
[0156] (1) Get the current time now;
[0157] (2) Determine the main direction of forest fire spread based on the latest fire simulation diagram at the end of the simulation T, and then set up isolation zones along the direction perpendicular to the main direction of spread;
[0158] (3) Continuously move the firebreak from far to near the fire line on the current fire spread simulation map, and then execute the following procedure:
[0159] a) Calculate the completion time of the firebreak construction as t, obtain the simulated forest fire spread at time t, and determine whether the forest fire has spread to the firebreak at time t;
[0160] b) If the spread has already exceeded the isolation zone, stop approaching and use the location of the isolation zone from the previous simulation as the feasible isolation zone.
[0161] (4) Set the forest fire spread rate coefficient at the feasible isolation zone to 0, and regenerate the fire simulation map for the time period T>=now through forest fire spread simulation.
[0162] (5) Based on the new fire simulation diagram, determine whether the existing firebreaks are sufficient to meet the requirements for blocking the spread of forest fire, i.e. whether the forest fire will continue to spread after these firebreaks are set up. If not, new firebreaks need to be set up. Therefore, jump back to step (2) of this process. If it meets the requirements, no new firebreaks need to be set up. Output all the firebreaks that have been set up and their locations.
[0163] It should be noted that the main direction of forest fire spread needs to be determined based on the forest fire spread map. For example, it can be the direction with the longest spread distance from the starting point, or it can be the direction with the largest burned area. In addition, when it is necessary to re-simulate the forest fire spread in step (4) above, the real-time forest fire spread simulation process in the first part can be repeated.
[0164] In this embodiment, the three firebreaks to be established through indirect fire extinguishing decision-making are as follows: Figure 8 As shown.
[0165] III. Forest Fire Spread Simulation and Early Warning System
[0166] The following section introduces the structure and design of a forest fire spread simulation and early warning system based on the previous two parts. For example... Figure 9 The diagram shown is a functional block diagram of this system.
[0167] (1) Resource Management Module
[0168] like Figure 9 As shown, the resource management module includes several parts: GIS resource management, model management, simulation result management, and user information resource management.
[0169] GIS resource management is responsible for adding, deleting, querying, and modifying the GIS geographic information data required for forest fire spread simulation. This resource includes various terrain data (elevation, vegetation distribution, road networks, distribution of important facilities, distribution of settlements, etc.), and supports users to query data according to different regional ranges. It also supports maintenance and updating of GIS geographic information through layer editing and other methods.
[0170] Model management is used to store wildfire spread models and simulation results.
[0171] User information resource management supports user creation, password management, information management, etc. Each user can upload their own GIS data and has certain permission information management for the data.
[0172] Simulation result management refers to the saving and management of forest fire spread results, path planning results, and various result files calculated by these modules, such as spread simulation, detection and early warning, and visualization, providing operations such as viewing, deleting, and adding.
[0173] (2) Forest fire spread simulation module
[0174] The forest fire spread module is the main module of the system, which includes several functions such as parameter input, spread simulation, and optimal fire extinguishing path.
[0175] Parameter settings are primarily configured through the visualization module. For selecting the spread area, the visualization module first displays a GIS geographic layer. Users can then freely zoom the map and select the area to simulate the forest fire spread. The parameter input module retrieves necessary information from the resource management module within the required terrain area. Additionally, users need to input real-time wind direction, select the coordinates of the ignition point, and set the maximum simulation time.
[0176] The spread simulation involves establishing a corresponding forest fire spread model based on the input parameters, performing the simulation using the method described in this invention, obtaining a corresponding result set, saving the result set in conjunction with the resource management module, and visually displaying the spread dynamics based on the result set. The above-mentioned spread simulation is implemented according to the simulation method described in the first part.
[0177] (3) Forest fire spread monitoring and assessment module
[0178] This module is used to conduct fire risk assessments based on the results of forest fire spread simulations to assist in decision-making.
[0179] The first function is monitoring and early warning, which uses designated stations to detect fires in real time. Once a fire is detected, the system automatically uses geographic information from the corresponding area to simulate the spread of wildfires and evacuates residents in areas at risk of wildfire spread.
[0180] The second function, fire loss assessment, is to comprehensively evaluate factors such as the area of the burned area, the type and area of vegetation in the burned area, important facilities in the burned area, and the location of residents in the burned area. The assessment results are used to classify the fire intensity level and scale, and corresponding early warnings and the dispatch and allocation of fire-fighting resources are carried out based on this level.
[0181] The third function is the determination of the optimal fire extinguishing path. Based on the results of the spread simulation, the fire extinguishing decision results (optimal fire extinguishing point and firebreak) can be calculated according to the fire extinguishing decision method in Part II mentioned above, and the shortest path to the key locations in these fire extinguishing decision results can be determined. This can be used to assist in the decision-making of fire extinguishing resource allocation.
[0182] The fourth function is fire-fighting resource allocation decision-making, including fire-fighting resource allocation ratio and fire-fighting resource allocation strategy. Resource allocation ratio is mainly based on the fire assessment level for corresponding preparation. Resource allocation strategy is to use the shortest path algorithm to allocate the required resources to the fire-fighting point in the way with the shortest distance and fastest arrival time.
[0183] (4) Visualization module
[0184] The visualization module includes GIS resource visualization, fire spread dynamics visualization, and query result visualization. GIS resource visualization primarily presents GIS data such as slope, terrain, and vegetation distribution through visualized geographic layers. This allows for selecting areas of fire spread and setting ignition points on the layers, and also helps rescue personnel plan more effective fire routes. Fire spread dynamics visualization displays the fire's trajectory over time, allowing for the current location of the fire's head, tail, and flanks. Query result visualization allows users to query various data from resource management, such as GIS resources, user information resources, and historical simulation results, and then visualize them.
[0185] Figure 10 The following is a diagram of the platform's architecture, combined with... Figure 10 This will illustrate two important system processes.
[0186] (1) System flow of forest fire spread simulation
[0187] The process of the forest fire spread simulation system is as follows: Figure 10 As shown, the process involves a resource management module, a visualization module, and a forest fire spread module. First, the corresponding GIS resource is selected. Then, the visualization module displays the corresponding topographic map. Next, the user selects the area on the map where the forest fire spread will occur and sets the ignition point, while also entering other necessary parameters (wind direction, wind speed, etc.). The area range information, along with other parameters, is passed to the forest fire spread simulation module. This module reads the topographic, vegetation, and other parameter data within the corresponding area and performs a forest fire spread simulation. After the simulation, the results are passed to the visualization module for dynamic display. Simultaneously, the results are saved as a JSON file through the resource management module as a log entry for future reference. The format of the forest fire spread ignition point result set is as follows:
[0188]
[0189]
[0190] (2) System process for monitoring, evaluation and early warning
[0191] The system process for monitoring and evaluation is as follows: Figure 10As shown, the monitoring and assessment module continuously monitors the relevant area. Once a fire is detected, it automatically retrieves the GIS resource information of the corresponding location through the resource management module and then performs a forest fire spread simulation. After the simulation, the results do not need to be visualized; they only need to be saved as a log entry. This result can then be queried through the resource management module to view the expected dynamics of the forest fire spread. In addition, the system automatically assesses fire damage based on the burned area from the forest fire spread simulation results, classifies the fire level according to the damage and fire area, and issues warnings and evacuations to residents in the affected area. It also generates corresponding fire suppression dispatch strategies based on the fire level, such as suggested fire extinguishing points for direct fire suppression or firebreak schemes for indirect fire suppression.
[0192] The embodiments described above are merely preferred embodiments of the present invention and are not intended to limit the invention. Those skilled in the art can make various changes and modifications without departing from the spirit and scope of the invention. Therefore, all technical solutions obtained through equivalent substitution or transformation fall within the protection scope of the present invention.
Claims
1. A method for real-time simulation of forest fire spread, characterized in that, include: S1. Obtain the three types of distribution data of slope aspect, slope and combustible type in the forest fire spread simulation area, as well as real-time wind direction, real-time wind speed and ignition point coordinates. After rasterizing the three types of distribution data, establish the raster network corresponding to the forest fire spread simulation area. S2. Locate the ignition point grid in the grid network based on the ignition point coordinates. Simulate the spread of wildfire by gradually expanding the grid network from the ignition point grid. Simulate the advancement of the fire area based on the shortest path algorithm and the boundary extension method, and obtain the ignition time of each grid in the grid network. The spread rate of each ignition grid to any neighboring grid is calculated by the wildfire spread model based on the slope aspect, slope, combustible material type, real-time wind direction and real-time wind speed of the grid. S3. Visualize the fire time of each grid in the grid network to display the fire simulation map at different times during the simulation period; The method for simulating the progression of forest fire spread, based on the shortest path algorithm and boundary extension method, is as follows: S201. Before performing the first step of the forest fire spread simulation, set the ignition time of the ignition point grid to the specified initial time and add the ignition point grid to the priority queue that is initially empty. S202. Take the grid with the earliest fire time from the current priority queue, delete it from the priority queue and add it to the result set, and use it as the extension point of the fire area boundary. S203. Traverse each neighboring grid in the eight neighborhoods of the expansion point that does not exist in the result set. For each current neighboring grid, firstly, the forest fire spread model calculates the spread speed of the forest fire from the expansion point to the current neighboring grid based on the slope aspect, slope, combustible material type, real-time wind direction, and real-time wind speed corresponding to the current neighboring grid. Then, it calculates the time required for the forest fire to spread from the expansion point to the current neighboring grid based on the spatial distance from the expansion point to the current neighboring grid. Finally, it obtains the ignition time of the current neighboring grid based on the ignition time of the expansion point itself. If the ignition time of the current neighboring grid is greater than the end time of the simulation period, the current neighboring grid is not added to the priority queue. Otherwise, it is necessary to determine whether the current neighboring grid already exists in the priority queue. If it does not exist, the expansion point is used as the ignition grid of the current neighboring grid, and the current neighboring grid is added to the priority queue and the corresponding ignition time is recorded. If it exists, the ignition time of this grid in the priority queue is updated to the earlier time between the calculated value of the current step size and the original recorded value, and the recorded ignition grid is updated. S204. If there are still grids in the priority queue, S202 and S203 are iterated and repeated for the next long simulation until there are no grids in the priority queue. The forest fire spread simulation is then completed. The final result records all the grids that caught fire and their corresponding ignition times during the entire simulation period.
2. The real-time simulation method for forest fire spread as described in claim 1, characterized in that, The forest fire spread model is an improved Wang Zhengfei model, in which the formula for calculating the spread rate of forest fire from the burned grid cell at the spread point to neighboring grid cells is: in The initial velocity of the forest fire's spread; The combustible coefficient is determined by the combustible type of the neighboring grid. For real-time wind direction, The slope aspect of the neighboring grid. The slope of the neighboring grid. The direction of spread from the burning grid to the neighboring grid. The real-time wind speed is given; the real-time wind direction, slope direction, and direction of spread are all calculated clockwise with due north as the 0-degree direction.
3. A fire extinguishing decision-making method, characterized in that, First, a result set recording the ignition time of each grid in the grid network is obtained according to the real-time simulation method of forest fire spread as described in claim 1. Then, a fire extinguishing decision is made by selecting one of two methods: direct fire extinguishing and indirect fire extinguishing. The direct fire suppression method is used to determine the optimal set of fire suppression points required to establish a direct fire suppression line in the simulated forest fire spread area. The determination method is shown in A1 to A4: A1. Form a directed graph by arranging all the grids in the result set according to the relationship between the ignited grid and the igniting grid. After setting an initial cost value for each grid, calculate the in-degree of each grid based on the directed graph, and add all grids with an in-degree of 0 to the initially empty grid queue. A2. Take a grid from the grid queue and delete it from the grid queue. Add the value of the currently taken grid to the corresponding ignition grid to update the value of the ignition grid. At the same time, decrement the in-degree of the ignition grid by 1. If the in-degree of the ignition grid is 0 after decrementing by 1, add the ignition grid to the grid queue. A3. If there are still grids in the grid queue, continue to iterate and repeat A2 until there are no grids in the grid queue, and then complete the iteration to obtain the result set after the cost value is updated. A4. For the result set after the value is updated, select a specified number of the best fire extinguishing points that meet the fire extinguishing time limit in descending order of value; the fire extinguishing time limit is that the firefighters can arrive at the fire extinguishing point before it is ignited and complete the fire extinguishing isolation zone setting in advance. The indirect fire extinguishing method is used to determine the optimal location for setting up a firebreak in the simulated forest fire area to prevent the spread of the forest fire at the current moment. The determination method is shown in B1 to B4: B1. Get the current time; B2. Determine the main direction of forest fire spread based on the latest fire simulation map at the end of the simulation. B3. Simulate and set up a firebreak of a preset length along the main direction of forest fire spread determined in B2, and gradually lengthen the firebreak from far to near the fire line in the current fire simulation map. Determine the optimal position closest to the fire line and where the forest fire has not yet spread to the firebreak when the firebreak is completed. Simulate and build a firebreak at the optimal position and set the forest fire spread rate coefficient at the firebreak to 0. Then, start the forest fire spread simulation again from the current moment to obtain a new fire simulation map after the current moment. B4. Based on the new fire scene simulation map obtained in B3, determine whether the current firebreaks meet the requirements for blocking the spread of wildfire. If they do, output all the current firebreaks and their locations. If they do not meet the requirements, re-execute B2 and B3 based on the new fire scene simulation map until the requirements for blocking the spread of wildfire are met, and then output all the current firebreaks and their locations.
4. The fire extinguishing decision-making method as described in claim 3, characterized in that, In A1, the initial cost value of each grid in the result set is set according to the importance of the location space of the grid. The cost value of the grid in the location space without important facilities is set to 1 by default, and the cost value of the grid in the location space with important facilities is increased.
5. The fire extinguishing decision-making method as described in claim 3, characterized in that, In A4, the fire extinguishing time is limited to satisfying the following formula: In the formula: This indicates the time required for firefighters to reach a fire location. This indicates the time when the fire was ignited at the extinguishing point. This indicates the time allotted for firefighters to set up the necessary firebreaks after they arrive at the fire site.
6. The fire extinguishing decision-making method as described in claim 3, characterized in that, In A4, the time required for firefighters to reach a fire extinguishing point is calculated based on the shortest path planning algorithm combined with road network information.
7. A forest fire spread simulation and early warning system, characterized in that, It includes a resource management module, a propagation simulation module, a monitoring and evaluation module, and a visualization module; The resource management module is used to manage the input parameters, user information resources, and simulation results required for forest fire spread simulation, and to provide corresponding resource permissions for each user. The spread simulation module is used to simulate forest fire spread in a designated forest fire spread simulation area according to the simulation method described in claim 1, based on the input parameters received by the resource management module, and is responsible for providing forest fire spread simulation results to other modules; The monitoring and assessment module is used to detect the fire situation in real time through monitoring equipment, and output the fire extinguishing decision results according to the fire extinguishing decision method described in claim 3 to assist in the scheduling of fire extinguishing resources, and to conduct loss assessment and fire level classification based on the forest fire spread simulation results. The visualization module is used to dynamically visualize the fire scene at different times during the simulation period based on the ignition time of each grid in the grid network.
8. The forest fire spread simulation and early warning system as described in claim 7, characterized in that, In the spread simulation module, the forest fire spread simulation area and ignition point are selected through the GIS map visualization interface provided by the visualization module.
9. The forest fire spread simulation and early warning system as described in claim 7, characterized in that, It also includes a fire-fighting resource scheduling decision module, which is used to assist in formulating fire-fighting resource allocation and fire-fighting resource scheduling strategies. The resource allocation is based on the assessed fire level to prepare resources, and the fire-fighting resource scheduling strategy is to use the shortest path algorithm to schedule the required resources to the fire-fighting point in the way with the shortest distance and the fastest arrival time.