Forest fire hazard level prediction method based on artificial intelligence

Abstract

The invention provides a forest fire hazard level prediction method based on artificial intelligence, and the method comprises the steps: building a forest fire hazard assessment model, carrying out the fire hazard assessment, generating a forest fire hazard level distribution diagram, and compiling an analysis report through the combustible carrying capacity data, the field fire source investigation result, the meteorological condition data and the like. The forest fire hazard level prediction method based on artificial intelligence has the advantages of being comprehensive, systematic, accurate and the like. Through comprehensive application of multi-source data and an artificial intelligence algorithm, accurate prediction and grading of forest fire risk are realized, and a scientific basis and decision support are provided for forest fire prevention work. The implementation of the method is helpful for improving the forest fire prevention and coping capacity and reducing the harm of the forest fire to the ecological environment and the life and property of people.

Description

Technical Field

[0001] The present invention relates to the field of artificial intelligence technology, and particularly to a method for predicting the forest fire danger level based on artificial intelligence. Background Art

[0002] Forest fires, as a serious natural disaster, pose a major threat to the ecological environment, the safety of human life and property, and social and economic activities. Traditional methods for assessing the risk of forest fires mainly rely on static statistical data, historical experience, and expert judgment. These methods often have the following limitations: Traditional methods usually only consider a small number of evaluation indicators, such as fuel load, meteorological conditions, etc., while ignoring the impact of other important factors, such as wildfire sources and terrain, on the risk of forest fires. This single-index evaluation method is difficult to comprehensively reflect the complexity and diversity of the risk of forest fires. Traditional methods often lack systematic integration and analysis of evaluation indicators, resulting in the lack of comprehensiveness and accuracy of evaluation results. In addition, the evaluation process usually lacks standardized processes and methods, making it difficult to conduct horizontal comparisons and comprehensive analyses of evaluation results in different regions. In view of the above defects, there is an urgent need for a method for predicting the forest fire danger level based on artificial intelligence that can comprehensively consider multiple factors and achieve systematicness. Summary of the Invention

[0003] In view of this, in order to solve the problems existing in the technical background, the present invention proposes a method for predicting the forest fire danger level based on artificial intelligence. By using fuel load data, wildfire source investigation results, and meteorological condition data, etc., a forest fire risk assessment model is established to conduct fire risk assessment, generate a forest fire risk level distribution map, and compile an analysis report. The specific technical solutions are as follows: A method for predicting the forest fire danger level based on artificial intelligence includes the following steps: Step 1: Use the scoring method to determine the weights of risk assessment indicators. The forest fire risk assessment indicators include four aspects: combustibles, meteorological conditions, wildfire sources, and terrain, a total of 15 items; Step 2, calculation of the risk index: The calculation formula for the risk index of a subcompartment is as follows: ; In the formula: HI is the forest fire risk index; Wi is the weight value of the i-th indicator in the secondary risk indicators; Hi ′ is the value of the i-th indicator after standardization; Step 3: Take the county as the unit and convert the subcompartment risk index distribution map within the county territory into standard grid cell data; Step 4: Calculation of the forest fire risk index at the township level. The forest fire risk index at the township level is the product of the average risk index of the standard grid cells within the jurisdiction and the proportion of the township's forest land area. The forest fire risk index at the county level is formed based on the weighted average of the township-level risk index and forest land area within its jurisdiction. Step 5: Based on the forest fire risk indices of the standard grid cells, township units, and county units within the regional scope, delimit the index ranges for different forest fire risk levels. According to the index ranges, the forest fire risk levels are divided into four levels, namely high, medium-high, medium-low, and low. Step 6: Assessment of key potential hazards of forest fires. According to the risk assessment factors, distribution of housing buildings, distribution of firebreaks, roads, and rivers, distribution of forest resources, distribution of wildfire ignition points, distribution of combustibles, etc., extract assessment indicators, calculate the hazard indices of each disaster-bearing body and hazard-causing potential indices according to their respective weights, and conduct hierarchical assessments to form assessment data on the distribution of hazards of each disaster-bearing body and the distribution of hazard-causing potential. Step 7: Forest fire risk assessment and zoning. Taking 30″ standard grid cells, township-level, and county-level as assessment units, assess the comprehensive risk, forest resource risk, building risk, population risk, and economic risk.

[0004] Step 8: Generate a forest fire risk level distribution map and compile an analysis report. Furthermore, the conversion method of the grid risk index: Overlay the small-class risk index distribution map with the standard grid, judge the spatial relationship between the center point of each grid and the small class, and the risk index value of the small class where the grid center point is located is the risk index value of that grid.

[0005] Furthermore, in Step 7, taking townships as units, obtain indicators such as comprehensive risk, forest resource risk, building risk, population risk, and economic risk, combine the risk distribution indicators and basic geographic information, calculate the exposure and vulnerability of disaster-bearing bodies, and then conduct assessments and zoning on the comprehensive risk, forest resource risk, building risk, population risk, and economic risk respectively.

[0006] Furthermore, in Step 7, the risk rating method is used for the comprehensive forest fire risk assessment. The risk assessment is carried out under the classical framework of the three elements of hazard-causing potential (H), exposure of disaster-bearing bodies (E), and vulnerability of disaster-bearing bodies (V). After scoring each of the three elements, the risk R value is calculated according to the following formula: ; Where: R: Comprehensive forest fire risk index. H: Hazard level value of forest fire (taking a 30″ standard grid or township as a unit, the hazard level during the fire prevention period). The forest fire hazard is divided into four levels: "high, medium-high, medium-low, low", and the corresponding values are 1, 2, 3, 4 respectively; E: Exposure level value of forest fire disaster-bearing bodies. The forest fire disaster-bearing bodies include forest resources, buildings, fire prevention facilities, population, and economy. Using the forest stock volume, the number of buildings and fire prevention facilities, the population number, and GDP within the assessment unit as indicators, the exposure level of the disaster-bearing bodies is divided as required. The exposure level is divided into four levels: "high, medium-high, medium-low, low", and the corresponding values are 1, 2, 3, 4 respectively. V: Vulnerability level value of forest fire disaster-bearing bodies. Using the proportion of forest stock volume of flammable tree species, the proportion of the number of flammable buildings, the proportion of the number of elderly and children, and the GDP per unit area within the assessment unit as indicators, the vulnerability level of the disaster-bearing bodies is divided. The vulnerability level is divided into four levels: "high, medium-high, medium-low, low", and the corresponding values are 1, 2, 3, 4 respectively.

[0007] Adopting the above technical solutions, the following beneficial effects are achieved: The forest fire danger level prediction method based on artificial intelligence of the present invention has the characteristics of comprehensiveness, systematicness, accuracy, etc. By comprehensively using multi-source data and artificial intelligence algorithms, accurate prediction and level classification of forest fire hazards are realized, providing a scientific basis and decision-making support for forest fire prevention work. The implementation of this method will help improve the forest fire prevention and response capabilities and reduce the harm of forest fires to the ecological environment and people's lives and property. Description of the Drawings

[0008] Figure 1 It is a flowchart of the forest fire danger level prediction method based on artificial intelligence of the present invention; Figure 2 It is a more obvious evaluation and zoning flowchart of the forest fire danger level prediction method based on artificial intelligence of the present invention.

[0009] Figure 3 It is a table of forest fire hazard assessment index system of the forest fire danger level prediction method based on artificial intelligence of the present invention. Detailed Embodiments

[0010] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

[0011] An artificial intelligence-based method for predicting the forest fire danger level, comprising the following steps: Step 1: Use the scoring method to determine the weights of the danger assessment indicators. The forest fire danger assessment indicators include four aspects: combustibles, meteorological conditions, wildfire sources, and terrain, with a total of 15 items, as shown in Figure 3 shown.

[0012] Step 2, calculation of the danger index: The calculation formula for the danger index of a subcompartment is as follows: ; In the formula: HI is the forest fire danger index; Wi is the weight value of the i-th indicator in the secondary danger indicators; Hi ′ is the value of the i-th indicator after standardization; Step 3: Take the county as the unit and convert the distribution map of the danger index of subcompartments within the county into standard grid cell data; Step 4, calculation of the forest fire danger index at the township level. The forest fire danger index at the township level is the product of the average value of the danger index of the standard grid within the jurisdiction and the proportion of the forest land area of the township. The forest fire danger index at the county level is formed by the weighted average of the forest land areas of the township-level danger indices within its jurisdiction; Step 5: According to the forest fire danger indices of the standard grid cells, township units, and county units within the regional scope, delimit the index intervals for different forest fire danger levels. Based on the index intervals, the forest fire danger levels are divided into four levels, namely high, medium-high, medium-low, and low; Step 6, assessment of key hidden dangers of forest fires. According to the danger assessment factors, distribution of housing buildings, distribution of firebreaks, roads, rivers, distribution of forest resources, distribution of wildfire source points, distribution of combustibles, etc., extract assessment indicators, calculate the hidden danger indices of each disaster-bearing body and disaster-causing hidden danger indices according to their respective weights and conduct hierarchical assessments to form assessment data on the distribution of hidden dangers of each disaster-bearing body and the distribution of disaster-causing hidden dangers; Step 7, forest fire risk assessment and zoning. Take the 30″ standard grid, township level, and county level as assessment units to assess the comprehensive risk, forest resource risk, building risk, population risk, and economic risk.

[0013] Step 8: Generate a forest fire danger level distribution map and compile an analysis report In this embodiment, the conversion method of the grid danger index: Superimpose the distribution map of the subcompartment danger index with the standard grid, judge the spatial relationship between the center point of each grid and the subcompartment, and the danger index value of the subcompartment where the grid center point is located is the danger index value of the grid.

[0014] See Figure 2As shown, in Step 7, indicators such as comprehensive risk, forest resource risk, building risk, population risk, and economic risk are obtained at the township level. Combining the hazard distribution index and basic geographic information, the exposure and vulnerability of disaster-bearing bodies are calculated. Then, the comprehensive risk, forest resource risk, building risk, population risk, and economic risk are respectively evaluated and zoned. In Step 7, the risk rating method is used for the comprehensive risk assessment of forest fires. The risk assessment is carried out under the classic framework of three elements: disaster-causing hazard (H), exposure of disaster-bearing bodies (E), and vulnerability of disaster-bearing bodies (V). After scoring the three elements respectively, the risk R value is calculated according to the following formula: ; Where: R: Comprehensive risk index of forest fires. H: Hazard level value of forest fire disaster-causing (taking a 30″ standard grid or township as the unit, hazard level during the fire prevention period). The forest fire hazard is divided into four levels: "high, medium-high, medium-low, low", and the corresponding values are 1, 2, 3, 4 respectively; E: Exposure level value of forest fire disaster-bearing bodies. The forest fire disaster-bearing bodies are forest resources, housing buildings, fire prevention facilities, population, and economy. The forest volume, number of housing buildings and fire prevention facilities, population number, and GDP within the assessment unit are used as indicators, and the exposure level of disaster-bearing bodies is divided according to requirements. The exposure level is divided into four levels: "high, medium-high, medium-low, low", and the corresponding values are 1, 2, 3, 4 respectively. V: Vulnerability level value of forest fire disaster-bearing bodies. The proportion of forest volume of flammable tree species, proportion of flammable building numbers, proportion of the number of the elderly and children, and GDP per unit area within the assessment unit are used as indicators to divide the vulnerability level of disaster-bearing bodies. The vulnerability level is divided into four levels: "high, medium-high, medium-low, low", and the corresponding values are 1, 2, 3, 4 respectively.

[0015] The present invention also includes the assessment of disaster reduction capabilities, which assesses the disaster reduction capabilities of forest fires, including government management capabilities, professional team rescue capabilities, monitoring and early warning capabilities, engineering protection capabilities, etc. Its work process is to collect basic data, extract disaster reduction capability indicators, calculate various indexes, calculate the disaster reduction capability index, summarize the disaster reduction capability index, and divide the disaster reduction capability level. The forest fire disaster reduction capability assessment indicators include four aspects: government management capabilities, professional team rescue capabilities, monitoring and early warning capabilities, and engineering protection capabilities.

[0016] Calculation of the disaster reduction capability index. The calculation formula for the forest fire disaster reduction capability index of a township is as follows: ; Where: P is the forest fire disaster reduction ability index; Wi is the weight value of the i-th secondary index of the forest fire disaster reduction ability; Pi ′ is the value of the i-th secondary index of the forest fire disaster reduction ability after standardization; n is the number of secondary indexes of the forest fire disaster reduction ability.

[0017] The forest fire disaster reduction ability index at the county level is formed by the weighted average of the disaster reduction ability indexes and forest land areas of the lower-level administrative units within its jurisdiction according to the following formula: ; Where: P is the regional forest fire disaster reduction ability index; Pi is the forest fire disaster reduction ability index of the i-th lower-level administrative unit; is the regional forest land area of the i-th lower-level administrative unit, with the unit of hm2; n is the number of lower-level administrative units.

[0018] Using the standard deviation method, the forest fire disaster reduction ability index is classified. The forest fire disaster reduction abilities of the counties (cities, districts) in a certain city are divided into five levels, namely high, medium-high, medium, medium-low, and low.

[0019] The work process of the key hidden danger assessment of forest fires is as follows: According to the risk assessment factors, the distribution of housing buildings, the distribution of firebreaks, roads, and rivers, the distribution of forest resources, the distribution of wildfire ignition points, the distribution of combustibles, etc., assessment indicators are extracted. The hidden danger indexes of each disaster-bearing body and the disaster-causing hidden danger indexes are calculated according to their respective weights and classified for assessment to form the assessment data of the hidden danger distribution of each disaster-bearing body and the disaster-causing hidden danger distribution. Taking the single housing building within 100 meters of the forest area and its edge as the center point, buffer zones are drawn with radii of 10 meters, 50 meters, and 100 meters respectively. The forest fire risk level map (standard grid) is overlaid. According to the forest fire risks of levels I to IV, the grids with different levels of risk within 10 meters, 10 - 50 meters, and 50 - 100 meters of the building are extracted respectively. The proportion of the grid area with different risk levels in the buffer zone area is calculated. Taking the three distance ranges as assessment indicators, the index weights are set, and the hidden danger values of the disaster-bearing bodies in different distance ranges are calculated. The sum of the hidden danger values in the three distance ranges is the hidden danger index of the single housing building. According to the forest fire housing building hidden danger index RIE, the hidden dangers of single housing buildings are divided into three levels, namely key, general, and minor. Taking the township as a unit, according to the ratio of the number of housing buildings with key, general, and minor hidden danger levels of single forest fires in the disaster-bearing body within the region to the total number of housing buildings in the region, the hidden danger level of the regional forest fire of the disaster-bearing body is determined. The hidden danger level of the regional forest fire of the disaster-bearing body is divided according to PRIEi. The hidden danger level of the regional forest fire of the disaster-bearing body represents the overall hidden danger level of the forest fire of the housing building disaster-bearing body in the region.

[0020] The method for predicting the forest fire danger level based on artificial intelligence of the present invention has the characteristics of comprehensiveness, systematicness, accuracy, etc. By comprehensively applying multi-source data and artificial intelligence algorithms, accurate prediction and level classification of the forest fire danger are realized, providing a scientific basis and decision-making support for forest fire prevention work. The implementation of this method will help improve the forest fire prevention and response capabilities and reduce the harm of forest fires to the ecological environment and people's lives and property.

[0021] The basic principles and main features of the present invention have been described above. Those skilled in the art should understand that the present invention is not limited by the above embodiments. What is described in the above embodiments and the specification only illustrates the principles of the present invention. Without departing from the spirit and scope of the present invention, various changes and improvements will occur to the present invention, and these changes and improvements all fall within the scope of the present invention claimed. The scope of the invention claimed is defined by the appended claims and their equivalents.

Claims

1. A method for predicting the forest fire danger level based on artificial intelligence, characterized in that, It includes the following steps: Step 1: Determine the weights of the risk assessment indicators using the scoring method. The forest fire risk assessment indicators include four aspects: combustibles, meteorological conditions, wildfire sources, and terrain, with a total of 15 items. Step 2, Calculate the risk index: The calculation formula for the risk index of a subcompartment is as follows: ； In the formula: HI is the forest fire risk index; Wi is the weight value of the i-th indicator in the secondary risk indicators; Hi ′ is the value of the i-th indicator after standardization. Step 3: Take the county as the unit and convert the risk index distribution map of the subcompartments within the county into standard grid cell data. Step 4: Calculate the forest fire risk index of the township. The forest fire risk index at the township level is the product of the average risk index of the standard grids within the jurisdiction and the proportion of the township's forest land area. The forest fire risk index at the county level is formed by the weighted average of the township-level risk index forest land areas within its jurisdiction. Step 5: According to the forest fire risk indices of the standard grid cells, township units, and county units within the regional scope, delimit the index intervals for different forest fire risk levels. Based on the index intervals, divide the forest fire risk levels into four levels, namely high, medium-high, medium-low, and low. Step 6: Conduct a key hidden danger assessment of forest fires. Extract assessment indicators according to risk assessment factors, the distribution of housing buildings, the distribution of firebreaks, roads, rivers, the distribution of forest resources, the distribution of wildfire source points, the distribution of combustibles, etc. Calculate the hidden danger indices of each disaster-bearing body and disaster-causing hidden danger indices according to their respective weights and conduct a graded assessment to form the assessment data of the hidden danger distribution of each disaster-bearing body and the disaster-causing hidden danger distribution. Step 7: Conduct a forest fire risk assessment and zoning. Take 30″ standard grid cells, township levels, and county levels as assessment units to assess the comprehensive risk, forest resource risk, building risk, population risk, and economic risk. Step 8: Generate a forest fire risk level distribution map and compile an analysis report.

2. The method for predicting the forest fire danger level based on artificial intelligence according to claim 1, wherein The conversion method of the grid risk index: Overlay the risk index distribution map of the subcompartments with the standard grid, judge the spatial relationship between the center point of each grid and the subcompartments, and the risk index value of the subcompartment where the center point of the grid is located is the risk index value of the grid.

3. The method for predicting the forest fire danger level based on artificial intelligence according to claim 1, wherein, In Step 7, take the township as the unit to obtain indicators such as comprehensive risk, forest resource risk, building risk, population risk, and economic risk. Combine the risk distribution indicators and basic geographic information to calculate the exposure and vulnerability of the disaster-bearing bodies, and then conduct an assessment and zoning of the comprehensive risk, forest resource risk, building risk, population risk, and economic risk respectively.

4. The method for predicting the forest fire danger level based on artificial intelligence according to claim 1 or 3, characterized in that, In Step 7, the risk level method is used for the comprehensive forest fire risk assessment. Under the classical framework of the three elements of disaster-causing risk (H), exposure of the disaster-bearing body (E), and vulnerability of the disaster-bearing body (V), after scoring each of the three elements, the risk R value is calculated according to the following formula: ； In the formula: R: The comprehensive forest fire risk index; H: The value of the disaster-causing risk level of forest fires (taking the 30″ standard grid or township as the unit, the risk level during the fire prevention period); the forest fire risk is divided into four levels: "high, medium-high, medium-low, low", and the corresponding values are 1, 2, 3, and 4 respectively; E: The value of the exposure level of forest fire disaster-bearing bodies. The forest fire disaster-bearing bodies are forest resources, buildings, fire prevention facilities, population, and economy. The forest volume, the number of buildings and fire prevention facilities, the population number, and GDP within the assessment unit are used as indicators to divide the exposure level of the disaster-bearing bodies according to requirements. The exposure level is divided into four levels: "high, medium-high, medium-low, low", and the corresponding values are 1, 2, 3, and 4 respectively. V: The value of the vulnerability level of forest fire disaster-bearing bodies. The proportion of forest volume of flammable tree species, the proportion of the number of flammable buildings, the proportion of the number of the elderly and children, and the GDP per unit area within the assessment unit are used as indicators to divide the vulnerability level of the disaster-bearing bodies. The vulnerability level is divided into four levels: "high, medium-high, medium-low, low", and the corresponding values are 1, 2, 3, and 4 respectively.

Images