Joint monitoring method and system for forest surface fire and crown fire based on multidimensional data

Abstract

The present invention discloses a method and system for joint monitoring of forest surface fires and crown fires based on multidimensional data. By understanding the past fire occurrences and frequencies in the monitoring area through historical fire data and historical geographic data, it helps to eliminate some interference data and interference situations in subsequent real-time monitoring. The influence of various fire risk factors on the occurrence and progress of fires obtained through a hierarchical analysis model and a fuzzy evaluation model is combined with real-time environmental data to generate a fire dynamic trend, which can reflect the possibility of fire occurrence and progress in the environment of the monitoring area. By comprehensively analyzing the real-time monitoring data of the monitoring area and the fire dynamic trend, the fire dynamic trend can be used to eliminate interference factors in the real-time monitoring data, thereby improving the monitoring and recognition rate of forest surface fires and crown fires, and overcoming the serious problem of missed and false alarms in existing forest fire monitoring methods.

Description

Technical Field

[0001] The present invention relates to the technical field of forest fire monitoring, and in particular to a method and system for jointly monitoring forest surface fires and crown fires based on multidimensional data. Background Art

[0002] Forest fires vary over time and space, and are divided into surface fires, crown fires, and mountain fires. The core issue in forest fire monitoring and early warning is early detection. Satellite remote sensing, video surveillance, and IoT sensing methods each have limitations in forest fire monitoring and early warning. For example, video surveillance has a low smoke detection rate, resulting in a high incidence of false alarms and missed targets. Furthermore, IoT sensing devices in the forest floor have limited coverage, making it difficult to detect crown fires.

[0003] There is an urgent need for a method to monitor forest surface fires and crown fires by combining multidimensional data. Summary of the Invention

[0004] In response to the problems raised by the background technology, the purpose of the present invention is to provide a joint monitoring method for forest surface fires and crown fires based on multidimensional data, which solves the problem that the existing forest fire monitoring methods have certain limitations, resulting in low monitoring and recognition rates for forest surface fires and crown fires, and serious missed and false alarms.

[0005] The present invention is achieved through the following technical solutions:

[0006] A first aspect of the present invention provides a method for jointly monitoring forest surface fires and crown fires based on multidimensional data, comprising the following steps:

[0007] Step S1: Acquire historical fire data and historical geographic data, perform comprehensive analysis on the historical fire data and the historical geographic data, and obtain a fire risk factor; wherein the fire risk factor includes a fire risk occurrence factor and a fire risk spread factor;

[0008] Step S2: constructing a hierarchical analysis model, inputting the fire risk occurrence factor and the fire risk spread factor into the hierarchical analysis model for analysis, and obtaining a fire risk weight;

[0009] Step S3: constructing a fuzzy evaluation model, inputting the fire risk weight into the fuzzy evaluation model for analysis, and obtaining a fire risk impact index;

[0010] Step S4: acquiring real-time environmental data of the monitored area, simulating the triggering mechanism of surface fire and crown fire, comprehensively analyzing the real-time environmental data and the fire impact index, and generating a fire dynamic trend;

[0011] Step S5: Acquire real-time monitoring data of the monitoring area, conduct a comprehensive analysis of the real-time monitoring data and the fire dynamic trend, and generate a surface fire and crown fire distribution map.

[0012] In the above technical solution, analyzing historical fire data helps understand the past occurrence and frequency of fires in the monitored area, helping to eliminate interference data and situations in subsequent real-time monitoring, thereby improving monitoring accuracy. Using a tomographic analysis model and a fuzzy evaluation model, fire risk factors and fire spread factors are analyzed, their weights determined, and a fire risk impact index (which determines the magnitude of their impact on fire) is used to accurately analyze the forest environment during subsequent simulations.

[0013] The dynamic trend of fire is generated by combining historical fire data, historical geographic data, hierarchical analysis model, and fuzzy evaluation model to obtain the degree of influence of various fire risk factors on the occurrence and progress of fire. It is generated based on real-time environmental data and can reflect the possibility of fire occurrence and progress in the monitored area environment.

[0014] By comprehensively analyzing the real-time monitoring data of the monitored area and the dynamic trends of fires, the interference factors in the real-time monitoring data can be eliminated through the dynamic trends of fires, thereby improving the monitoring and recognition rate of forest surface fires and crown fires, and overcoming the serious problems of missed reports and false alarms in existing forest fire monitoring methods.

[0015] In an optional embodiment, performing a comprehensive analysis based on the historical fire data and the historical geographic data to obtain a fire risk factor includes the following steps:

[0016] Step S11, dividing the historical fire data into historical forest fire data and historical non-forest fire data according to the burning range of the historical fire data and the historical geographical data;

[0017] Step S12: constructing a spatial clustering model, inputting the historical forest fire data into the spatial clustering model for analysis to obtain fire spatial correlation; wherein the fire spatial correlation includes global fire spatial correlation and local fire spatial correlation;

[0018] Step S13: classifying fire influencing factors into positive correlation factors, negative correlation factors, and unrelated factors according to the global fire spatial correlation;

[0019] Step S14: Based on the spatial correlation of local fires, the positive correlation factor and the negative correlation factor are analyzed to obtain a fire risk occurrence factor and a fire risk spread factor.

[0020] In an optional embodiment, constructing a hierarchical analysis model includes: taking forest fire risk analysis as the first level, taking fire occurrence and fire spread as the second level, taking social factors, meteorological factors, vegetation factors and terrain factors as the third level, and taking the fire risk factors as the fourth level.

[0021] In an optional embodiment, inputting the fire risk occurrence factor and the fire risk spread factor into the hierarchical analysis model for analysis to obtain the fire risk weight includes the following steps:

[0022] Step S21: Calculate the relative importance of each level element in the hierarchical analysis model using a dominance analysis method;

[0023] Step S22: constructing a judgment matrix using a consistent matrix method according to the relative importance;

[0024] Step S23: performing hierarchical single sorting on the judgment matrix to obtain hierarchical single sorting weights of each level in the hierarchical analysis model;

[0025] Step S24: comprehensively calculate the hierarchical single ranking weights of each level in the hierarchical analysis model to obtain the fire risk weight.

[0026] In an optional embodiment, before using the dominance analysis method to calculate the relative importance between the elements of each level in the hierarchical analysis model, the method further includes analyzing the correlation of the fire risk factors.

[0027] In an optional embodiment, constructing a fuzzy evaluation model includes: taking the fire risk factor as a factor set, taking the fire risk development state as a level set, and taking the fire risk weight as a weight set; wherein the fire risk development state includes steady surface fire, rapid surface fire, intermittent crown fire, steady crown fire and rapid crown fire.

[0028] In an optional embodiment, inputting the fire risk weight into the fuzzy evaluation model for analysis to obtain the fire risk impact index comprises the following steps:

[0029] Step S31, constructing a membership function based on the historical forest fire data;

[0030] Step S32: determining the membership degree of the fire risk factor to the fire risk development state according to the membership function;

[0031] Step S33: Combining the membership degree and the fire risk weight to obtain a fire risk impact index.

[0032] In an optional embodiment, the real-time monitoring data includes smoke data, sound data and fire point data.

[0033] In an optional embodiment, performing a comprehensive analysis of the real-time monitoring data and the fire dynamic trend to generate a surface fire crown fire distribution map includes the following steps:

[0034] Step S51: Analyze the smoke data and determine the first fire progress by the smoke color; analyze the sound data and determine the second fire progress by the sound spectrum; analyze the fire point data and determine the third fire progress by the fire point status;

[0035] Step S52: Correcting the first fire process, the second fire process, and the third fire process respectively according to the fire dynamic trend, and coupling the corrected first fire process, the second fire process, and the third fire process to obtain a final fire process;

[0036] Step S53: Mark the final fire process on a map to generate a surface fire and crown fire distribution map.

[0037] A second aspect of the present invention provides a combined forest surface fire and crown fire monitoring system based on multidimensional data, comprising:

[0038] A fire risk factor module, which is used to obtain historical fire data and historical geographic data, and conduct a comprehensive analysis based on the historical fire data and the historical geographic data to obtain a fire risk factor; wherein the fire risk factor includes a fire occurrence factor and a fire spread factor;

[0039] A hierarchy analysis module, wherein the hierarchy analysis module is used to construct a hierarchy analysis model, input the fire risk occurrence factor and the fire risk spread factor into the hierarchy analysis model for analysis, and obtain a fire risk weight;

[0040] A fuzzy evaluation module is used to construct a fuzzy evaluation model, input the fire risk weight into the fuzzy evaluation model for analysis, and obtain a fire risk impact index;

[0041] A fire trend module is used to obtain real-time environmental data of the monitored area, simulate the triggering mechanism of surface fire and crown fire, conduct a comprehensive analysis of the real-time environmental data and the fire impact index, generate a fire dynamic trend, and construct a fire trend index based on the fire dynamic trend;

[0042] The fire monitoring module is used to obtain real-time monitoring data of the monitoring area, conduct a comprehensive analysis of the real-time monitoring data and the fire trend index, and generate a surface fire and crown fire distribution map.

[0043] Compared with the prior art, this application has the following advantages and beneficial effects:

[0044] Analyzing historical fire data helps understand past fire occurrences and frequency in the monitored area, helping to eliminate interference data and situations in subsequent real-time monitoring, thereby improving monitoring accuracy. Using a tomographic analysis model and a fuzzy evaluation model, we analyze fire risk factors and fire spread factors, determine their weights, and determine the fire risk impact index, which determines the magnitude of their impact on fires. This allows for accurate analysis of the forest environment during subsequent simulations.

[0045] The dynamic trend of fire is generated by combining historical fire data, historical geographic data, hierarchical analysis model, and fuzzy evaluation model to obtain the degree of influence of various fire risk factors on the occurrence and progress of fire. It is generated based on real-time environmental data and can reflect the possibility of fire occurrence and progress in the monitored area environment.

[0046] By comprehensively analyzing the real-time monitoring data of the monitored area and the dynamic trends of fires, the interference factors in the real-time monitoring data can be eliminated through the dynamic trends of fires, thereby improving the monitoring and recognition rate of forest surface fires and crown fires, and overcoming the serious problems of missed reports and false alarms in existing forest fire monitoring methods. BRIEF DESCRIPTION OF THE DRAWINGS

[0047] In order to more clearly illustrate the technical solutions of the exemplary embodiments of the present invention, the following briefly introduces the drawings required for use in the examples. It should be understood that the following drawings only illustrate certain embodiments of the present invention and should not be considered as limiting the scope. A person of ordinary skill in the art can also derive other relevant drawings based on these drawings without inventive effort. In the drawings:

[0048] Figure 1 A schematic flow chart of a method for jointly monitoring forest surface fires and crown fires based on multidimensional data provided in Example 1 of the present invention;

[0049] Figure 2 A schematic diagram of constructing a fire risk factor according to Example 1 of the present invention;

[0050] Figure 3 A structural diagram of a hierarchical analysis model is provided for Example 1 of the present invention. DETAILED DESCRIPTION

[0051] In order to make the objectives, technical solutions and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with examples and drawings. The exemplary embodiments of the present invention and their descriptions are only used to explain the present invention and are not intended to limit the present invention.

[0052] Example 1

[0053] Figure 1 A flow chart of a method for joint monitoring of forest surface fire and crown fire based on multidimensional data provided in Example 1 of the present invention is shown as follows: Figure 1 As shown in FIG, the joint monitoring method of forest surface fire and crown fire based on multidimensional data includes the following steps:

[0054] Step S1: Acquire historical fire data and historical geographic data, perform comprehensive analysis on the historical fire data and the historical geographic data, and obtain a fire risk factor.

[0055] Figure 2 This is a schematic diagram of the construction of the fire risk factor provided in Example 1 of the present invention, as shown in FIG. Figure 2 shown.

[0056] The fire risk factors include fire occurrence factors and fire spread factors.

[0057] In the field of forest fire monitoring, historical fire data and historical geographic data are of great significance for fire monitoring in the monitoring area. Among them, the occurrence of forest fires is not uniformly distributed in time and space.

[0058] From a temporal perspective, forest fires have a distinct seasonal distribution, and their occurrence is closely related to changes in solar radiation throughout the day. Spatially, forest fires are also unevenly distributed due to factors such as topography, weather, and vegetation distribution.

[0059] In summary, analyzing historical fire data is helpful for understanding the past occurrence and frequency of fires in the monitored area, and helps to eliminate some interference data and interference situations in subsequent real-time monitoring, thereby improving the accuracy of monitoring.

[0060] Specifically, the comprehensive analysis of the historical fire data and the historical geographic data to obtain the fire risk factor includes the following steps:

[0061] Step S11 : dividing the historical fire data into historical forest fire data and historical non-forest fire data according to the burning range of the historical fire data and the historical geographical data.

[0062] Among them, the fire location in the historical fire data may be a non-forest area. In order to improve the accuracy of forest fire identification, it is necessary to filter out the historical fire data belonging to non-forest areas. Therefore, in the present invention, the non-forest area fires in the historical fire data are excluded through historical geographic data to improve the accuracy of forest area fire analysis.

[0063] It should be noted that in this disclosure, non-forest areas are limited to fires that occurred entirely within non-forest areas, and do not include fires that started in non-forest areas and subsequently spread to forest areas. Therefore, the classification method for dividing historical fire data into historical forest fire data and historical non-forest fire data is as follows: if the land cover type within the burned area does not include either forestland or grassland, the historical fire data is classified as historical non-forest fire data; otherwise, it is classified as historical forest fire data.

[0064] Step S12: construct a spatial clustering model, input the historical forest fire data into the spatial clustering model for analysis, and obtain the fire spatial correlation.

[0065] The fire spatial correlation includes global fire spatial correlation and local fire spatial correlation.

[0066] Due to the correlation between spaces, existing analyses of fire impacts are limited to analyzing the influencing factors of terrain, ignoring the correlation between spaces at the geographic level. However, for monitoring forest surface fires and crown fires, spatial correlations are crucial for analyzing their spread. These correlations determine whether a fire can spread and whether a surface fire is likely to transform into a crown fire. Therefore, the present invention uses spatial correlations as a basis for screening fire risk spread factors.

[0067] Specifically, the spatial clustering model includes global spatial clustering and local spatial clustering.

[0068] In this paper, the Moran's I index is selected as the research indicator for global spatial clustering, the Z statistic is selected as the test indicator, and the number of fires in historical forest fire data is used as the analysis data. The Moran's I index is calculated based on the historical forest fire data under various fire influencing factors.

[0069] If the Moran's I index is less than zero, it indicates that there is a global negative spatial correlation in the occurrence of forest fires under this influencing factor. If the Moran's I index is greater than zero, it indicates that there is a global positive spatial correlation in the occurrence of forest fires under this influencing factor. If the Moran's I index is equal to zero, it indicates that there is no global spatial correlation in the occurrence of forest fires under this influencing factor. The Moran's I index can be used to obtain the global spatial correlation of fires.

[0070] The Moran's I index was expressed in scatter plots and saliency maps, and these were analyzed to determine the local spatial clustering patterns of fires. These clustering patterns include "high-high," "high-low," "low-high," and "low-low." These clustering patterns are considered the spatial correlation of local fires.

[0071] Step S13: Classify fire influencing factors into positive correlation factors, negative correlation factors and unrelated factors according to the fire spatial correlation.

[0072] According to the global spatial correlation of fire, the factors affecting the occurrence of forest fire with global negative spatial correlation are defined as positive correlation factors, the factors affecting the occurrence of forest fire with global negative spatial correlation are defined as negative correlation factors, and the factors affecting the occurrence of forest fire with no global spatial correlation are defined as irrelevant factors.

[0073] Step S14: Based on the spatial correlation of local fires, the positive correlation factor and the negative correlation factor are analyzed to obtain a fire risk occurrence factor and a fire risk spread factor.

[0074] Among them, in the local fire spatial correlation with a clustering form of "high", it means that the attributes of the studied area are similar to those of the adjacent areas, and both are areas with high fire clustering. Under this clustering form, forest fires are prone to both occurrence and spread. Therefore, the corresponding positive correlation factors and negative correlation factors are defined as fire risk occurrence factors and fire risk spread factors at the same time.

[0075] In the local fire spatial correlation with the clustering form of "high and low", it means that the attributes of the study area are opposite to those of the adjacent areas. The study area is an area with high fire clustering, and the areas surrounding the study area are areas with low fire clustering. Under this clustering situation, forest fires are prone to occur, but not prone to spread. Therefore, the corresponding positive and negative correlation factors are defined as fire risk factors.

[0076] In the local fire spatial correlation with the clustering form of "low-high", it means that the attributes of the study area are opposite to those of the adjacent areas. The study area is an area with low fire clustering, and the areas surrounding the study area are areas with high fire clustering. Under this clustering situation, forest fires are prone to spread, but fires are not likely to occur. Therefore, the corresponding positive and negative correlation factors are defined as fire risk spread factors.

[0077] In the local fire spatial correlation with a clustering form of "low", it means that the attributes of the studied area are similar to those of the adjacent areas, and both are areas with low fire clustering. Under this clustering form, it is not easy for forest fires to occur or spread.

[0078] Based on the above situation, the fire attributes of positive and negative correlation factors are assigned through the spatial correlation of local fires, and they are divided into fire risk occurrence factors and fire risk spread factors. This step is of great significance for analyzing the progress of fire risk.

[0079] It should be noted that the influencing factors can be either fire risk occurrence factors or fire risk spread factors, and their classification is not an exclusive state.

[0080] Step S2: constructing a hierarchical analysis model, inputting the fire risk occurrence factor and the fire risk spread factor into the hierarchical analysis model for analysis, and obtaining a fire risk weight.

[0081] Among them, constructing a hierarchical analysis model includes: taking forest fire risk analysis as the first level, fire occurrence and fire spread as the second level, social factors, meteorological factors, vegetation factors and terrain factors as the third level, and the fire risk factors as the fourth level.

[0082] Figure 3 A schematic diagram of the structure of a hierarchical analysis model is provided for Example 1 of the present invention, as shown in FIG. Figure 3 As shown in the figure, the first layer is forest fire risk analysis, the second layer is fire occurrence and spread, the third layer is social factors, meteorological factors, vegetation factors, and topographic factors, and the fourth layer is fire risk factors. Fire risk factors under social factors include road location and building location, meteorological factors include air pressure, temperature, relative humidity, precipitation, wind speed, and daily duration, vegetation factors include forest volume and vegetation type, and topographic factors include slope, aspect, slope position, and altitude.

[0083] It should be noted that since forest fire risk analysis is the goal of constructing a hierarchical analysis model, it is considered the first level. The core purpose of this invention is to monitor the occurrence and progression of forest surface fires and crown fires, so fire occurrence and spread are considered the second level. Fire occurrence and spread are influenced by social factors, meteorological factors, vegetation factors, and topographic factors, so they are considered the third level. Specific fire risk factors among these social, meteorological, vegetation, and topographic factors are considered the fourth level. The aforementioned structural component hierarchical analysis model is integrated to achieve the calculation of fire risk weights.

[0084] Specifically, inputting the fire risk occurrence factor into the hierarchical analysis model for analysis to obtain the fire risk weight of the fire risk factor includes the following steps:

[0085] Step S21: Calculate the relative importance of the elements at each level in the hierarchical analysis model using a dominance analysis method.

[0086] Among them, the relative importance includes the relative importance between the location of the highway and the location of the building under social factors; the relative importance between air pressure, temperature, relative humidity, precipitation, wind speed and daily duration under meteorological factors; the relative importance between forest volume and vegetation type under vegetation factors; the relative importance between slope, aspect, position and altitude under terrain factors; the relative importance between social factors, meteorological factors, vegetation factors and terrain factors under rapid crown fire; the relative importance between social factors, meteorological factors, vegetation factors and terrain factors under steady crown fire. Importance; the relative importance of social factors, meteorological factors, vegetation factors, and terrain factors under rapid crown fires; the relative importance of social factors, meteorological factors, vegetation factors, and terrain factors under intermittent crown fires; the relative importance of social factors, meteorological factors, vegetation factors, and terrain factors under rapid surface fires; the relative importance of social factors, meteorological factors, vegetation factors, and terrain factors under steady surface fires; the importance of rapid crown fires, steady crown fires, intermittent crown fires, rapid surface fires, and steady surface fires in forest fire risk analysis.

[0087] The calculation formula of the advantage analysis method is as follows:

[0088]

[0089] in, Represents the lower element x i The average contribution value of , y represents the upper element, p represents the number of elements, x h represents the set of elements consisting of the k elements in the lower layer, is the contribution value, is the correlation coefficient with the upper layer elements.

[0090] In an optional embodiment, before using the dominance analysis method to calculate the relative importance between the elements of each level in the hierarchical analysis model, the method further includes analyzing the correlation of the fire risk factors.

[0091] like Figure 3 As shown, slope, aspect, position, and altitude have a certain influence on forest volume and vegetation type. There is a certain correlation between them, so they are considered indirect influencing factors under the vegetation factor. When calculating relative importance, it is also necessary to analyze the relative importance of sunshine duration, forest volume, vegetation type, slope, aspect, position, and altitude. Calculating the correlation of fire risk factors before calculating relative importance helps improve the hierarchical analysis model.

[0092] Step S22: constructing a judgment matrix using a consistent matrix method according to the relative importance.

[0093] Among them, the fire risk factor of the i-th row is compared with the fire risk factor of the j-th column, and a ij Defined as the matrix element in row i and column j, then Compare the fire risk factor in row j with the fire risk factor in column k, and jk is defined as the matrix element in row j and column k. ij and a jk Need to meet a ij *a jk =a ik .

[0094] The fire risk factors are compared pairwise to construct a judgment matrix.

[0095] Step S23: performing hierarchical single sorting on the judgment matrix to obtain hierarchical single sorting weights of each level in the hierarchical analysis model.

[0096] Normalize the judgment matrix to obtain the ranking weight of the relative importance of elements at the same level under the elements at the previous level.

[0097] Step S24: comprehensively calculate the hierarchical single ranking weights of each level in the hierarchical analysis model to obtain the fire risk weight.

[0098] The comprehensive calculation involves multiplying the ranking weights of the elements at each level to obtain the fire risk weight for that fire risk factor. For example, the fire risk weight for a highway location is calculated as the ranking weight of the highway location * the ranking weight of the social factor * the ranking weight of the aggressive crown fire.

[0099] Step S3: construct a fuzzy evaluation model, input the fire risk weight into the fuzzy evaluation model for analysis, and obtain a fire risk impact index.

[0100] Among them, constructing a fuzzy evaluation model includes: taking the fire risk factor as a factor set, taking the fire risk development state as a level set, and taking the fire risk weight as a weight set; wherein, the fire risk development state includes steady surface fire, rapid surface fire, intermittent crown fire, steady crown fire and rapid crown fire.

[0101] It should be noted that the core purpose of the invention is to monitor the occurrence and progress of forest surface fires and crown fires, so the fire risk development status is taken as a level set; since the fire risk factor is an influencing factor that affects the fire risk development status, it is taken as a factor set, and the fire risk weight corresponding to the fire risk factor is taken as a weight set.

[0102] Since the present invention is for monitoring forest surface fires and crown fires, the fire risk development status is divided into five types: steady surface fire, rapid surface fire, intermittent crown fire, steady crown fire and rapid crown fire. By analyzing the influence of fire risk factors and fire risk weights on the fire risk development status, the fire risk impact index of fire risk factors on the fire risk development status is obtained.

[0103] Specifically, inputting the fire risk weight into the fuzzy evaluation model for analysis to obtain the fire risk impact index includes the following steps:

[0104] Step S31: Constructing a membership function based on the historical forest fire data.

[0105] Among them, the fire risk factors obtained in step S14 can be divided into fire risk occurrence factors and fire risk spread factors according to the progress classification, and fire risk factors can be divided into positive correlation factors and negative correlation factors according to their effects on fire risk. The fire risk factors are traversed and the effects and progress of each fire risk factor in the historical forest fire data are counted in turn. The number of times the fire risk factor belongs to the fire risk occurrence factor in the historical forest fire data is recorded as FireHappen, the number of times it belongs to the fire risk spread factor is recorded as FireSpread, the number of times the positive correlation factor is recorded as CorPos, and the number of times the negative correlation factor is recorded as CorNeg. The forest fire intensity and flame spread speed corresponding to each fire risk factor in the historical forest fire data are recorded, and they are recorded as FI and FV respectively. The membership function is constructed by the classification and number of fire risk factors in the historical forest fire data. The membership function is constructed as follows:

[0106]

[0107] Where i represents the i-th fire risk factor, U i is the membership degree of the i-th fire risk factor.

[0108] When forest fire intensity is less than 750kW / m², surface fires cannot transition to crown fires. Within this range, the fire risk factor primarily belongs to the surface fire category, resulting in a higher degree of membership for both rapid and steady surface fires and a lower degree of membership for crown fires. When forest fire intensity ranges from 750kW / m² to 3500kW / m², transition is possible, requiring analysis of the flame spread rate. A flame spread rate less than 5km / h is defined as low, while a flame spread rate greater than 5km / h is defined as medium. Fire intensity greater than 3500kW / m² is defined as high. At this point, the membership between crown fire and surface fire requires consideration of both positive and negative correlation factors. Positive correlation factors may increase the likelihood of fire spread, leading to a higher probability of rapid and steady crown fires.

[0109] Step S32: determining the membership degree of the fire risk factor to the fire risk development state according to the membership function.

[0110] Substituting specific data into the above membership function can determine the membership degree corresponding to the fire risk factor.

[0111] Step S33: Combining the membership degree and the fire risk weight to obtain a fire risk impact index.

[0112] The fire risk impact index of the fire risk factor is obtained by multiplying the membership degree and the fire risk weight.

[0113] Step S4: Acquire real-time environmental data of the monitored area, simulate the triggering mechanism of surface fire and crown fire, conduct a comprehensive analysis of the real-time environmental data and the fire impact index, and generate a fire dynamic trend.

[0114] Among them, the purpose of the present invention is to monitor forest fire conditions in real time. In order to solve the inaccuracy of monitoring fire conditions in the monitoring area in the existing technology, the present invention analyzes and processes historical forest fire data to obtain the influence index of each fire risk factor on the occurrence and progress of fire.

[0115] Acquire real-time environmental data from the monitored area. By simulating the triggering mechanisms of surface and crown fires, analyze this data and the Fire Impact Index to determine if a fire has occurred and how it is progressing. This step aims to capture the dynamics of fires under specific conditions, providing an environmental basis for subsequent real-time monitoring.

[0116] Specifically, the triggering mechanism of surface fire and crown fire is as follows:

[0117] When the forest fire intensity is less than 750kW / m 2When the fire intensity is 750kW / m 2 Up to 3500kW / m 2 When the flame spread speed is less than 5km / h, it is defined as low spread probability; when the flame spread speed is above 5km / h, it is defined as medium spread probability; when the forest fire intensity is greater than 3500kW / m 2 , it is defined as having a high probability of spread.

[0118] Step S5: Acquire real-time monitoring data of the monitoring area, conduct a comprehensive analysis of the real-time monitoring data and the fire dynamic trend, and generate a surface fire and crown fire distribution map.

[0119] The real-time monitoring data includes smoke data, sound data and fire point data.

[0120] In existing technologies, forest fire monitoring is typically performed using video surveillance and understory IoT sensing devices. However, video surveillance has a low smoke recognition rate and is prone to false alarms and missed alarms. Furthermore, understory IoT sensing devices have low coverage and are unable to monitor crown fires. The core concept of this invention is to collect real-time environmental data to simulate the current forest environment, combining smoke data, sound data, and fire point data to monitor forest fire status. This multi-dimensional, real-time monitoring data and the forest environment improve the monitoring capability and accuracy of forest surface and crown fires, reducing false alarms and missed alarms.

[0121] Step S51: Analyze the smoke data and determine the first fire progress by the smoke color; analyze the sound data and determine the second fire progress by the sound spectrum; analyze the fire point data and determine the third fire progress by the fire point status.

[0122] Forest fire smoke can come in various colors, including white, black, and mixed. Smoke color generally categorizes fire progression into surface fire, crown fire, and surface-to-crown fire. In dense forests, or in low-density forests where crown fires are absent, burning primarily consists of understory weeds and dead branches and leaves, creating thick white smoke. Due to the high oil content of combustible materials, typically found in dense pine forests, the formation of black smoke indicates a crown fire. Therefore, if white smoke is detected via video, the initial fire progression is surface fire; if black smoke is detected, the initial fire progression is crown fire; and if mixed smoke is detected, the initial fire progression is surface fire-to-crown fire.

[0123] Burning trees produce sound. By analyzing the frequency-amplitude waveform of the sound produced by burning trees, we can categorize fire processes into three types: surface fire, crown fire, and surface fire transitioning to crown fire. This classification method is not a core concept of this invention, so we will use existing classification methods.

[0124] Fire point data can reflect the flame height and flame range. The flame height is positively correlated with forest fire intensity. Therefore, the fire progress can be reflected by fire point data, and the third fire progress can be determined by the flame height and flame range.

[0125] Step S52: Correct the first fire process, the second fire process, and the third fire process respectively according to the fire dynamic trend, and couple the corrected first fire process, the second fire process, and the third fire process to obtain a final fire process.

[0126] Among them, the dynamic trend of fire is the degree of influence of various fire risk factors on the occurrence and progress of fire obtained through historical fire data and historical geographical data through hierarchical analysis model and fuzzy evaluation model. It is generated on the basis of real-time environmental data. It can reflect the possibility of fire occurrence and progress in the environment of the monitored area.

[0127] Smoke data, sound data, and fire point data are all affected and interfered by certain environmental factors. The present invention uses the fire dynamic trend to correct the first fire process, the second fire process, and the third fire process respectively to eliminate the interference of various environmental factors on fire identification. The correction factor is calculated as follows:

[0128]

[0129] Wherein, P is the probability of occurrence of fire process, which is used as a correction factor in the present invention, PI is a positive correlation factor, PN is a negative correlation factor, n is the number of correlation factors, and m is the number of positive correlation factors.

[0130] The coupling process is as follows:

[0131] The first, second, and third fire processes were quantified. The quantification method was to assign fixed values of 30, 25, 20, 15, and 10 to each fire process, respectively, for rapid crown fire, steady crown fire, intermittent crown fire, rapid surface fire, and steady surface fire, based on the fire process's progress. Values within the range of [10, 15] were considered steady surface fires, within the range of [15, 20] for rapid surface fires, within the range of [20, 25] for intermittent crown fires, within the range of [25, 30] for steady crown fires, and values above 30 for rapid crown fires. After correcting the first, second, and third fire processes using the correction factor, they were compared pairwise. If they fell within the same range, the coupling was successful; if not, the coupling failed. The successfully coupled fire process was considered the final fire process. If all failed, the data needed to be retested.

[0132] Step S53: Mark the final fire process on a map to generate a surface fire and crown fire distribution map.

[0133] Example 2

[0134] Embodiment 2 of the present invention provides a joint monitoring system for forest surface fires and crown fires based on multidimensional data. The joint monitoring system includes:

[0135] The fire risk factor module is used to obtain historical fire data and historical geographic data, and conduct a comprehensive analysis based on the historical fire data and the historical geographic data to obtain a fire risk factor; wherein the fire risk factor includes a fire occurrence factor and a fire spread factor.

[0136] A hierarchical analysis module is used to construct a hierarchical analysis model, input the fire risk occurrence factor and the fire risk spread factor into the hierarchical analysis model for analysis, and obtain the fire risk weight.

[0137] The fuzzy evaluation module is used to construct a fuzzy evaluation model, input the fire risk weight into the fuzzy evaluation model for analysis, and obtain a fire risk impact index.

[0138] A fire trend module is used to obtain real-time environmental data of the monitored area, simulate the triggering mechanism of surface fire and crown fire, conduct a comprehensive analysis of the real-time environmental data and the fire impact index, generate a fire dynamic trend, and construct a fire trend index based on the fire dynamic trend.

[0139] The fire monitoring module is used to obtain real-time monitoring data of the monitoring area, conduct a comprehensive analysis of the real-time monitoring data and the fire trend index, and generate a surface fire and crown fire distribution map.

[0140] The specific implementation methods described above further illustrate the objectives, technical solutions and beneficial effects of the present invention in detail. It should be understood that the above description is only a specific implementation method of the present invention and is not intended to limit the scope of protection of the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. A joint monitoring method for forest surface fire and crown fire based on multidimensional data, characterized in that: The steps include: Step S1: Acquire historical fire data and historical geographic data, perform comprehensive analysis on the historical fire data and the historical geographic data, and obtain a fire risk factor; wherein the fire risk factor includes a fire risk occurrence factor and a fire risk spread factor; Step S2: constructing a hierarchical analysis model, inputting the fire risk occurrence factor and the fire risk spread factor into the hierarchical analysis model for analysis, and obtaining a fire risk weight; Step S3: constructing a fuzzy evaluation model, inputting the fire risk weight into the fuzzy evaluation model for analysis, and obtaining a fire risk impact index; Step S4: acquiring real-time environmental data of the monitored area, simulating the triggering mechanism of surface fire and crown fire, comprehensively analyzing the real-time environmental data and the fire impact index, and generating a fire dynamic trend; Step S5: Acquire real-time monitoring data of the monitoring area, conduct a comprehensive analysis of the real-time monitoring data and the fire dynamic trend, and generate a surface fire and crown fire distribution map.

2. The method for joint monitoring of forest surface fire and crown fire based on multidimensional data according to claim 1, characterized in that: Combining the historical fire data and the historical geographic data for comprehensive analysis to obtain a fire risk factor includes the following steps: Step S11, dividing the historical fire data into historical forest fire data and historical non-forest fire data according to the burning range of the historical fire data and the historical geographical data; Step S12: constructing a spatial clustering model, inputting the historical forest fire data into the spatial clustering model for analysis to obtain fire spatial correlation; wherein the fire spatial correlation includes global fire spatial correlation and local fire spatial correlation; Step S13: classifying fire influencing factors into positive correlation factors, negative correlation factors, and unrelated factors according to the global fire spatial correlation; Step S14: Based on the spatial correlation of local fires, the positive correlation factor and the negative correlation factor are analyzed to obtain a fire risk occurrence factor and a fire risk spread factor.

3. The method for joint monitoring of forest surface fire and crown fire based on multidimensional data according to claim 1, characterized in that: Constructing a hierarchical analysis model includes: taking forest fire risk analysis as the first level, taking fire occurrence and fire spread as the second level, taking social factors, meteorological factors, vegetation factors and terrain factors as the third level, and taking the fire risk factors as the fourth level.

4. The method for joint monitoring of forest surface fire and crown fire based on multidimensional data according to claim 3, characterized in that: Inputting the fire risk occurrence factor and the fire risk spread factor into the hierarchical analysis model for analysis to obtain the fire risk weight includes the following steps: Step S21: Calculate the relative importance of each level element in the hierarchical analysis model using a dominance analysis method; Step S22: constructing a judgment matrix using a consistent matrix method according to the relative importance; Step S23: performing hierarchical single sorting on the judgment matrix to obtain hierarchical single sorting weights of each level in the hierarchical analysis model; Step S24: comprehensively calculate the hierarchical single ranking weights of each level in the hierarchical analysis model to obtain the fire risk weight.

5. The method for joint monitoring of forest surface fire and crown fire based on multidimensional data according to claim 4, characterized in that: Before using the dominance analysis method to calculate the relative importance of the elements at each level in the hierarchical analysis model, the method also includes analyzing the correlation of the fire risk factors.

6. The method for joint monitoring of forest surface fire and crown fire based on multidimensional data according to claim 2, characterized in that: Constructing a fuzzy evaluation model includes: taking the fire risk factor as a factor set, taking the fire risk development state as a level set, and taking the fire risk weight as a weight set; wherein the fire risk development state includes steady surface fire, rapid surface fire, intermittent crown fire, steady crown fire and rapid crown fire.

7. The method for joint monitoring of forest surface fire and crown fire based on multidimensional data according to claim 6, characterized in that: Inputting the fire risk weight into the fuzzy evaluation model for analysis to obtain the fire risk impact index comprises the following steps: Step S31, constructing a membership function based on the historical forest fire data; Step S32: determining the membership degree of the fire risk factor to the fire risk development state according to the membership function; Step S33: Combining the membership degree and the fire risk weight to obtain a fire risk impact index.

8. The method for joint monitoring of forest surface fire and crown fire based on multidimensional data according to claim 1, characterized in that: The real-time monitoring data includes smoke data, sound data and fire point data.

9. The method for joint monitoring of forest surface fire and crown fire based on multidimensional data according to claim 8, characterized in that: Comprehensively analyzing the real-time monitoring data and the fire dynamic trend to generate a surface fire crown fire distribution map includes the following steps: Step S51: Analyze the smoke data and determine the first fire progress by the smoke color; analyze the sound data and determine the second fire progress by the sound spectrum; analyze the fire point data and determine the third fire progress by the fire point status; Step S52: Correcting the first fire process, the second fire process, and the third fire process respectively according to the fire dynamic trend, and coupling the corrected first fire process, the second fire process, and the third fire process to obtain a final fire process; Step S53: Mark the final fire process on a map to generate a surface fire and crown fire distribution map.

10. A joint monitoring system for forest surface fire and crown fire based on multidimensional data, characterized by: include: A fire risk factor module, which is used to obtain historical fire data and historical geographic data, and conduct a comprehensive analysis based on the historical fire data and the historical geographic data to obtain a fire risk factor; wherein the fire risk factor includes a fire occurrence factor and a fire spread factor; A hierarchy analysis module, wherein the hierarchy analysis module is used to construct a hierarchy analysis model, input the fire risk occurrence factor and the fire risk spread factor into the hierarchy analysis model for analysis, and obtain a fire risk weight; A fuzzy evaluation module is used to construct a fuzzy evaluation model, input the fire risk weight into the fuzzy evaluation model for analysis, and obtain a fire risk impact index; A fire trend module is used to obtain real-time environmental data of the monitored area, simulate the triggering mechanism of surface fire and crown fire, conduct a comprehensive analysis of the real-time environmental data and the fire impact index, generate a fire dynamic trend, and construct a fire trend index based on the fire dynamic trend; The fire monitoring module is used to obtain real-time monitoring data of the monitoring area, conduct a comprehensive analysis of the real-time monitoring data and the fire trend index, and generate a surface fire and crown fire distribution map.

Images