A Forest Fire Risk Assessment System Based on Multi-Source Data

Abstract

This invention discloses a forest fire risk assessment system based on multi-source data, relating to the field of forest fire safety technology. It includes a data acquisition module, an information processing module, and a large database. The large database generates a dataset containing multiple fire risk characteristics based on historical forest fire data and manually input data. The data acquisition module includes a remote sensing image unit and multiple environmental monitoring units. The remote sensing image unit acquires remote sensing image data of the corresponding forest area and transmits it to the information processing module. The multiple environmental monitoring units are distributed in the corresponding forest area, acquire environmental data of the corresponding local area, and transmit the environmental data to the information processing module. This invention significantly improves the accuracy, real-time performance, and scientific rigor of forest fire monitoring and prevention by introducing multi-source data fusion technology, combined with a dynamic risk assessment model and risk-oriented analysis.

Description

Technical Field

[0001] This invention relates to the field of forest fire safety technology, specifically a forest fire risk assessment system based on multi-source data. Background Technology

[0002] Forest fires are a serious natural disaster that threatens the global ecological environment and socio-economic development. Their suddenness and rapid spread place extremely high demands on forest resource protection and disaster prevention. Traditional forest fire monitoring technologies mainly rely on single data sources, such as ground monitoring equipment, meteorological data, or remote sensing satellite imagery. While this method can provide information about fire occurrences to some extent, it has significant limitations.

[0003] A search revealed an invention patent (publication number: CN118861899A) that discloses a forest fire risk assessment system. This patent collects meteorological, topographical, and vegetation data in real time and employs advanced data fusion technology to process data from different sources in a unified manner, forming a comprehensive data model. The risk assessment model, constructed using machine learning algorithms, can accurately predict the probability and severity of fires.

[0004] In existing technologies, traditional methods rely on a single data source, making it difficult to comprehensively reflect the various inducing factors of fires. Furthermore, they are insufficient in terms of real-time performance and dynamism. Due to the lag in data acquisition and processing, the monitoring system cannot dynamically update fire risks and cannot reflect the latest status of fire risks in a timely manner. In particular, the system is not effective in early fire monitoring and warning. Therefore, this invention proposes a forest fire risk assessment system based on multi-source data. Summary of the Invention

[0005] The purpose of this invention is to provide a forest fire risk assessment system based on multi-source data to solve the problems mentioned in the background art.

[0006] This invention can be achieved through the following technical solution: a forest fire risk assessment system based on multi-source data, including a data acquisition module, an information processing module, and a large database;

[0007] The big data database generates a dataset containing multiple fire risk characteristics based on historical forest fire data and manually input data, and the big data database updates the dataset regularly.

[0008] The data acquisition module includes a remote sensing image unit and an environmental monitoring unit;

[0009] The remote sensing image unit is used to acquire remote sensing image data of the corresponding forest area and transmit it to the information processing module;

[0010] Multiple environmental monitoring units are distributed in corresponding forest areas, acquire environmental data of corresponding local areas, and transmit the environmental data to the information processing module;

[0011] The information processing module includes a receiving unit, a processing unit, and a modeling unit;

[0012] The modeling unit establishes a risk model based on the topography, environment, and vegetation types of the corresponding forest area by selecting fire risk characteristics from the dataset.

[0013] The receiving unit is used to receive remote sensing image data and environmental data from various environmental monitoring units, and after spatial and temporal matching of the environmental data from various environmental monitoring units with the remote sensing image data, it transmits the data to the processing unit.

[0014] The processing unit generates corresponding levels of influence factors based on the spatial and temporal matching results of various environmental data and remote sensing image data. At the same time, the processing unit matches various environmental data and remote sensing image data with the risk model. By the degree of matching between various environmental data and remote sensing image data and various fire risk features in the risk model, the number and correlation of actual risk features of the corresponding forest area are obtained.

[0015] In various environmental data and remote sensing image data, the more feature points that match the characteristics of fire risk, the higher the fire risk.

[0016] Furthermore, the correlation with actual risk characteristics includes:

[0017] Real-time data Falling into the corresponding data range threshold After that, the closer it gets... The greater the correlation, the stronger the correlation.

[0018] The processing unit conducts risk assessment based on the quantity and correlation of actual risk characteristics, combined with influencing factors.

[0019] A further technical improvement of the present invention is that: the information processing module divides the forest area into multiple identification areas based on the geographical location of each environmental monitoring unit;

[0020] The remote sensing image unit divides the remote sensing image data into multiple sub-image data based on each identification region;

[0021] When performing spatial and temporal matching, the receiving unit matches each sub-image data with the environmental data of the corresponding recognition area.

[0022] A further technical improvement of the present invention is that: the processing unit identifies whether the sub-image data of the corresponding identification area and the corresponding environmental data have the same actual risk characteristics, and classifies the impact factors into levels;

[0023] When the actual risk characteristics of sub-image data and environmental data in the same identification area are the same, the impact factor is level one;

[0024] When the actual risk characteristics of sub-image data and environmental data in the same identification area are different, the impact factor is level two;

[0025] Impact factors are used to regulate risk assessment, and the first-level impact factors have a greater regulatory effect on risk assessment than the second-level impact factors.

[0026] A further technical improvement of the present invention is that: the information processing module assigns values ​​to each fire risk feature in the risk model, and the information processing module calculates the influencing factors by combining the sub-image data of each identified area with the actual risk features in the environmental data, using the following formula:

[0027] ;

[0028] In the formula, R is the risk assessment value of the corresponding identified area;

[0029] n is the total number of actual risk characteristics;

[0030] The weight of the i-th actual risk feature represents the relative importance of that actual risk feature to the fire risk.

[0031] Assign a value to the i-th fire risk characteristic;

[0032] Let be the impact factor level for the i-th actual risk characteristic. In this embodiment, the impact factor levels include Level 1 and Level 2.

[0033] Let be the correlation degree of the i-th actual risk feature;

[0034] This is normal data corresponding to actual risk characteristics;

[0035] Real-time data for the i-th actual risk feature;

[0036] This represents the absolute distance between the actual risk characteristic and the normal data of the corresponding characteristic, indicating the degree to which the current actual risk characteristic deviates from the normal data of the corresponding characteristic.

[0037] The normalized range of the i-th actual risk feature is derived from historical forest fire data. It is used to limit the upper and lower limits of the corresponding risk feature and to standardize the deviation value.

[0038] A further technical improvement of the present invention is that: the information processing module establishes a risk assessment value distribution map matching the corresponding forest area based on the risk assessment value of each identified area;

[0039] Furthermore, the information processing module identifies the risk gradient of each identified area relative to its neighboring identified areas based on the risk assessment value, performs risk-oriented analysis, and determines whether identified areas with risk assessment values ​​higher than the risk threshold show a concentration trend and the direction of their spread through a preset risk threshold.

[0040] A further technical improvement of this invention lies in the following: During risk-oriented analysis in the information processing module, the formula for calculating the risk gradient is: ;

[0041] In the formula, To identify the risk gradient from region i to region j, i.e. the direction and intensity of risk change;

[0042] , These are the risk assessment values ​​for identified region j and identified region i, respectively.

[0043] Furthermore, the information processing module normalizes the risk gradient using the following formula:

[0044] ;

[0045] In the formula: The normalized risk gradient is represented by m, which is the number of all neighboring identification regions of identification region i. This represents the absolute value of the risk gradient.

[0046] Furthermore, the information processing module calculates the risk orientation of the corresponding forest area based on the normalized risk gradient, using the following formula:

[0047] ;

[0048] In the formula, To identify the wind direction vector of region i, indicating the direction of risk concentration or spread;

[0049] Let i be the set of neighboring recognition regions.

[0050] Let be the unit direction vector from recognition region i to recognition region j.

[0051] A further technical improvement of the present invention is that the risk-oriented analysis step performed by the information processing module includes:

[0052] S1. Extract the risk assessment value of each identified area from the risk assessment value distribution map. ;

[0053] S2. Calculate the risk gradient between each identification region and its neighboring identification regions. ;

[0054] S3, Risk gradient for each identified region Normalization is performed to obtain the normalized gradient. ;

[0055] S4. Based on the normalized gradient and unit direction vector Calculate the overall wind direction for each identified area. ;

[0056] S5, By assessing the risk value High-risk areas are identified by comparing them with preset risk thresholds.

[0057] S6. Generate risk guidance maps using visualization tools to mark the direction of risk concentration or spread.

[0058] A further technical improvement of the present invention is that: the information processing module is also used to... Determine the fire spread direction θ, and assess the fire spread rate R' considering wind direction changes, using the following formula:

[0059] ,

[0060] in, It is the reaction intensity, measured in kilowatts per square meter. ε is the fuel bed density, in kilograms per cubic meter; ε is the effective heating coefficient. It is the heat of combustion of fuel, measured in kilojoules per kilogram. It is a wind speed correction factor. =C'×W 2 , where C' is an empirical constant and W is the wind speed; It is the slope correction factor. =D'×tan(S), where D' is an empirical constant and S is the slope; It is a wind direction correction factor. =E×cos(θ-W d ), where E is an empirical constant, W d It's the wind direction.

[0061] More preferably, θ = γ × θ1 + β × θ2, where θ1 is The corresponding propagation direction, θ2 is the actual propagation direction, γ+β=1.

[0062] Compared with the prior art, the present invention has the following beneficial effects:

[0063] This invention significantly improves the accuracy, real-time performance, and scientific rigor of forest fire monitoring and prevention by introducing multi-source data fusion technology and combining it with dynamic risk assessment models and risk-oriented analysis. By integrating remote sensing image data with environmental monitoring data, a comprehensive assessment of fire risk is achieved. Furthermore, the cross-validation and supplementation of multi-source data significantly enhances the accuracy of risk assessment and reduces the risk of misjudgment and omission that may occur with a single data source.

[0064] Meanwhile, the risk distribution map generated by this invention based on each identified area can intuitively reflect the spatial distribution characteristics of forest fire risk. By visually presenting the risk assessment values ​​of different areas, high-risk areas can be quickly located, providing data support for the scientific allocation of fire prevention resources. This regionalized risk monitoring method helps to achieve refined management of fire risk.

[0065] Furthermore, by combining the calculation of risk gradient and risk direction vector, the system can effectively predict the spread direction and expansion trend of fire. Through quantitative analysis of risk gradient, it can identify the fire spread path, provide scientific guidance for designing fire isolation belts and deploying fire fighting forces, and improve fire emergency response capabilities.

[0066] The fire spread rate assessment scheme is based on the Rothermel model, takes into account wind direction changes, and combines the potential fire spread direction with the actual fire spread direction to more accurately predict the fire spread direction and speed trend, providing a scientific basis for fire prevention and emergency response. Attached Figure Description

[0067] To facilitate understanding by those skilled in the art, the present invention will be further described below with reference to the accompanying drawings.

[0068] Figure 1 This is a system block diagram of the present invention. Detailed Implementation

[0069] To further illustrate the technical means and effects of the present invention in achieving its intended purpose, the following detailed description of the specific implementation methods, structures, features, and effects of the present invention, in conjunction with the accompanying drawings and preferred embodiments, is provided.

[0070] Example 1

[0071] Please see Figure 1As shown, the present invention provides a forest fire risk assessment system based on multi-source data, including a data acquisition module, an information processing module, and a large database;

[0072] The big data database generates a dataset that includes multiple fire risk characteristics based on historical forest fire data and manually input data, and the big data database updates the dataset regularly.

[0073] Historical forest fire data includes fire type, corresponding forest topography (including topographic slope, land use type, and the impact of topography on fire spread speed and affected area), climate (including temperature, humidity, precipitation, wind speed, and wind direction), spatial distribution of fire, fire occurrence time, and forest vegetation characteristics (including the relationship between vegetation type, density, health status, etc. and fire spread). Manually input data covers meteorological changes and changes in remote sensing images.

[0074] The data acquisition module includes a remote sensing image unit and an environmental monitoring unit;

[0075] The remote sensing image unit is used to acquire remote sensing image data of the corresponding forest area and transmit it to the information processing module. The remote sensing image data includes:

[0076] Heat source data, using infrared imaging technology, identifies whether there are high-temperature areas in forest regions;

[0077] Remote sensing of smoke data to monitor whether smoke is spreading in the atmosphere;

[0078] NDVI vegetation index data is used to monitor the health status of forest vegetation and indirectly reflect humidity status by analyzing changes in the spectral characteristics of vegetation.

[0079] Multiple environmental monitoring units are distributed in corresponding forest areas, acquire environmental data of corresponding local areas, and transmit the environmental data to the information processing module;

[0080] Environmental data includes temperature data, humidity data, and localized smoke data;

[0081] The information processing module includes a receiving unit, a processing unit, and a modeling unit;

[0082] The modeling unit establishes a risk model based on the topography, environment, and vegetation types of the corresponding forest area by selecting fire risk characteristics from the dataset.

[0083] The receiving unit is used to receive remote sensing image data and environmental data from various environmental monitoring units, and after spatial and temporal matching of the environmental data from various environmental monitoring units with the remote sensing image data, it transmits the data to the processing unit.

[0084] Based on the spatial and temporal matching results of various environmental data and remote sensing image data, the processing unit generates influence factors at corresponding levels. At the same time, the processing unit matches various environmental data and remote sensing image data with the risk model. By the degree of matching between various environmental data and remote sensing image data and various fire risk features in the risk model, the number and correlation of actual risk features of the corresponding forest area are obtained.

[0085] In various environmental data and remote sensing image data, the more feature points that match the characteristics of fire risk, the higher the fire risk.

[0086] Each fire risk characteristic is set as a range threshold for the corresponding type of data. When various environmental data and remote sensing image data are used, when the corresponding real-time data are used... The threshold range of data falling into the corresponding category Once the feature points are successfully matched, they can be used as actual risk features.

[0087] Furthermore, the correlation with actual risk characteristics includes:

[0088] Real-time data Falling into the corresponding data range threshold After that, the closer it gets... The greater the correlation, the better, especially if the data is real-time. Data exceeding the threshold of the corresponding category If the correlation reaches its maximum, it indicates a significantly increased risk of fire.

[0089] The processing unit conducts risk assessment based on the quantity and correlation of actual risk characteristics, combined with influencing factors.

[0090] The information processing module divides the forest area into multiple identification zones based on the geographical location of each environmental monitoring unit;

[0091] The remote sensing image unit divides the remote sensing image data into multiple sub-image data based on each identification area;

[0092] When the receiving unit performs spatial and temporal matching, it matches each sub-image data with the environmental data of the corresponding recognition area.

[0093] The processing unit identifies whether the sub-image data of the corresponding identification area and the corresponding environmental data have the same actual risk characteristics, and classifies the impact factors into levels.

[0094] When the actual risk characteristics of the sub-image data and environmental data of the same identification area are the same, the influence factor is level one. That is, when a certain actual risk characteristic appears in both the sub-image data and environmental data of the corresponding identification area, its corresponding influence factor is level one.

[0095] When the actual risk characteristics of the sub-image data and environmental data of the same identification area are different, the impact factor is level two. When a certain actual risk characteristic does not appear in the sub-image data and environmental data of the corresponding identification area at the same time, its corresponding impact factor is level two.

[0096] Impact factors are used to regulate risk assessment, and the first-level impact factors have a greater regulatory effect on risk assessment than the second-level impact factors.

[0097] The information processing module assigns values ​​to each fire risk feature in the risk model, and combines the actual risk features in the sub-image data and environmental data of each identified area with the influencing factors to calculate the following formula:

[0098] ;

[0099] In the formula, R is the risk assessment value of the corresponding identified area;

[0100] n is the total number of actual risk characteristics;

[0101] The weight of the i-th actual risk feature represents the relative importance of that actual risk feature to the fire risk.

[0102] Assign a value to the i-th fire risk characteristic;

[0103] Let be the impact factor level for the i-th actual risk characteristic. In this embodiment, the impact factor levels include Level 1 and Level 2.

[0104] Let be the correlation degree of the i-th actual risk feature;

[0105] This is normal data corresponding to actual risk characteristics;

[0106] Real-time data for the i-th actual risk feature;

[0107] This represents the absolute distance between the actual risk characteristic and the normal data of the corresponding characteristic, indicating the degree to which the current actual risk characteristic deviates from the normal data of the corresponding characteristic.

[0108] The normalized range of the i-th actual risk feature is derived from historical forest fire data and is used to limit the upper and lower limits of the corresponding risk feature and to standardize the deviation value.

[0109] In this embodiment, the primary impact factor is set to 1, the secondary impact factor is set to 0.5, and the actual risk characteristics include:

[0110] temperature:

[0111] =0.8, =0.5, =0.4, =0.5, =0.9, =1;

[0112] humidity:

[0113] =0.4, =0.6, =0.3, =0.3, =0.7, =1;

[0114] smoke:

[0115] =0.9, =0.5, =0.5, =0.2, =1, =0.5;

[0116] The temperature section is as follows:

[0117] 0.5 × 0.8 × (0.9 × 1 + 0.3 / 0.4) = 0.5 × 0.8 × (0.9 + 0.75) = 0.66;

[0118] The humidity section is as follows:

[0119] 0.3 × 0.4 × (0.7 × 1 + 0.2 / 0.3) = 0.3 × 0.4 × (0.7 + 0.67) = 0.1632;

[0120] The smoke section is as follows:

[0121] 0.2 × 0.9 × (1.0 × 0.5 + 0.4 / 0.5) = 0.2 × 0.9 × (0.5 + 0.8) = 0.234;

[0122] In summary:

[0123] R = 0.66 + 0.1632 + 0.234 = 1.0572;

[0124] Based on the risk assessment values ​​of each identified area, the information processing module establishes a risk assessment value distribution map that matches the corresponding forest area.

[0125] Furthermore, the information processing module identifies the risk gradient of each identified area relative to its neighboring identified areas based on the risk assessment value, performs risk-oriented analysis, and determines whether identified areas with risk assessment values ​​higher than the risk threshold show a concentration trend and the direction of their spread through a preset risk threshold.

[0126] The steps involved in risk-oriented analysis by the information processing module include:

[0127] S1. Extract the risk assessment value of each identified area from the risk assessment value distribution map. ;

[0128] S2. Calculate the risk gradient between each identification region and its neighboring identification regions. ;

[0129] The formula for calculating the risk gradient is: ;

[0130] In the formula, To identify the risk gradient from region i to region j, i.e. the direction and intensity of risk change;

[0131] , These are the risk assessment values ​​for identified region j and identified region i, respectively.

[0132] S3, Risk gradient for each identified region Normalization is performed to obtain the normalized gradient. ;

[0133] The formula used is:

[0134] ;

[0135] In the formula: The normalized risk gradient is represented by m, which is the number of all neighboring identification regions of identification region i. This represents the absolute value of the risk gradient.

[0136] S4. Based on the normalized gradient and unit direction vector Calculate the overall wind direction for each identified area. ;

[0137] The formula is:

[0138] ;

[0139] In the formula, To identify the wind direction vector of region i, indicating the direction of risk concentration or spread;

[0140] Let i be the set of neighboring recognition regions.

[0141] Let be the unit direction vector from recognition region i to recognition region j.

[0142] S5, By assessing the risk value High-risk areas are identified by comparing them with preset risk thresholds.

[0143] S6. Generate risk guidance maps using visualization tools to mark the direction of risk concentration or spread.

[0144] In summary, the technical solution of the present invention has the following beneficial effects:

[0145] This invention significantly improves the accuracy, real-time performance, and scientific rigor of forest fire monitoring and prevention by introducing multi-source data fusion technology and combining it with dynamic risk assessment models and risk-oriented analysis. By integrating remote sensing image data with environmental monitoring data, a comprehensive assessment of fire risk is achieved. Furthermore, the cross-validation and supplementation of multi-source data significantly enhances the accuracy of risk assessment and reduces the risk of misjudgment and omission that may occur with a single data source.

[0146] Meanwhile, the risk distribution map generated by this invention based on each identified area can intuitively reflect the spatial distribution characteristics of forest fire risk. By visually presenting the risk assessment values ​​of different areas, high-risk areas can be quickly located, providing data support for the scientific allocation of fire prevention resources. This regionalized risk monitoring method helps to achieve refined management of fire risk.

[0147] Furthermore, by combining the calculation of risk gradient and risk direction vector, the system can effectively predict the spread direction and expansion trend of fire. Through quantitative analysis of risk gradient, it can identify the fire spread path, providing scientific guidance for designing firebreaks and deploying fire-fighting forces, and improving fire emergency response capabilities.

[0148] Example 2

[0149] In practical applications, the models may be more complex, as microclimate factors such as real-time changes in wind direction can also affect the fire spread rate. Therefore, considering wind direction changes, and based on the potential fire spread direction obtained in Example 1, in this example, the information processing module of the forest fire risk assessment system based on multi-source data further evaluates the fire spread rate R', as shown in the following formula:

[0150]

[0151] in, It is the reaction intensity, measured in kilowatts per square meter. ε is the fuel bed density, in kilograms per cubic meter; ε is the effective heating coefficient. It is the heat of combustion of fuel, measured in kilojoules per kilogram. It is a wind speed correction factor. =C'×W 2 , where C' is an empirical constant and W is the wind speed; It is the slope correction factor. =D'×tan(S), where D' is an empirical constant and S is the slope; It is a wind direction correction factor. =E×cos(θ-W d ), where E is an empirical constant, W d It's the wind direction.

[0152] For example, =500 kilowatts per square meter, =0.2W 2 , =0.1×tan(S), =0.15×cos(θ-W d ), =10 kg / m³, ε=0.5 =18 kJ / kg, W=5 m / s, S=10°, θ=45°, W d =30°, calculate:

[0153] ≈34.24 meters per minute.

[0154] It is worth noting that the fire spread rate assessment in this embodiment can be applied to predictions both before and during a fire. However, before a fire occurs, the assessment is based solely on... Determine the direction of fire spread θ; after the fire occurs, according to The fire spread direction θ is determined by the actual spread direction, where θ = γ × θ1 + β × θ2, and θ1 is... The corresponding propagation direction, θ2 is the actual propagation direction, γ+β=1.

[0155] The fire spread rate assessment scheme in this embodiment of the invention is based on the Rothermel model, takes into account wind direction changes, and combines the potential fire spread direction with the actual fire spread direction to more accurately predict the fire spread direction and speed trend, providing a scientific basis for fire prevention and emergency response.

[0156] The above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention in any way. Although the present invention has been disclosed above with reference to preferred embodiments, it is not intended to limit the present invention. Any person skilled in the art can make some modifications or alterations to the above-disclosed technical content to create equivalent embodiments without departing from the scope of the present invention. Any simple modifications, equivalent changes and alterations made to the above embodiments based on the technical essence of the present invention without departing from the scope of the present invention shall still fall within the scope of the present invention.

Claims

1. A forest fire risk assessment system based on multi-source data, comprising a data acquisition module, an information processing module and a large database, characterized in that: the large database generates a data set comprising a plurality of fire risk characteristics based on historical forest fire data and artificial input data; the data acquisition module comprises a remote sensing image unit and a plurality of environmental monitoring units; the remote sensing image unit is used to obtain remote sensing image data corresponding to the forest area and transmit it to the information processing module; a plurality of environmental monitoring units are distributed in the corresponding forest area and obtain environmental data of the corresponding local area, and each environmental data is transmitted to the information processing module; the information processing module comprises a receiving unit, a processing unit and a modeling unit; the modeling unit establishes a risk model by selecting the corresponding fire risk characteristics in the data set based on the topography, environment and vegetation species of the corresponding forest area; the receiving unit is used to receive remote sensing image data and environmental data of each environmental monitoring unit, and after spatial matching and time matching of the environmental data of each environmental monitoring unit and the remote sensing image data, it is transmitted to the processing unit; the processing unit generates corresponding level influence factors based on the spatial matching and time matching results of each environmental data and remote sensing image data, and the processing unit matches each environmental data and remote sensing image data with the risk model to obtain the number and correlation degree of the actual risk characteristics of the corresponding forest area; the number of actual risk characteristics is obtained by: the correlation degree of the corresponding actual risk characteristics includes: the processing unit performs risk assessment based on the number and correlation degree of the actual risk characteristics and in combination with the influence factors, and the formula is: the information processing module divides the forest area into a plurality of identified areas based on the geographical location of each environmental monitoring unit; the remote sensing image unit divides the remote sensing image data into a plurality of sub-image data based on each identified area; when the receiving unit performs spatial matching and time matching, each sub-image data is matched with the environmental data of the corresponding identified area; the processing unit grades the influence factors based on whether the sub-image data and the environmental data of the corresponding identified area have the same actual risk characteristics; when the actual risk characteristics of the sub-image data and the environmental data of the same identified area are the same, the influence factor is first level; when the actual risk characteristics of the sub-image data and the environmental data of the same identified area are different, the influence factor is second level; the information processing module establishes a risk assessment value distribution map matching the risk assessment value of the corresponding forest area based on the risk assessment value of each identified area; the information processing module identifies the risk gradient of each identified area relative to its adjacent identified area based on the risk assessment value, and performs risk-oriented analysis; and the information processing module determines whether the identified area with a risk assessment value higher than the risk threshold value presents a concentration trend and its spread direction through a preset risk threshold value. The steps of the information processing module for risk-oriented analysis include: S6, generating a risk-oriented map through a visualization tool and marking the risk concentration or spread direction; the information processing module normalizes the risk gradient, and the formula used is: ​ ​ ​ ​ ​ ​ ​ ​ ​ Each fire risk feature is set as a range threshold of corresponding category data When each environment data and remote sensing image data, when the real-time data of corresponding category Falls into the range threshold of corresponding category data After that, the feature point matching is successful, and is used for actual risk features; ​ Real-time data Falls within corresponding category data range threshold The closer it is to The greater the degree of association; ​ ； wherein R is the risk assessment value corresponding to the identified area; n is the total number of actual risk characteristics; is the weight of the ith actual risk characteristic; is the assigned value of the ith fire risk characteristic; is the impact factor level of the ith actual risk characteristic; is the correlation degree of the ith actual risk characteristic; is the normal data corresponding to the actual risk characteristic; is the real-time data of the ith actual risk characteristic; is the absolute distance of the actual risk characteristic deviating from the normal data corresponding to the characteristic; is the normalized range of the ith actual risk characteristic. 2.The forest fire risk assessment system based on multi-source data according to claim 1, wherein, ​ ​ ​ 3. The forest fire risk assessment system based on multi-source data according to claim 2, characterized in that, ​ ​ ​ 4. The forest fire risk assessment system based on multi-source data according to claim 3, characterized in that, ​ 5. The forest fire risk assessment system based on multi-source data according to claim 4, characterized in that, ​ ​ 6. The forest fire risk assessment system based on multi-source data according to claim 5, wherein, ​ S1, extracting the risk assessment value of each identified region from the risk assessment value distribution map ; S2, calculating a risk gradient between each identified area and its adjacent identified area ; S3, risk gradient for each identified area normalization to obtain a normalized gradient ; S4. calculating a normalized gradient and unit direction vectors , calculating a comprehensive wind direction guide for each identified area ; S5、by comparing the risk assessment value with the preset risk threshold, screening out high-risk areas; ​ 7. The forest fire risk assessment system based on multi-source data according to claim 6, characterized in that, When the information processing module performs risk-oriented analysis, the formula for calculating the risk gradient is: ; wherein to identify a risk gradient, i.e. the direction and intensity of the change of risk, from the identification area i to the identification area j; , are the risk assessment values of the identification area j and the identification area i, respectively. ​ ； wherein: is the normalized risk gradient; m is the number of all neighboring identified regions of the identified region i; is the absolute value of the risk gradient; The information processing module calculates the risk orientation of the corresponding forest region based on the normalized risk gradient, and the formula is: ； wherein is the wind direction orientation vector of the identified area i, indicating the direction in which the risk is concentrated or spreading; is the set of neighboring identified areas of the identified area i; is the unit direction vector from the identified area i to the identified area j. 8.The forest fire risk assessment system based on multi-source data according to claim 7, wherein, The information processing module is further configured to determine the fire spread direction θ and evaluate the fire spread speed R' considering the change of the wind direction, according to the following formula: determine the fire spread direction θ and evaluate the fire spread speed R' considering the change of the wind direction, according to the following formula: ， wherein, is the reaction intensity, in kilowatts per square meter; is the fuel bed density, in kilograms per cubic meter; ε is the effective heating coefficient; is the fuel ignition heat, in kilojoules per kilogram; is the wind speed correction factor, = C' x W 2 , where C' is an empirical constant and W is the wind speed; is the slope correction factor, = D' x tan(S), where D' is an empirical constant and S is the slope; is the wind direction correction factor, = E x cos(θ - W d ), where E is an empirical constant and W d is the wind direction. 9.The forest fire risk assessment system based on multi-source data according to claim 8, wherein, θ = γ x θ1+ β x θ2, θ1 is corresponding to the direction of spread, θ2 is the actual direction of spread, γ + β = 1.

Images