Ground-air collaborative fire barrier opening method based on digital twinning and neural network

Abstract

The invention discloses a ground-air cooperative fire barrier opening method based on digital twinborn and neural networks, and belongs to the technical field of forest steppe fire prevention. Geographic information, weather and fire monitoring data are collected and preprocessed, and a digital twinborn model is constructed to simulate meteorological elements. And then constructing an evaluation index system to evaluate the fire risk, and dividing risk grades. An isolation belt is planned by adopting a genetic algorithm, and a parameter combination is explored through selection, crossing and mutation operations. During air-ground collaborative operation, the unmanned aerial vehicle guides construction, the helicopter lifts materials, and the scheme is monitored and adjusted in real time. The fire retardant effect is evaluated through simulation and actual data, the efficiency and the cost effectiveness are set up, and all links are optimized according to results. According to the method, the fire risk is accurately assessed, the isolation belt is scientifically planned, the ground-air collaborative operation efficiency is improved, dynamic monitoring and optimization are realized, the cost is reduced, and an effective means is provided for forest steppe fire prevention and control.

Description

Technical Field

[0001] The present invention relates to the field of forest and grassland fire prevention technology, and in particular to a method for establishing a ground-to-air coordinated fire isolation zone based on digital twins and neural networks. Background Art

[0002] Globally, forest and grassland fires are frequent, posing a serious threat to ecosystems, economic development, and human life and property. Traditional firebreak design methods have numerous insurmountable flaws in the complex and ever-changing forest and grassland environments.

[0003] Traditional methods for data collection and processing rely primarily on manual field measurements and simple remote sensing monitoring. Manual measurements are not only inefficient and limited in coverage, but are also susceptible to terrain and weather conditions, making it difficult to obtain real-time, comprehensive data. Early remote sensing technology, however, had low resolution and could not accurately identify vegetation types, densities, or subtle features of topography. This resulted in a lack of accurate data support for subsequent firebreak planning. Traditional data processing methods struggled to effectively integrate and analyze large amounts of multi-source data, failing to fully tap into the underlying information and making it difficult to accurately assess fire risks.

[0004] Traditional fire risk assessment methods often consider only a single or limited number of factors, such as vegetation type and distribution, while ignoring the combined influence of multiple factors, including topography, weather, and human activity. The lack of a scientific and systematic assessment indicator system makes fire risk assessments inaccurate and incomplete. Furthermore, traditional assessment methods, mostly based on empirical evidence and simple mathematical models, are unable to adapt to the dynamic changes in forest and grassland environments. This makes it difficult to accurately predict the likelihood and spread of fires in advance, and consequently, fails to provide a scientific basis for the planning of firebreaks.

[0005] Traditional approaches to planning containment zones often rely on experience, failing to fully consider the patterns of fire spread and the varying risks across different areas. The width, length, and location of containment zones lack scientific justification, resulting in some containment zones failing to effectively prevent fire spread or wasting resources due to excessive or excessive width. Furthermore, traditional planning methods fail to fully utilize modern information technology, making it difficult to simulate and optimize different planning schemes and unable to adjust planning strategies in a timely manner based on actual conditions.

[0006] During construction, the traditional ground-to-air collaborative operation model suffers from serious shortcomings. A lack of efficient communication and coordination mechanisms between ground-based construction equipment and aerial work forces results in poor information transfer, leading to low operational efficiency. For example, when helicopters are lifting equipment and supplies, improper coordination with ground-based construction equipment can cause delays or inaccurate placement, impacting the installation of fire barriers. Furthermore, traditional construction methods make it difficult to monitor construction progress and quality in real time, making it impossible to promptly detect and correct deviations during construction. This can lead to inconsistent quality of barrier zones, compromising fire prevention effectiveness.

[0007] With global warming and the expansion of human activities, the frequency and severity of forest and grassland fires are increasing. Increased extreme climate events, such as high temperatures, droughts, and strong winds, are making forest and grassland vegetation drier and more flammable, significantly increasing the risk of fire. Furthermore, increasing human development activities around forests and grasslands are increasing the potential risk points for fires. Against this backdrop, traditional firebreaks are no longer sufficient to meet the growing demand for forest fire prevention. An innovative approach integrating advanced technologies is urgently needed to enhance forest fire prevention capabilities and reduce the losses caused by fires. Summary of the Invention

[0008] The purpose of this invention is to provide a ground-to-air collaborative fire isolation zone opening method based on digital twins and neural networks, so as to solve the problems of inaccurate data, unscientific planning, low collaborative efficiency, etc. in traditional fire isolation zone opening methods, realize precise prevention and control of forest and grassland fires, improve the efficiency and quality of fire isolation zone opening, and reduce the losses caused by fire.

[0009] To achieve the above objectives, the present invention provides a method for establishing a ground-to-air coordinated fire isolation zone based on digital twins and neural networks, comprising the following steps:

[0010] S1. Data collection and preprocessing: Collect geographic information, meteorological, and fire monitoring data, and then perform coordinate unification, format conversion, quality control, and denoising operations on various types of data;

[0011] S2. Digital Twin Model Construction: First, a geometric model that accurately reproduces geographic features is constructed. Then, physical properties are added to the geometric model. A meteorological model is constructed using the Navier-Stokes equations to simulate meteorological elements. Finally, the simulation results are compared with actual observations to calibrate the model parameters.

[0012] S3. Fire Risk Assessment: Build an assessment index system covering vegetation, topography, meteorology, and human factors, use AHP to determine weights and conduct consistency checks; use a neural network combined with the Rothermel model to assess risk, classify it into low, medium, and high risk levels, and visualize it in the digital twin model;

[0013] S4. Isolation zone planning: Using a genetic algorithm, we define a fitness function evaluation scheme, optimize chromosomes through selection, crossover, and mutation operations, and explore parameter combinations for isolation zone widths through binary or real-number coded mutation. Finally, we present the planning results and simulate and analyze their fire-blocking effects, optimizing as needed.

[0014] S5. Ground-air collaborative operations: Drones guide ground construction equipment to plan routes and monitor construction, while helicopters hoist equipment and materials and receive instructions from ground workstations. Simultaneously, equipment operation and construction progress are monitored in real time, plans are adjusted based on the results, and isolation zone planning is regularly reassessed and optimized.

[0015] S6. Effect evaluation and feedback: Construct evaluation indicators for fire-retardant effect, opening efficiency, and cost-effectiveness, evaluate through simulation and record actual data, and optimize each link based on the feedback of the results.

[0016] Preferably, in step S2, the geometric model is constructed based on geographic information data, and a forest-grassland geometric model is constructed using three-dimensional modeling software to accurately restore the geographical features of the terrain. The tree crown model is constructed using three-dimensional reconstruction technology based on point cloud data, and the grassland is modeled using texture mapping.

[0017] Preferably, in step S2, a meteorological model is established, and computational fluid dynamics methods are used to describe wind field motion using the Navier-Stokes equations, which are combined with real-time meteorological data to solve and simulate the distribution and changes of meteorological elements;

[0018] In the Cartesian coordinate system, the Navier-Stokes equations are as follows:

[0019] Continuity equation:

[0020]

[0021] Momentum equation:

[0022]

[0023]

[0024] Where ρ is the fluid density, t is time, u, v, and w are the velocity components of the fluid in the x, y, and z directions, respectively, p is the pressure, and μ is the dynamic viscosity, which are the external force components in the directions. The wind speed and direction at different locations are obtained by solving the equations using a numerical algorithm;

[0025] The simulation results of the digital twin model are compared and verified with the actual observation data. The model is calibrated according to the verification results and the parameters are adjusted to make the simulation results more consistent with the actual situation.

[0026] Preferably, in step S3, constructing the evaluation index system includes the following steps:

[0027] Consider vegetation factors, topographic factors, meteorological factors, and human factors;

[0028] The analytic hierarchy process (AHP) is used to determine the weight of each evaluation indicator. The judgment matrix is constructed through expert scoring to calculate the relative weight of each indicator. The judgment matrix needs to be tested for consistency. The calculation formula of the consistency index CI is:

[0029]

[0030] Among them, λ max is the maximum eigenvalue of the judgment matrix, n is the order of the judgment matrix; the random consistency index RI has a corresponding standard value according to the matrix order n.

[0031] Preferably, in step S3, the risk assessment is performed using a neural network combined with a Rothermel model, including the following steps:

[0032] Use neural network models to assess fire risk, use assessment indicators as neural network input, and train the neural network using a large amount of historical fire data and corresponding assessment indicator data;

[0033] The Rothermel model is used to describe the surface fire spread rate, and the formula is:

[0034]

[0035] R is the fire spread rate;

[0036] I g is the potential energy release rate per unit area, and the calculation formula is H c is the calorific value of the fuel, M is the fuel load, ρ b is the bulk density of the fuel, φ is the proportion of the effective combustion part, t d is the burning duration;

[0037] ε is the energy transmission efficiency factor, which is related to fuel type and terrain factors; ω is the wind combustion factor, which is a function of wind speed and is expressed as ω = 1 + a·V, where V is the wind speed and a is a coefficient related to fuel and terrain; ω is the environmental resistance factor, which reflects the obstruction of terrain and slope to fire spread; by inputting different vegetation, meteorological and terrain parameters into the model, the speed and direction of fire spread can be predicted, providing a reference for fire risk assessment.

[0038] Preferably, in step S3, the fire risk is divided into three levels: low, medium, and high. According to the output of the neural network model, a threshold h is set to divide the risk level. The value range of the threshold h is set between 0 and 1. The specific steps are as follows:

[0039] Collect data on forest and grassland fires in different regions, seasons, and climate conditions, as well as data on various environmental parameters in the period before the fires;

[0040] Through correlation analysis, we determined the environmental factors that have a greater impact on fire risk, and then used cluster analysis to preliminarily divide the fire risk level intervals;

[0041] Organize experts in the field of forest fire prevention to evaluate and discuss the analysis results, comprehensively consider actual application needs and risk prevention and control goals, and determine the threshold h that is suitable for the local area;

[0042] The threshold h is taken as the lower threshold h1 and the upper threshold h2 respectively;

[0043] When the risk value output by the neural network model is less than h1, the area is judged to be of low risk level;

[0044] When the risk value output by the neural network model is between h1 and h2, the area is classified as medium risk level;

[0045] When the risk value output by the neural network model is greater than h2, the area is identified as a high-risk level.

[0046] Preferably, in step S4, binary coding exploration of isolation zone width parameter combinations and fire-retardant effect analysis includes the following steps:

[0047] Determine the use of binary code to represent the width of the isolation zone and clarify the actual width change corresponding to each bit of code;

[0048] Based on the genetic algorithm, a set of initial chromosome populations including parameters such as isolation zone width is randomly generated;

[0049] Mutate the encoding of the isolation band width in the chromosome;

[0050] Substitute the mutated chromosome into the fitness function to calculate the fitness value. The fitness function is:

[0051]

[0052] Fitness represents the fitness value. The larger the value, the better the corresponding isolation zone planning scheme. α and β are weight coefficients. A blocked The area of the isolation belt that successfully blocked the spread of fire obtained through the digital twin model simulation. The larger the value, the more significant the role of the isolation belt in preventing the spread of fire.total is the total area to which the fire may spread when no isolation belt is set up, reflecting the relative effect of the isolation belt in preventing the spread of fire; C is the total cost of setting up the isolation belt. The lower the C value, the more beneficial it is to improve the fitness value;

[0053] Chromosomes are selected according to their fitness values, and chromosomes with high fitness are retained to enter the next generation;

[0054] Through continuous iteration, repeated mutation, fitness calculation and selection operations, a better combination of isolation zone width parameters is gradually screened out;

[0055] The digital twin model visualizes the planning scheme corresponding to the iteratively obtained optimal barrier width parameter combination. The barrier is highlighted with different colors or line styles, clearly showing its location, direction, and width information, and annotating the relevant attributes of the barrier under that width.

[0056] Using digital twin models, we simulated the optimal width of isolation zones under various fire scenarios.

[0057] By comparing the simulation results under different scenarios, the fire-retardant effect of the isolation belt of this width is comprehensively evaluated, and its advantages and disadvantages in different environments are analyzed to provide a basis for optimization.

[0058] Preferably, the real number coding exploration of isolation zone width parameter combinations and fire retardant effect analysis includes the following steps:

[0059] Use real number coding to directly represent the width of the isolation zone;

[0060] Generate an initial chromosome population containing parameters such as isolation zone width through genetic algorithm;

[0061] Set the mutation range, encode the width of the isolation band in each chromosome, and mutate with a certain mutation probability;

[0062] Using fitness function Calculate the fitness value of the mutated chromosome and use the digital twin model to simulate the fire spread of isolation belts of different widths in a specific fire scenario to obtain A blocked and A total , combined with the cost estimate to get C, and thus the fitness value;

[0063] Chromosomes are selected based on fitness values, and chromosomes with high fitness are retained for the next generation. After multiple rounds of iterations, the chromosomes in the population are continuously approaching a more optimal isolation zone width parameter.

[0064] Display the planning scheme corresponding to the optimal isolation zone width parameters after iterative optimization in the digital twin model;

[0065] Using the digital twin model, we simulated and tested the optimal width of isolation zones in various fire scenarios, recorded the spread of the fire at different times, and calculated key indicators of the isolation zones' success in blocking the fire.

[0066] By comparing simulation data under different scenarios, we conduct an in-depth analysis of the fire-retardant effect of the isolation belt of this width under the influence of different environmental factors, providing detailed data support for subsequent solution improvements.

[0067] Preferably, in step S5, the steps of real-time monitoring and adjustment are as follows:

[0068] Utilize ground sensor networks, drone monitoring, and helicopter-mounted equipment to collect comprehensive, real-time data on the firebreak opening process;

[0069] The ground workstation conducts comprehensive analysis on the collected multi-source data;

[0070] Adjust the operation plan in a timely manner based on the data analysis and evaluation results, and continuously monitor the adjusted operation status.

[0071] Preferably, in step S6, the evaluation indicators include the fire-blocking effect, opening efficiency and cost-effectiveness of the isolation belt; wherein the fire-blocking effect of the isolation belt is measured by simulating the success rate of the isolation belt in preventing the spread of fire when a fire occurs by the digital twin model, which is specifically expressed as

[0072] Success rate = number of times fire spread was successfully prevented / total number of simulations × 100%;

[0073] The efficiency of opening is to calculate the length or area of the isolation zone opened per unit time;

[0074] Cost-effectiveness is to compare the cost of opening a median strip with the fire losses reduced by the opening of the median strip, and evaluate the cost-effectiveness ratio.

[0075] Therefore, the present invention adopts a method for establishing a ground-air coordinated fire isolation zone based on digital twins and neural networks using the above structure, which has the following beneficial effects:

[0076] (1) The present invention acquires high-precision geographic information, meteorological and fire monitoring data through multi-source data collection and preprocessing, and uses the digital twin model to build a virtual environment that is highly similar to the real scene. Combined with the neural network and Rothermel model, it can accurately assess the fire risk and scientifically plan the location, width, length and other parameters of the fire isolation zone.

[0077] (2) The ground-air collaborative system architecture of the present invention achieves close coordination between drones, helicopters, and ground construction equipment, greatly improving construction efficiency and shortening the time required to open fire isolation zones.

[0078] (3) This invention utilizes ground sensor networks, drone monitoring, and helicopter-mounted equipment to achieve real-time dynamic monitoring of the fire isolation zone opening process and the forest and grassland environment. Based on the monitoring results, the operation plan is adjusted in a timely manner to dynamically optimize the isolation zone.

[0079] (4) The present invention avoids unnecessary waste of resources and reduces opening costs by scientifically planning isolation zones. At the same time, accurate fire risk assessment and effective isolation zone setting can significantly reduce losses in the event of a fire.

[0080] (5) The fire risk assessment based on neural network and isolation zone planning based on genetic algorithm in the present invention provides intelligent support for decision-making. It can quickly analyze large amounts of data, automatically generate the optimal isolation zone planning scheme, and dynamically adjust it according to real-time data. This not only reduces the burden of manual decision-making, but also improves the accuracy and timeliness of decision-making.

[0081] The technical solution of the present invention is further described in detail below through the accompanying drawings and embodiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0082] Figure 1 This is a flow chart of a method for establishing a ground-to-air coordinated fire isolation zone based on digital twins and neural networks according to the present invention;

[0083] Figure 2 This is a schematic diagram of the ground-to-air collaborative system architecture of the present invention's ground-to-air collaborative fire isolation zone opening method based on digital twins and neural networks. DETAILED DESCRIPTION

[0084] The technical solution of the present invention is further described below with reference to the accompanying drawings and embodiments.

[0085] Unless otherwise defined, the technical or scientific terms used in the present invention shall have the usual meanings understood by persons of ordinary skill in the field to which the present invention belongs. The words "first", "second" and similar terms used in the present invention do not indicate any order, quantity or importance, but are only used to distinguish different components. Words such as "include" or "comprise" mean that the elements or objects preceding the word include the elements or objects listed after the word and their equivalents, without excluding other elements or objects. Words such as "connect" or "connected" are not limited to physical or mechanical connections, but may include electrical connections, whether direct or indirect. "Up", "down", "left", "right" and the like are only used to indicate relative positional relationships. When the absolute position of the object being described changes, the relative positional relationship may also change accordingly.

[0086] Example

[0087] like Figure 1-2 As shown, the present invention provides a method for opening a ground-air coordinated fire isolation zone based on digital twins and neural networks, comprising the following steps:

[0088] S1. Data collection and preprocessing: Collect geographic information, meteorological, and fire monitoring data, and then perform coordinate unification, format conversion, quality control, and denoising operations on various types of data;

[0089] S2. Digital Twin Model Construction: First, a geometric model that accurately reproduces geographic features is constructed. Then, physical properties are added to the geometric model. A meteorological model is constructed using the Navier-Stokes equations to simulate meteorological elements. Finally, the simulation results are compared with actual observations to calibrate the model parameters.

[0090] S3. Fire Risk Assessment: Build an assessment index system covering vegetation, topography, meteorology, and human factors, use AHP to determine weights and conduct consistency checks; use a neural network combined with the Rothermel model to assess risk, classify it into low, medium, and high risk levels, and visualize it in the digital twin model;

[0091] S4. Isolation zone planning: Using a genetic algorithm, we define a fitness function evaluation scheme, optimize chromosomes through selection, crossover, and mutation operations, and explore parameter combinations for isolation zone widths through binary or real-number coded mutation. Finally, we present the planning results and simulate and analyze their fire-blocking effects, optimizing as needed.

[0092] S5. Ground-air collaborative operations: Drones guide ground construction equipment to plan routes and monitor construction, while helicopters hoist equipment and materials and receive instructions from ground workstations. Simultaneously, equipment operation and construction progress are monitored in real time, plans are adjusted based on the results, and isolation zone planning is regularly reassessed and optimized.

[0093] S6. Effect evaluation and feedback: Construct evaluation indicators for fire-retardant effect, opening efficiency, and cost-effectiveness, evaluate through simulation and record actual data, and optimize each link based on the feedback of the results.

[0094] like Figure 2 As shown in the figure, the ground-air collaborative system architecture includes ground monitoring, execution subsystem, and air monitoring and support subsystem, among which:

[0095] Ground monitoring and execution subsystem, as follows:

[0096] Sensor network: A large number of sensor nodes are deployed in forest and grassland areas, including temperature, humidity, smoke, infrared sensors, etc., and the collected data is transmitted to the ground base station in real time through wireless ad hoc networks.

[0097] Ground workstations: Receive sensor network data and perform preliminary processing and analysis. Using digital twin models and real-time data, they visualize the current state of forests and grasslands, including vegetation distribution and meteorological conditions. Furthermore, fire risk levels for different areas are calculated using fire risk assessment algorithms.

[0098] Ground construction equipment: including bulldozers, brush cutters and other mechanical equipment used to create fire isolation zones, which are remotely controlled by ground workstations based on decision instructions or operated by operators according to planned paths.

[0099] Air surveillance and support subsystem, as follows:

[0100] Drone monitoring: Multiple drones equipped with high-definition cameras, thermal imagers, lidar, and other equipment are deployed. They patrol along pre-set routes, capturing real-time images and videos and transmitting them back to ground stations. Neural network algorithms are used to analyze images and quickly identify potential fire sources and signs.

[0101] Helicopter support: When a fire breaks out, helicopters arrive quickly on the scene. They can carry firefighting supplies for aerial extinguishing, lift large ground-based construction equipment to hard-to-reach areas, and assist in establishing firebreaks. Equipped with communication equipment, helicopters communicate with ground stations in real time to receive command instructions.

[0102] In step S1, data collection and preprocessing include geographic information data collection, meteorological data collection, fire monitoring data collection and data preprocessing, wherein:

[0103] Geographic information data collection, as follows:

[0104] Use satellite remote sensing to obtain topographic data of large areas with a resolution of 10-30 meters, and extract information such as vegetation type and land cover through image interpretation;

[0105] Use LiDAR for high-precision terrain mapping to generate centimeter-level digital elevation models (DEMs) that accurately reflect terrain undulations;

[0106] Low-altitude mapping by drones obtains high-resolution images and terrain data of local areas, supplementing the lack of satellite remote sensing and lidar data.

[0107] Meteorological data collection is as follows:

[0108] Multiple weather monitoring stations are set up in and around the area to monitor wind speed, wind direction, temperature, humidity, air pressure and other meteorological parameters in real time, and transmit the data to ground stations via wireless communication networks;

[0109] Meteorological satellite data is used to obtain regional macro-meteorological information, such as weather system movement trends and precipitation distribution, and integrated with ground meteorological monitoring station data to improve accuracy and comprehensiveness.

[0110] Fire monitoring data collection is as follows:

[0111] Using a network of ground-based smoke and infrared sensors, early signs of fire, such as smoke concentration and abnormally high temperature, can be monitored in real time;

[0112] Satellite thermal infrared sensors are used to monitor fire hotspots over large areas to determine the approximate location and extent of the fire.

[0113] Drones equipped with thermal imagers can conduct close-range, high-precision monitoring of suspected fire areas, accurately identifying the location and size of the fire.

[0114] Data preprocessing is as follows:

[0115] Process geographic information data by unifying coordinates and converting formats so that it can be integrated into the digital twin model;

[0116] Conduct quality control on meteorological data, remove outliers and erroneous data, and use Kriging interpolation to supplement missing data to ensure continuity;

[0117] The fire monitoring data is denoised, and the fire images taken by drones are processed using image enhancement algorithms to highlight the characteristics of the fire source.

[0118] In step S2, the digital twin model construction includes geometric model construction, physical model construction, and model verification and calibration, where:

[0119] Geometry model construction:

[0120] Based on geographic information data, 3D modeling software was used to construct geometric models of forests and grasslands, accurately recreating topography, mountains, rivers, lakes, and other geographical features. Different modeling methods were used for different vegetation types. For example, tree crown models were constructed using 3D reconstruction technology based on point cloud data, while grasslands were modeled using texture mapping.

[0121] Physical model construction:

[0122] Add physical properties to the digital twin model, including vegetation combustion characteristics (ignition point, calorific value, burning rate, etc.), soil thermal conductivity characteristics, water evaporation characteristics, etc., determined through experimental data and theoretical calculations;

[0123] A meteorological model was established using computational fluid dynamics (CFD) methods. The Navier-Stokes equations were used to describe wind field motion. The model was then solved using real-time meteorological data to simulate the distribution and changes of meteorological elements such as wind, temperature, and humidity. In the Cartesian coordinate system, the Navier-Stokes equations are expressed as follows:

[0124] Continuity equation:

[0125]

[0126] Momentum equation:

[0127]

[0128] Where ρ is the fluid density, t is time, u, v, and w are the velocity components of the fluid in the x, y, and z directions, respectively, p is the pressure, and μ is the dynamic viscosity, which are the external force components in the directions. The wind speed and direction at different locations are obtained by solving the equations using a numerical algorithm;

[0129] Compare and verify the digital twin model's simulation results with actual observational data. Based on the verification results, calibrate the model and adjust parameters to better align the simulation results with the actual situation. If the model-simulated wind speed deviates from the measured wind speed, adjust the roughness parameters in the meteorological model, etc., until the error is within an acceptable range.

[0130] Model validation and calibration:

[0131] Consider vegetation factors (vegetation type, density, moisture content), terrain factors (slope, aspect, altitude), meteorological factors (wind speed, wind direction, temperature, humidity), and human factors (human activity density, fire source distribution).

[0132] The analytic hierarchy process (AHP) was used to determine the weight of each evaluation indicator. A judgment matrix was constructed through expert scoring to calculate the relative weight of each indicator. The judgment matrix was subject to consistency testing. The calculation formula for the consistency index CI is:

[0133]

[0134] Among them, λ max is the maximum eigenvalue of the judgment matrix, n is the order of the judgment matrix; the random consistency index RI has a corresponding standard value according to the matrix order n.

[0135] The consistency ratio CR is expressed as

[0136]

[0137] When CR<1, the judgment matrix is considered to have acceptable consistency, otherwise it needs to be adjusted.

[0138] In step S3, the risk is assessed using a neural network combined with the Rothermel model, including the following steps:

[0139] Use neural network models to assess fire risk, use assessment indicators as neural network input, and train the neural network using a large amount of historical fire data and corresponding assessment indicator data;

[0140] The Rothermel model is used to describe the surface fire spread rate, and the formula is:

[0141]

[0142] R is the fire spread rate (m / min);

[0143] I g is the potential energy release rate per unit area (kW / m 2 ), the calculation formula is H c is the calorific value of the fuel (kJ / kg), M is the fuel load (kg / m 2 ), ρ b is the bulk density of the fuel (kg / m 3 ), φ is the proportion of the effective combustion part, t d is the burning duration (min);

[0144] ε is the energy transmission efficiency factor, which is related to fuel type and terrain factors; ω is the wind combustion factor, which is a function of wind speed and is expressed as ω = 1 + a·V, where V is the wind speed and a is a coefficient related to fuel and terrain; ω is the environmental resistance factor, which reflects the obstruction of terrain and slope to fire spread; by inputting different vegetation, meteorological and terrain parameters into the model, the speed and direction of fire spread can be predicted, providing a reference for fire risk assessment.

[0145] The fire risk is divided into three levels: low, medium, and high. According to the output of the neural network model, a threshold h is set to divide the risk level. The value range of the threshold h is set between 0 and 1. The specific steps are as follows:

[0146] Collect data on forest and grassland fires in different regions, seasons, and climate conditions, as well as data on various environmental parameters in the period before the fires;

[0147] Through correlation analysis, we determined the environmental factors that have a greater impact on fire risk, and then used cluster analysis to preliminarily divide the fire risk level intervals;

[0148] Organize experts in the field of forest fire prevention to evaluate and discuss the analysis results, comprehensively consider actual application needs and risk prevention and control goals, and determine the threshold h that is suitable for the local area;

[0149] The threshold h is taken as the lower threshold h1 and the upper threshold h2 respectively;

[0150] When the risk value output by the neural network model is less than h1, the area is judged to be of low risk level;

[0151] When the risk value output by the neural network model is between h1 and h2, the area is classified as medium risk level;

[0152] When the risk value output by the neural network model is greater than h2, the area is identified as a high-risk level.

[0153] In step S4, the isolation zone planning includes planning principles, planning methods based on optimization algorithms, and planning result presentation and analysis, where:

[0154] Planning principles: Adhere to the principles of "fire prevention, fire control, and ease of deployment." Isolation zones should be established where they can effectively prevent the spread of fire, such as along natural or artificial barriers such as ridgelines, rivers, and roads. Considering ease of construction, areas with flat terrain and relatively sparse vegetation should be selected. Leveraging digital twin model information, factors such as fire risk level, topography, and vegetation distribution should be comprehensively considered. Prioritize the planning and implementation of isolation zones in high-risk areas, with their width and density appropriately increased.

[0155] Planning methods based on optimization algorithms:

[0156] Genetic algorithm (GA) is used to plan isolation zones. The isolation zone position, length, width and other parameters are encoded to form chromosomes.

[0157] Define the fitness function to evaluate the quality of the planning scheme. The fitness function is:

[0158]

[0159] Fitness represents the fitness value. The larger the value, the better the corresponding isolation zone planning scheme. α and β are weight coefficients. A blocked The area of the isolation belt that successfully blocked the spread of fire obtained through the digital twin model simulation. The larger the value, the more significant the role of the isolation belt in preventing the spread of fire. total is the total area to which the fire may spread when no isolation belt is set up, reflecting the relative effect of the isolation belt in preventing the spread of fire; C is the total cost of setting up the isolation belt. The lower the C value, the more beneficial it is to the improvement of the fitness value.

[0160] The chromosomes are iteratively optimized through genetic operations of selection, crossover, and mutation. The chromosome encoding corresponds to the planning parameters of the isolation zone, and the genes in the chromosome represent the location, length, and width of the isolation zone.

[0161] Assume that the parent chromosome P1=[a1,a2,…,a n ] and P2=[b1,b2,…,b n ], where a i and b i , i=1,2,…,n, are the specific parameter values related to the isolation zone planning;

[0162] Randomly select a crossover point k, where 1 < k < n. Then the two offspring chromosomes C1 and C2 generated after crossover are as follows:

[0163] C1 = [a1, a2, …, a k , b k+1 , b k+2 , …, b n ;

[0164] C2 = [b1, b2, …, b k , a k+1 , a k+2 , …, a n ;

[0165] The mutation operation can introduce new gene features. For the gene representing the width of the isolation belt, if binary coding is used, the corresponding coding in the chromosome is "0010", which represents an isolation belt width of 8 meters according to the coding rule (assuming that the width change amount corresponding to each binary coding is 2 meters). In the actual simulation of the fire scenario, when the isolation belt is 8 meters wide, in the case of high wind speed and flammable vegetation, part of the fire breaks through the isolation belt, indicating that the width setting is conservative. In the mutation operation, if the "1" in the third position becomes "0", the coding becomes "0000", and the corresponding isolation belt width becomes 0 meters, which does not meet the actual requirements and has a low fitness value; if the "0" in the fourth position becomes "1", the coding becomes "0011", and the corresponding isolation belt width becomes 10 meters. After re - simulation, the 10 - meter - wide isolation belt successfully blocks the spread of the fire, the increase in construction cost is small, far lower than the loss caused by the out - of - control fire, and the fitness value is significantly improved, exploring a more appropriate width.

[0166] If real - number coding is used, assume that the original gene value of the isolation belt width is 10 meters, and the mutation range is set to fluctuate by 20% up and down based on the original value. During the mutation operation, a random mutation coefficient between - 0.2 and 0.2 is generated.假定生成的变异系数为0.15,那么变异后的隔离带宽度基因值为10×(1 + 0.15) = 11.5 meters. <000\\(0387>

[0167] Substitute this new width value into the isolation belt planning scheme and use the digital twin model to simulate the fire scenario. The results show that the 11.5 - meter - wide isolation belt can not only effectively block the fire, but also, compared with the 10 - meter - wide isolation belt, the increase in construction cost is relatively small, and the fire - blocking effect is further improved. After calculating the fitness function, it is found that the fitness value of the mutated scheme is higher than the original scheme, verifying that through the mutation operation under real - number coding, a better isolation belt width parameter has been successfully found, achieving the goals of effective fire blocking and reasonable cost control, and improving the overall optimization effect of the planning scheme.

[0168] Display and analysis of the planning results:

[0169] The optimized isolation zone planning scheme is visualized in the digital twin model, intuitively presenting information such as the isolation zone's location, direction, and width, and different levels of isolation zones are distinguished by different colors or line styles.

[0170] Using digital twin models, we simulated and analyzed planning schemes to evaluate the fire-blocking effectiveness of containment zones under different fire scenarios. For example, we simulated the spread of fire with and without containment zones under different wind speeds and directions, comparing the effectiveness of containment zones in suppressing fire spread. Based on these simulation results, we further optimized and adjusted the planning scheme.

[0171] In step S5, the ground-air collaborative operation is specifically as follows:

[0172] Drone-guided ground construction equipment:

[0173] The drone flies over the planned isolation zone, capturing real-time images and transmitting them back to the ground workstation. Using image processing algorithms, it identifies terrain features and obstacles, planning safe and efficient routes for ground construction equipment.

[0174] Ground-based construction equipment receives route information transmitted by drones and follows the route through its autonomous driving system or human operators. Drones continuously monitor construction progress and quality, promptly identifying and correcting deviations.

[0175] Helicopter lifting equipment and materials:

[0176] In areas with complex terrain and difficult to reach by ground construction equipment, helicopters can lift small construction equipment to designated locations and can also lift fire-fighting materials to the vicinity of the fire scene.

[0177] During the lift, the helicopter communicates with the ground station in real time, receiving flight instructions and safety alerts. Based on the digital twin model's terrain information and weather data, the ground station plans a safe flight path for the helicopter, avoiding obstacles and inclement weather.

[0178] Real-time monitoring and adjustment:

[0179] Utilize ground sensor networks, drone monitoring, and helicopter-mounted equipment to monitor the progress and quality of fire isolation zone opening in real time.

[0180] Adjust the work plan based on real-time monitoring results. If the construction progress lags behind, add construction equipment or adjust the construction process. If the isolation zone quality does not meet the standards, rework the work promptly.

[0181] During real-time monitoring, it's also important to monitor the equipment's operating status. By installing sensors on construction equipment, drones, and helicopters to collect data on vibration, temperature, fuel consumption, and other factors, machine learning algorithms can be used to develop a model for predicting equipment failures. If any signs of equipment anomaly are detected, maintenance can be scheduled in advance to prevent sudden equipment failures during operation, which could impact the establishment of firebreaks.

[0182] At the same time, taking into account the dynamic changes in the forest and grassland environment, the fire risk assessment model and isolation zone planning algorithm are re-run at regular intervals using the latest geographic information data, meteorological data and fire monitoring data. The existing isolation zones are adjusted and improved based on the assessment results, such as widening isolation zones with insufficient width and replanning and opening isolation zones with unreasonable locations.

[0183] In step S6, the effect evaluation and feedback are as follows:

[0184] Constructing evaluation metrics:

[0185] The fire-blocking effectiveness of isolation barriers is measured by the success rate of isolation barriers in preventing fire spread during fire simulations using the digital twin model. Success rate = number of fire spread successfully blocked / total number of simulations × 100%. The average fire-blocking time of isolation barriers in different fire scenarios can also be evaluated, which is the average time from the onset of a fire to its complete containment by the isolation barrier.

[0186] Construction efficiency measures the length or area of the isolation zone created per unit time. Furthermore, considering equipment utilization efficiency, the ratio of actual equipment operating time to total operating time can be calculated to assess the efficiency of equipment resource utilization. A low ratio indicates issues such as idle equipment or excessive waiting times during construction, necessitating further optimization of the construction process.

[0187] Cost-effectiveness: Compare the costs of establishing isolation zones with the reduced fire losses (including forest resource losses, ecological damage losses, and firefighting costs) resulting from their establishment to assess the cost-benefit ratio. Forest resource losses can be estimated based on different tree species, forest age, and area, combined with market timber prices and ecological service valuation methods. Ecological damage losses are quantified using ecological economic models, taking into account the long-term impacts of fires on soil, water resources, and biodiversity.

[0188] Evaluation Methodology:

[0189] Using a digital twin model, we conducted multiple fire simulations to evaluate the fire-blocking effectiveness of isolation zones under various fire scenarios. Each simulation detailed the path and speed of fire spread, as well as the barrier's effectiveness. Statistical analysis of the extensive simulation data revealed the barrier's fire-blocking success rate and average fire-blocking time.

[0190] Record construction time, equipment usage, and other data during the actual construction process to calculate construction efficiency. Using GPS positioning systems and working status monitoring devices installed on construction equipment, the equipment's working trajectory, working hours, and workload at each stage are recorded in real time, allowing accurate calculation of the length or area of the isolation zone opened per unit time, as well as equipment utilization efficiency.

[0191] By compiling post-fire loss data and combining it with data on the costs of establishing isolation zones, we calculated a cost-benefit ratio. We collaborated with forestry departments and ecological and environmental monitoring agencies to obtain assessment reports on the damage to forest resources and the ecological environment following the fire. We also compiled detailed information on all costs during the construction process and conducted a comprehensive cost-benefit analysis.

[0192] Feedback and Optimization:

[0193] Based on the effectiveness evaluation results, any problems will be fed back into the system. If the fire barrier effect of the isolation zone is not ideal, the cause will be analyzed and the planning and construction process will be optimized accordingly.

[0194] If the opening efficiency is low, analyze whether there is room for optimization in the construction process and whether it is necessary to replace more efficient construction equipment.

[0195] If the cost-effectiveness ratio is not high, investigate measures to reduce costs or improve benefits.

[0196] Therefore, the present invention adopts the above-mentioned ground-to-air collaborative fire isolation zone opening method based on digital twins and neural networks to solve the problems of inaccurate data, unscientific planning, and low collaborative efficiency in traditional fire isolation zone opening methods, thereby achieving precise prevention and control of forest and grassland fires, improving the efficiency and quality of fire isolation zone opening, and reducing losses caused by fires.

[0197] 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 the same. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that they can still modify or replace the technical solutions of the present invention with equivalents, and these modifications or equivalent replacements cannot cause the modified technical solutions to deviate from the spirit and scope of the technical solutions of the present invention.

Claims

1. A method for establishing ground-air coordinated fire isolation zones based on digital twins and neural networks, characterized by: The following steps are involved: S1. Data collection and preprocessing: Collect geographic information, meteorological, and fire monitoring data, and then perform coordinate unification, format conversion, quality control, and denoising operations on various types of data; S2. Digital Twin Model Construction: First, a geometric model that accurately reproduces geographic features is constructed. Then, physical properties are added to the geometric model. A meteorological model is constructed using the Navier-Stokes equations to simulate meteorological elements. Finally, the simulation results are compared with actual observations to calibrate the model parameters. S3. Fire Risk Assessment: Build an assessment index system covering vegetation, topography, meteorology, and human factors, use AHP to determine weights and conduct consistency checks; use a neural network combined with the Rothermel model to assess risk, classify it into low, medium, and high risk levels, and visualize it in the digital twin model; S4. Isolation zone planning: Using a genetic algorithm, we define a fitness function evaluation scheme, optimize chromosomes through selection, crossover, and mutation operations, and explore parameter combinations for isolation zone widths through binary or real-number coded mutation. Finally, we present the planning results and simulate and analyze their fire-blocking effects, optimizing as needed. S5. Ground-air collaborative operations: Drones guide ground construction equipment to plan routes and monitor construction, while helicopters hoist equipment and materials and receive instructions from ground workstations. Simultaneously, equipment operation and construction progress are monitored in real time, plans are adjusted based on the results, and isolation zone planning is regularly reassessed and optimized. S6. Effect evaluation and feedback: Construct evaluation indicators for fire-retardant effect, opening efficiency, and cost-effectiveness, evaluate through simulation and record actual data, and optimize each link based on the feedback of the results.

2. The method for establishing a ground-to-air coordinated fire isolation zone based on digital twins and neural networks according to claim 1 is characterized by: In step S2, the geometric model is constructed based on the geographic information data, and the forest-grassland geometric model is constructed using three-dimensional modeling software to accurately restore the geographical features of the terrain. The tree crown model is constructed using three-dimensional reconstruction technology based on point cloud data, and the grassland is modeled using texture mapping.

3. The method for establishing a ground-to-air coordinated fire isolation zone based on digital twins and neural networks according to claim 2 is characterized by: In step S2, a meteorological model is established, and computational fluid dynamics methods are used to describe wind field motion using the Navier-Stokes equations, which are combined with real-time meteorological data to solve and simulate the distribution and changes of meteorological elements; In the Cartesian coordinate system, the Navier-Stokes equations are as follows: Continuity equation: Momentum equation: Where ρ is the fluid density, t is time, u, v, and w are the velocity components of the fluid in the x, y, and z directions, respectively, p is the pressure, and μ is the dynamic viscosity, which are the external force components in the directions. The wind speed and direction at different locations are obtained by solving the equations using a numerical algorithm; The simulation results of the digital twin model are compared and verified with the actual observation data. The model is calibrated according to the verification results and the parameters are adjusted to make the simulation results more consistent with the actual situation.

4. The method for establishing a ground-to-air coordinated fire isolation zone based on digital twins and neural networks according to claim 1 is characterized by: In step S3, constructing the evaluation index system includes the following steps: Consider vegetation factors, topographic factors, meteorological factors, and human factors; The analytic hierarchy process (AHP) is used to determine the weight of each evaluation indicator. The judgment matrix is constructed through expert scoring to calculate the relative weight of each indicator. The judgment matrix needs to be tested for consistency. The calculation formula of the consistency index CI is: Among them, λ max is the maximum eigenvalue of the judgment matrix, n is the order of the judgment matrix; the random consistency index RI has a corresponding standard value according to the matrix order n.

5. The method for establishing a ground-to-air coordinated fire isolation zone based on digital twins and neural networks according to claim 1 is characterized by: In step S3, the risk is assessed using a neural network combined with the Rothermel model, including the following steps: Use neural network models to assess fire risk, use assessment indicators as neural network input, and train the neural network using a large amount of historical fire data and corresponding assessment indicator data; The Rothermel model is used to describe the surface fire spread rate, and the formula is: R is the fire spread rate; I g is the potential energy release rate per unit area, and the calculation formula is H c is the calorific value of the fuel, M is the fuel load, ρ b is the bulk density of the fuel, φ is the proportion of the effective combustion part, t d is the burning duration; ε is the energy transmission efficiency factor, which is related to fuel type and terrain factors; ω is the wind combustion factor, which is a function of wind speed and is expressed as ω = 1 + a·V, where V is the wind speed and a is a coefficient related to fuel and terrain; ω is the environmental resistance factor, which reflects the obstruction of terrain and slope to fire spread; by inputting different vegetation, meteorological and terrain parameters into the model, the speed and direction of fire spread can be predicted, providing a reference for fire risk assessment.

6. The method for establishing a ground-to-air coordinated fire isolation zone based on digital twins and neural networks according to claim 1 is characterized by: In step S3, the fire risk is divided into three levels: low, medium, and high. According to the output of the neural network model, a threshold h is set to divide the risk level. The value range of the threshold h is set between 0 and 1. The specific steps are as follows: Collect data on forest and grassland fires in different regions, seasons, and climate conditions, as well as data on various environmental parameters in the period before the fires; Through correlation analysis, we determined the environmental factors that have a greater impact on fire risk, and then used cluster analysis to preliminarily divide the fire risk level intervals; Organize experts in the field of forest fire prevention to evaluate and discuss the analysis results, comprehensively consider actual application needs and risk prevention and control goals, and determine the threshold h that is suitable for the local area; The threshold h is taken as the lower threshold h1 and the upper threshold h2 respectively; When the risk value output by the neural network model is less than h1, the area is judged to be of low risk level; When the risk value output by the neural network model is between h1 and h2, the area is classified as medium risk level; When the risk value output by the neural network model is greater than h2, the area is identified as a high-risk level.

7. The method for establishing a ground-to-air coordinated fire isolation zone based on digital twins and neural networks according to claim 1 is characterized by: In step S4, binary coding is used to explore isolation zone width parameter combinations and analyze fire-blocking effects, including the following steps: Determine the use of binary code to represent the width of the isolation zone and clarify the actual width change corresponding to each bit of code; Based on the genetic algorithm, a set of initial chromosome populations including parameters such as isolation zone width is randomly generated; Mutate the encoding of the isolation band width in the chromosome; Substitute the mutated chromosome into the fitness function to calculate the fitness value. The fitness function is: Fitness represents the fitness value. The larger the value, the better the corresponding isolation zone planning scheme. α and β are weight coefficients. A blocked The area of the isolation belt that successfully blocked the spread of fire obtained through the digital twin model simulation. The larger the value, the more significant the role of the isolation belt in preventing the spread of fire. total is the total area to which the fire may spread when no isolation belt is set up, reflecting the relative effect of the isolation belt in preventing the spread of fire; C is the total cost of setting up the isolation belt. The lower the C value, the more beneficial it is to improve the fitness value; Chromosomes are selected according to their fitness values, and chromosomes with high fitness are retained to enter the next generation; Through continuous iteration, repeated mutation, fitness calculation and selection operations, a better combination of isolation zone width parameters is gradually screened out; The digital twin model visualizes the planning scheme corresponding to the iteratively obtained optimal barrier width parameter combination. The barrier is highlighted with different colors or line styles, clearly showing its location, direction, and width information, and annotating the relevant attributes of the barrier under that width. Using digital twin models, we simulated the optimal width of isolation zones under various fire scenarios. By comparing the simulation results under different scenarios, the fire-retardant effect of the isolation belt of this width is comprehensively evaluated, and its advantages and disadvantages in different environments are analyzed to provide a basis for optimization.

8. The method for establishing a ground-to-air coordinated fire isolation zone based on digital twins and neural networks according to claim 1 is characterized by: Real number coding exploration of isolation zone width parameter combinations and fire retardant effect analysis includes the following steps: Use real number coding to directly represent the width of the isolation zone; Generate an initial chromosome population containing parameters such as isolation zone width through genetic algorithm; Set the mutation range, encode the width of the isolation band in each chromosome, and mutate with a certain mutation probability; Using fitness function Calculate the fitness value of the mutated chromosome and use the digital twin model to simulate the fire spread of isolation belts of different widths in a specific fire scenario to obtain A blocked and A total , combined with the cost estimate to get C, and thus the fitness value; Chromosomes are selected based on fitness values, and chromosomes with high fitness are retained for the next generation. After multiple rounds of iterations, the chromosomes in the population are continuously approaching a more optimal isolation zone width parameter. Display the planning scheme corresponding to the optimal isolation zone width parameters after iterative optimization in the digital twin model; Using the digital twin model, we simulated and tested the optimal width of isolation zones in various fire scenarios, recorded the spread of the fire at different times, and calculated key indicators of the isolation zones' success in blocking the fire. By comparing simulation data under different scenarios, we conduct an in-depth analysis of the fire-retardant effect of the isolation belt of this width under the influence of different environmental factors, providing detailed data support for subsequent solution improvements.

9. The method for establishing a ground-to-air coordinated fire isolation zone based on digital twins and neural networks according to claim 1 is characterized by: In step S5, the steps of real-time monitoring and adjustment are as follows: Utilize ground sensor networks, drone monitoring, and helicopter-mounted equipment to collect comprehensive, real-time data on the firebreak opening process; The ground workstation conducts comprehensive analysis on the collected multi-source data; Adjust the operation plan in a timely manner based on the data analysis and evaluation results, and continuously monitor the adjusted operation status.

10. The method for establishing a ground-to-air coordinated fire isolation zone based on digital twins and neural networks according to claim 1 is characterized by: In step S6, the evaluation indicators include the fire-blocking effect, opening efficiency and cost-effectiveness of the isolation belt; the fire-blocking effect of the isolation belt is measured by simulating the success rate of the isolation belt in preventing the spread of fire when the digital twin model simulates the fire, which is specifically expressed as Success rate = number of times fire spread was successfully prevented / total number of simulations × 100%; The efficiency of opening is to calculate the length or area of the isolation zone opened per unit time; Cost-effectiveness is to compare the cost of opening a median strip with the fire losses reduced by the opening of the median strip, and evaluate the cost-effectiveness ratio.

Images