Expressway network fire risk measuring and calculating method, medium and terminal

Abstract

The invention is suitable for the technical field of fire prevention and control, and relates to an expressway network fire risk measuring and calculating method, medium and terminal, and the method comprises the steps: determining the composition and geographic position information of an expressway network, and obtaining a road network and a satellite remote sensing image; dividing risk units and acquiring latitude and longitude coordinates of the risk units; identifying vehicle types and numbers of different risk units based on target detection, and measuring and calculating the average length, spacing and speed of the vehicle types; calculating the annual average accident rate of each vehicle model under different road section types in combination with historical fire data; measuring and calculating the number of times of occurrence of different grades of fire in each risk unit annually; meanwhile, the influence of fire rescue accessibility on the fire loss is considered, and a loss amount adjustment value is calculated; and finally, calculating the fire risk value of each risk unit according to the annual average fire frequency and the adjusted loss amount, and carrying out visual presentation according to the risk level. The expressway fire risk level visualization is realized, the fire risk measurement and calculation accuracy is improved, and a scientific basis is provided for fire rescue station site selection and risk prevention and control.

Description

Technical Field

[0001] This invention belongs to the field of fire prevention and control technology, and in particular relates to a method, medium and terminal for calculating fire risk in highway networks. Background Technology

[0002] In existing technologies, fire risk research mostly focuses on regional fire risk, while research on highway fire risk is extremely limited. Highways are linear, isolated from the outside world, forming independent long-line spaces, and their fire risk characteristics differ from those of general regional fires. Therefore, it is necessary to calculate the fire risk of highway networks. In fire risk calculation techniques, most studies only calculate the fire risk level of the research object or risk unit, but do not visualize it. This prevents fire risk from being presented in a clear and intuitive way, hindering decision-makers and drivers from understanding the distribution of fire risk in a timely manner and taking targeted prevention and control measures.

[0003] Patent CN117114423B discloses a method and apparatus for predicting fire risks and assessing losses in industrial plants, including: analysis and assessment of potential indoor and outdoor fire risks in industrial plants; numerical simulation analysis of indoor and outdoor fires and smoke in industrial plants; and fire loss assessment based on simulated fire impact range. This patent pertains to regional fire risk assessment, which differs from highway network fire risk assessment. Furthermore, this application only calculates the fire risk level, without visualizing it, and therefore fails to address the problems existing in the prior art.

[0004] Therefore, how to provide a method for assessing fire risk in highway networks, and how to visualize the fire risk level of highways in a clear and intuitive way to improve the accuracy of fire risk assessment in highway networks, is an urgent problem to be solved by those in this technical field. Summary of the Invention

[0005] To address the shortcomings of existing technologies, the purpose of this invention is to provide a method for assessing fire risks in highway networks. This method solves the problem that existing technologies for assessing fire risks in highway networks do not achieve visualization of risk levels, resulting in low accuracy of fire risk assessments and hindering decision-makers and drivers from timely understanding the distribution of fire risks. In addition, this invention also provides a medium and terminal for assessing fire risks in highway networks.

[0006] To solve the above-mentioned technical problems, the present invention adopts the following technical solution:

[0007] In a first aspect, the present invention provides a method for assessing fire risk in highway networks, comprising the following steps:

[0008] S10. Determine the composition and geographical location information of the expressway network, obtain the road network map and satellite remote sensing images of the expressway network, determine the terrain type and road segment type of each expressway, and obtain the location information of their boundaries.

[0009] S20. Divide the expressway network into risk units and obtain their latitude and longitude coordinates;

[0010] S30. Use target detection method to identify satellite remote sensing images of different risk units, count the vehicle types and their quantities in different risk units, and calculate the average total number of different vehicle types in each risk unit under different time type scenarios, the average driving speed of various vehicle types on highways under different time type scenarios, and the average annual traffic flow of different risk units on the highway network.

[0011] S40. Calculate the annual average number of fires with different consequences and risk levels for different vehicle types on different road sections, the average mileage of different vehicle types on the expressway network, the annual average number of fires with different consequences and risk levels in different risk units of the expressway network, and the average loss per fire with different consequences and risk levels on the expressway network.

[0012] S50, Calculate the accessibility of fire and rescue services in different risk units of the expressway network;

[0013] S60. Calculate the fractional adjustment value of highway fire accident loss taking into account fire rescue accessibility, the annual average loss value of fire accidents in different risk units of the highway network, and the fire risk value of different risk units of the highway network.

[0014] S70. Determine the fire risk level of different risk units in the expressway network and their corresponding RGB values, as well as the latitude and longitude coordinate translation direction and location of different risk units in the expressway network, and visualize the risk values ​​of the risk units in the expressway network.

[0015] Furthermore, the specific steps in S10 are as follows:

[0016] S101. Determine the composition and geographical location information of the expressway network: Define the expressway network as consisting of N expressways, obtain the length L(n) of the nth expressway, and treat the up and down directions of the expressway as different roads. The latitude and longitude coordinates of the starting position of the up road are (x(n,ls),y(n,ls)), and the latitude and longitude coordinates of its ending position are (x(n,lf),y(n,lf)). The latitude and longitude coordinates of the starting position of the down road are (x(n,rs),y(n,rs)), and the latitude and longitude coordinates of its ending position are (x(n,rf),y(n,rf)).

[0017] S102. Obtain road network map and satellite remote sensing map historical images: Based on the geographical location information, obtain road network data and high-resolution satellite remote sensing images of the expressway network, and obtain high-resolution satellite remote sensing map historical images of different times in the past three years or more. Filter and integrate the historical images according to the time type that can reflect changes in traffic flow, and define the number of satellite remote sensing map historical images of the Tth time type after integration as T(z).

[0018] S103. Determine the terrain type, road segment type and their boundary information: Based on the satellite remote sensing images, road network maps and field surveys, determine the terrain type and road segment type of each expressway, as well as the boundary lines where the types change, and obtain the latitude and longitude coordinates of the center point of each boundary line; define roads with the same up and down directions, the same terrain type and the same road segment type in the same expressway as roads of the same road segment type, number the boundary lines of different road segment types of the nth expressway in a preset order, and obtain the latitude and longitude coordinates (x(b,k),y(b,k)) of the center point of the kth boundary line of the bth road segment type.

[0019] Furthermore, in S20, it is defined that there are B different road segment types for the nth expressway. The expressways of different road segment types are divided equally according to the interval l to form multiple risk units, where the remaining part with a length less than l is a separate risk unit; each service area and each toll station is a separate risk unit; all risk units on the same expressway are numbered according to the order of the uphill expressway first and the downhill expressway second, and according to the chainage number from small to large; the latitude and longitude coordinates of each risk unit are obtained, where the latitude and longitude coordinates of the jth risk unit of the nth expressway are (x(n,j),y(n,j)).

[0020] Furthermore, in step S30, the specific steps are as follows:

[0021] S301. Segment the historical satellite remote sensing imagery of highways according to risk unit boundaries, construct a dataset labeled with vehicle types and bounding boxes, and identify the vehicle types and quantities within each risk unit using a target detection algorithm; for units that cannot be identified due to occlusion or tunnels, calculate the number of vehicle types using linear interpolation, with the formula as follows:

[0022]

[0023] Where q(z,α,a) and q(z,β,a) represent the total number of type a vehicles in the two completely identifiable risk units adjacent to the j-th risk unit on the n-th highway at the z-th time of the T-th time type;

[0024] S302. Classify data by time type and calculate the average number of vehicle types within each risk unit. The formula is as follows:

[0025]

[0026] Where T(z) represents the number of historical images of the Tth time type in all satellite remote sensing map historical images of the highway network, and q(z,j,a) represents the total number of vehicle types of the ath type on the jth risk unit of the nth highway at the zth time of the Tth time type.

[0027] S303. Through orthorectification and image rotation, based on the vehicle bounding box pixel coordinates and map resolution, the average vehicle body length l(a) is calculated using the following formula:

[0028]

[0029]

[0030] Assuming the time type is T, the total number of samples of type a vehicle in the dataset obtained on the nth highway is V. The top-left pixel coordinates of the bounding box in the vth sample are (X(v,lu),Y(v,lu)), and the bottom-right pixel coordinates are (X(v,rd),Y(v,rd)). The top-left pixel coordinates of the bounding box in the (v+1)th sample are (X(v+1),Y(v+1,lu)), and the bottom-right pixel coordinates are (X(v+1),Y(v+1,rd)). The resolution of the obtained satellite remote sensing map is e meters / pixel. If the vehicle spacing between the vth sample and the (v+1)th sample is D(v), and the average spacing of type a vehicle on the nth highway in the Tth time type is D(T,a), then the calculation formula is as follows:

[0031]

[0032] ;

[0033] S304. The average driving speed of each vehicle type within a risk unit on the highway is obtained through expert experience or historical speed measurement data. For non-speed measurement units, the nearest value is used for assignment.

[0034] Suppose that in the T-th time type scenario, a total of I(a) data points on the speed of vehicle type a are obtained in the j-th risk unit of the n-th highway, where the speed of the vehicle in the i(a)-th data point is V(i(a)). The average speed of vehicle type a in the j-th risk unit of the n-th highway in the T-th time type is defined as V(T,j,a), calculated using the above method. Its calculation formula is as follows:

[0035] ;

[0036] S305. Taking into account time type, average speed, vehicle length, and vehicle spacing, calculate the daily average traffic flow Q(T,j,a) and annual average traffic flow Q(n,j,a) for each vehicle type. The formulas are as follows:

[0037]

[0038]

[0039] Where 24 represents 24h / d, V(T,j,a) is in km / h, D(T,a) and l(a) are in km, the average daily traffic flow is in veh / d, D(1), D(2) and D(3) represent the total number of working days, weekends and statutory holidays in the entire historical image statistical period of the highway satellite remote sensing map, respectively, and Y(n) represents the number of years in the entire historical image statistical period of the highway satellite remote sensing map.

[0040] Furthermore, in step S40, the specific steps are as follows:

[0041] S401. Based on historical fire data, count the number of times F(b,a,c) of type a vehicle occurring in a fire of type c consequences on type b section of type n highway within a period of q years, and calculate the average annual number of fires AF(b,a,c) as follows:

[0042]

[0043] S402. Summing all road segment types of the nth expressway, calculate the average annual number of fires YAF(n,a,c) of type a vehicle and class c fires on that expressway, using the following formula:

[0044]

[0045] Where B represents the total number of road segment types on the nth expressway;

[0046] S403. Based on the section length L(n,i) of each toll station on the nth expressway and the average daily traffic volume Q(n,i,a) of each vehicle type, calculate the average mileage L(n,a) of each vehicle type, using the following formula:

[0047]

[0048] Where m represents the number of toll stations, including the starting point and the ending point;

[0049] S404. Define the annual accident rate P(n,a,c) of type a vehicle causing a fire of class c on the nth highway as follows:

[0050]

[0051] Where P(n,a,c) is in units of times per million vehicle-kilometers;

[0052] S404. Define the number of times a type a vehicle experiences a fire of class c per year in the j-th risk unit of the n-th highway as T(j,a,c), as follows:

[0053]

[0054] Wherein, P(n,a,c) is the annual accident rate of type a fire of type c on type a vehicle on type n highway, Q(n,j,a) is the annual traffic volume of type a vehicle on type j risk unit of type n highway, and l(n,j) represents the length of type j risk unit of type n highway.

[0055] S405. Using linear interpolation, convert the losses caused by fire accidents into loss fractions. Assume the period for statistically analyzing historical fire data is q years. Number the fire accidents within this period in chronological order. If the number of seriously injured people in the h-th fire accident is Z(h), the number of deaths is S(h), and the direct property loss is C(h), then convert the casualties and direct property losses into loss fractions according to the following formula:

[0056]

[0057]

[0058]

[0059] Where DZ(h), DS(h), and DC(h) are the values ​​of the number of seriously injured, the number of dead, and the direct property loss caused by the fire accident after being converted into loss fractions, respectively. The loss fraction D(h) for the h-th fire accident is defined by the following formula:

[0060] ;

[0061] S406. Assume that during the entire statistical period, the total number of times a type a fire of class c occurs on the nth highway is G(a,c), and the loss fraction of the gth occurrence of a type a fire of class c is D(a,g,c). Define the average loss fraction per occurrence of a type a fire of class c on the nth highway as AD(a,c), as follows:

[0062] .

[0063] Furthermore, in step S50, all time scenario samples are classified according to time type, and time samples of the same type are numbered in chronological order. Assuming that the samples of time type T consist of z time scenarios, and the reachability score corresponding to the j-th risk unit in time scenario T(t) is A(T(t)), then the average reachability score corresponding to the j-th risk unit in time type T is A(T,j), as shown in the following formula:

[0064]

[0065] Assuming that the Tth time type has a total of T(d) days during the entire measurement period, the average accessibility score of the jth risk unit can be obtained as A(j), as shown in the following formula:

[0066]

[0067] Suppose that in a highway network, the total number of risk units on the nth highway is N(j), and its overall average fire and rescue accessibility is TA(N), as shown in the following formula:

[0068] .

[0069] Furthermore, in S60, the specific steps are as follows:

[0070] S601. Assume that in the j-th risk unit of the n-th highway, the adjustment coefficient for fire rescue accessibility to the amount of fire accident loss is f(j). In this risk unit, the average loss fraction adjustment value for a type a fire of class c is TD(a,c), as shown in the following formula:

[0071]

[0072] S602. Define the average annual loss value of a Class c fire occurring on the a-type vehicle in the j-th risk unit of the n-th highway as S(j,a,c), as follows:

[0073]

[0074] The average annual loss value of a fire accident in the j-th risk unit of the n-th highway is defined as S(n,,j), and is expressed as follows:

[0075]

[0076] Here, 5 represents the risk level of a highway fire accident, including minor fire, general fire, major fire, serious fire and extremely serious fire, and A represents the number of vehicle types traveling on the nth highway.

[0077] S603. Define the average annual risk value of a fire accident occurring in the j-th risk unit of the n-th highway as R(n,j), as follows:

[0078] When S(n,j)≤β(1), the fire risk level of this risk unit is Level I, and the corresponding risk value is:

[0079]

[0080] When β(1)<S(n,j)≤β(2), the fire risk level of this risk unit is Level II, and the corresponding risk value is:

[0081]

[0082] When β(2)<S(n,j)≤β(3), the fire risk level of this risk unit is Level III, and the corresponding risk value is:

[0083]

[0084] When S(n,j)>β(3), the fire risk level of this risk unit is Level IV, and the corresponding risk value is:

[0085]

[0086] The values ​​of β(1), β(2) and β(3) are determined according to the loss fractions corresponding to different fire risk levels of highways, β(1)=3, β(2)=6, β(3)=8.

[0087] Furthermore, in S70, the specific steps are as follows:

[0088] S701. Based on the preset risk value range, the highway network risk units are divided into low, general, relatively large, and major fire risk units. Based on the RGB color coordinates (blue, green, yellow, orange, and red) corresponding to the key risk values ​​0, 40, 60, 80, and 100, the RGB values ​​of any risk value are calculated using a nonlinear interpolation method to achieve dynamic mapping between risk values ​​and colors.

[0089] S702. Using the latitude and longitude coordinate translation method, assume that the j-th risk unit of the n-th highway corresponds to point E on the map. j Its latitude and longitude coordinates are (x j ,y j Point E represents the risk unit in the upbound direction of the highway, while point E represents the corresponding risk unit in the downbound direction. p Its latitude and longitude coordinates are (x p ,y p ), point E j Along vector Translation distance in the direction Point E p Along vector Translation distance in the direction , Determined based on the map scale, if the vector Let θ be the azimuth angle of the direction in which the vector is located. Then, when θ ∈ [0°, 180°), the vector... The azimuth angle of the direction is (θ+180°). When θ∈[180°,360°), then the vector... The azimuth angle of the direction is (θ-180°), and the formula for the azimuth angle θ is as follows:

[0090]

[0091] Assume point E j Along vector Translation distance in the direction Afterwards, it becomes a point. Its latitude and longitude coordinates are Point E p Along vector Translation distance in the direction Afterwards, it becomes a point. Its latitude and longitude coordinates are The formula for the translated latitude and longitude coordinates is as follows:

[0092]

[0093]

[0094]

[0095]

[0096] S703. Based on the latitude and longitude coordinates of each risk unit after coordinate translation, mark the position of each risk unit in the satellite remote sensing map, mark the corresponding position with color according to the RGB value coordinates of the risk value of the risk unit, and save it, so as to realize the visualization of the risk values ​​of different risk units of the highway.

[0097] In a second aspect, the present invention also provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the method described above.

[0098] Thirdly, the present invention also provides an electronic terminal, comprising: a processor and a memory; the memory is used to store a computer program, and the processor is used to execute the computer program stored in the memory to cause the terminal to perform the method described above.

[0099] Compared with existing technologies, the method, medium, and terminal for calculating fire risk in highway networks provided by this invention have at least the following advantages:

[0100] Most existing studies only calculate the fire risk level of the research object or risk unit, but fail to visualize it. This prevents fire risks from being presented clearly and intuitively, hindering decision-makers and drivers from understanding the fire risk distribution and taking targeted prevention and control measures. This invention integrates various dynamic and static factors to construct a multi-dimensional, multi-scenario risk assessment model, significantly improving the scientific rigor and accuracy of fire risk calculation. By combining complex calculation results with Geographic Information Systems (GIS) and satellite remote sensing maps, and utilizing latitude and longitude coordinate translation and RGB color mapping technology, this invention achieves a visualized and intuitive presentation of the fire risk levels of different risk units along highways. This greatly facilitates decision-makers, managers, and drivers in understanding the risk distribution. Furthermore, the calculation results of this invention not only provide the overall road network risk level but also detail the risk values ​​for each highway, each specific road segment (risk unit), and different vehicle types and consequences. This helps traffic management departments and fire and rescue agencies to conduct refined management and scientific decision-making, effectively improving the overall fire prevention and control capabilities and emergency response efficiency of the highway network. Attached Figure Description

[0101] To more clearly illustrate the solution of the present invention, a brief introduction will be given to the drawings used in the description of the embodiments below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0102] Figure 1 A flowchart of a method for calculating fire risk in a highway network, provided as an embodiment of the present invention;

[0103] Figure 2 This is a schematic diagram of a method for calculating vehicle body length and vehicle spacing based on satellite remote sensing images in a method for assessing fire risk in a highway network provided by an embodiment of the present invention.

[0104] Figure 3 This is a schematic diagram of the latitude and longitude coordinate translation of a highway risk unit in a highway network fire risk assessment method provided in an embodiment of the present invention. Detailed Implementation

[0105] To facilitate understanding of the present invention, a more complete description will be given below with reference to the accompanying drawings. Preferred embodiments of the invention are shown in the drawings. However, the invention can be implemented in many different forms and is not limited to the embodiments described herein. Rather, these embodiments are provided to provide a thorough and complete understanding of the disclosure of the invention.

[0106] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains. The terminology used herein in the description of the invention is for the purpose of describing particular embodiments only and is not intended to be limiting of the invention.

[0107] This invention provides a method for assessing fire risk in highway networks, applied in the research process of highway fire risk. The method for assessing fire risk in highway networks includes the following steps:

[0108] S10. Determine the composition and geographical location information of the expressway network, obtain the road network map and satellite remote sensing images of the expressway network, determine the terrain type and road segment type of each expressway, and obtain the location information of their boundaries; S20. Divide the expressway network into risk units and obtain their latitude and longitude coordinates; S30. Use target detection methods to identify satellite remote sensing images of different risk units, count the vehicle types and their quantities in different risk units, calculate the average total number of different vehicle types in each risk unit under different time type scenarios, the average driving speed of various vehicle types on the expressway under different time type scenarios, and the average annual traffic flow of different risk units of the expressway network; S40. Calculate the average number of fires with different consequences and risk levels occurring annually for different vehicle types on different road segments. The system includes: S50, calculating the accessibility of fire rescue services for different risk units in the highway network; S60, calculating the adjusted value of the highway fire accident loss score considering fire rescue accessibility; S70, calculating the annual average loss value of fire accidents in different risk units in the highway network; S80, determining the fire risk level and its corresponding RGB value for different risk units in the highway network; S90, visualizing the risk value of the highway network risk units.

[0109] This invention enables visualization of highway fire risk levels, improves the accuracy of fire risk assessment, and provides a scientific basis for fire and rescue station site selection and risk prevention.

[0110] To enable those skilled in the art to better understand the present invention, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings.

[0111] This invention provides a method for calculating the fire risk of highway networks, which is applied in the research process of highway fire risk, combined with... Figures 1 to 3 In this embodiment, the method for assessing the fire risk of the highway network includes the following steps:

[0112] S10. Determine the composition and geographical location information of the expressway network, obtain the road network map and satellite remote sensing images of the expressway network, determine the terrain type and road segment type of each expressway, and obtain the location information of their boundaries.

[0113] Specifically, in this embodiment, step S10 involves the following steps:

[0114] S101. The expressway network can be determined according to administrative divisions or it can be composed of multiple custom expressways. The names, numbers, lengths and geographical boundaries of each expressway in the expressway network are determined by using online maps and combining them with field surveys. The latitude and longitude coordinates of the center point of the geographical boundary line are obtained by using surveying instruments or online map latitude and longitude query platforms.

[0115] Specifically, the number of expressways in the expressway network is defined as N, and the length of the nth expressway is L(n). The up and down directions of the expressway are considered as different roads. The latitude and longitude coordinates of the starting position of the up road are (x(n,ls),y(n,ls)), and the latitude and longitude coordinates of its ending position are (x(n,lf),y(n,lf)). The latitude and longitude coordinates of the starting position of the down road are (x(n,rs),y(n,rs)), and the latitude and longitude coordinates of its ending position are (x(n,rf),y(n,rf)).

[0116] S102. Based on the location information of the expressway network, obtain the road network data of the expressway network in the road network map open platform, and use the satellite remote sensing map open platform to obtain high-resolution satellite remote sensing images of each expressway in the expressway network. It is necessary to be able to clearly identify the types of vehicles traveling on the expressway. Obtain historical satellite remote sensing map images of each expressway to be measured at different times in the past three years or more, and filter and integrate the historical images according to the time sequence and different time types. The division of time types should be able to reflect the obvious changes in traffic flow on the same section of the expressway.

[0117] Specifically, the number of historical images of the highway network of the Tth time type obtained after filtering and integrating the historical images of the highway satellite remote sensing map according to different time types using the above method is defined as T(z).

[0118] In this embodiment, the road network map of the expressway network can be obtained using OpenStreetMap, and the historical images of the expressway network to be measured can be obtained using open platforms such as Jilin-1, Tianditu, and Xingtu Cloud. The period of the historical images is from January 2022 to October 2025. The historical images are sorted in chronological order and classified into three time types: weekdays, weekends, and statutory holidays. There are 3 types of time types, where T∈{1,2,3}. The numbers in this set represent the time types {weekdays, weekends, statutory holidays}. If, after filtering, there are 6 groups of historical images whose time type belongs to statutory holidays, then 3(z) = 6 groups.

[0119] S103. Based on the satellite remote sensing images and road network maps of the expressway network to be evaluated, determine the terrain type and road segment type of the area where each expressway is located, and determine the boundary line where the terrain type and road segment type change. If it cannot be determined by satellite remote sensing images or road network maps, combine it with field surveys and use measuring instruments or online map latitude and longitude query platforms to obtain the latitude and longitude coordinates of the center point of the boundary line.

[0120] Specifically, roads with the same direction of travel, the same terrain type, and the same road segment type on the same expressway are defined as roads of the same road segment type. Suppose that there are B different road segment types on the nth expressway. Following the order of the road segment from the uphill road to the downhill road, the boundary lines of each road segment type are numbered in ascending order of the station number. The latitude and longitude coordinates of the center point of the kth boundary line of the bth road segment type are obtained as (x(b,k),y(b,k)).

[0121] In this embodiment, the terrain types of highways can generally be divided into plains, plateaus, hills, basins, and mountains, and can be added or removed according to actual conditions. Highway segment types can generally be divided into tunnels, bridges, ramps, service areas, toll stations, and ordinary roads, and can be added or removed according to actual conditions. Assuming that, based on satellite remote sensing images of the highway network combined with field surveys, the terrain type of a certain highway network is determined to be mainly plains and mountains, and the highway segments can be divided into tunnels, bridges, ramps, service areas, toll stations, and ordinary roads, then there are a total of 12 segment types: tunnels, bridges, ramps, service areas, toll stations, and ordinary roads in plain areas, and tunnels, bridges, ramps, service areas, toll stations, and ordinary roads in mountainous areas. Using Baidu online maps and through field surveys, the boundary lines of each segment type in the highway network are determined, and the latitude and longitude coordinates of each boundary point are obtained using the Baidu Map coordinate picking system.

[0122] S20. Divide the highway network into risk units and obtain their latitude and longitude coordinates.

[0123] Specifically, in this embodiment, the boundary lines of different road segment types of each expressway are used as boundaries. Different road segments of the same expressway are divided into equal risk units at a certain fixed interval. Risk units in each road segment that are less than the specified interval are considered as one risk unit. Each service area and toll station is considered as an independent risk unit. The latitude and longitude coordinates of each risk unit are obtained using measuring instruments or online map latitude and longitude query platforms.

[0124] In this embodiment, the nth expressway is defined as having B different road segment types. The expressways of different road segment types are divided into risk units according to the interval l. The risk units on the same expressway are numbered sequentially according to the order of the uphill expressway and the downhill expressway, based on the chainage from small to large. The latitude and longitude coordinates of the jth risk unit of the nth expressway are (x(n,j),y(n,j)).

[0125] S30. Use target detection to identify satellite remote sensing images of different risk units, count the vehicle types and their quantities in different risk units, and calculate the average total number of different vehicle types in each risk unit under different time type scenarios, the average driving speed of various vehicle types on highways under different time type scenarios, and the average annual traffic flow of different risk units on the highway network.

[0126] Specifically, in this embodiment, step S30 consists of the following steps:

[0127] S301. The filtered and integrated historical satellite remote sensing images of highways are segmented according to the boundaries of the risk units, obtaining satellite remote sensing images of different risk units. An initial dataset is established by filtering the satellite remote sensing images of different risk units. Vehicle types and their bounding boxes in the satellite remote sensing images are labeled using annotation tools to establish a highway vehicle type recognition dataset. Then, this dataset is trained using object detection algorithms (such as YOLO and SSD) to generate a highway vehicle type recognition model. This model, combined with Python programming, is used to identify vehicle types in the historical satellite remote sensing images of highway risk units, and the total number of vehicles of each type is counted on different risk unit sections of the highway network. If the number of vehicles cannot be accurately identified due to tree obstruction, shadow coverage, or tunnels in the satellite remote sensing images of highway risk units, linear interpolation is used to calculate the vehicle type and number.

[0128] Specifically, the total number of vehicle types of type a on the nth highway at the z-th time of the T-th time type is defined as q(z,j,a). If the number of vehicles cannot be accurately identified due to tree obstruction, shadow coverage, or tunnels in the satellite remote sensing image of the highway risk unit, then the total number of vehicle types of type a on it is q(z,j,a), calculated according to the following formula:

[0129]

[0130] Where q(z,α,a) and q(z,β,a) represent the total number of type a vehicles in the two completely identifiable risk units adjacent to the j-th risk unit at the z-th time of the T-th time type, respectively.

[0131] In this embodiment, the vehicle types traveling on the highway can be divided into cars, buses, light trucks, heavy trucks, trailer trucks, tankers and other vehicles. The initial dataset is labeled with satellite remote sensing images of different risk units according to different vehicle types using the labeling tool Labelimg, thus establishing a highway vehicle type recognition dataset. The total number of cars in the two adjacent risk units of tunnel A on a certain highway is 6 and 10 respectively. Therefore, the total number of cars in the risk unit of tunnel A is (6+10) / 2=8.

[0132] S302. The data on the number of different vehicle types in each risk unit of the highway network at different times are divided into T time types according to the chronological order. The average value of the data in the same time type is calculated using the arithmetic mean method. From this, the average total number of each type of vehicle in each risk unit under different time type scenarios can be obtained.

[0133] Specifically, the average total number of vehicle types of type a on the j-th risk unit of the n-th highway in the T-th time type is defined as q(T,j,a), calculated using the above method. The calculation formula is as follows:

[0134]

[0135] Where T(z) represents the number of historical images of the Tth time type in all historical images of the satellite remote sensing map of the highway network, and q(z,j,a) represents the total number of vehicle types of the ath type on the jth risk unit of the nth highway at the zth time of the Tth time type.

[0136] S303, such as Figure 2The dataset is established by selecting risk units without curves on highway sections as shown. Topographic distortion of satellite remote sensing map images is eliminated by orthorectification. The orientation of all satellite remote sensing images in the dataset is rotated to the vertical direction. The vehicle type and its bounding box of the vehicle in the dataset are obtained by using a trained highway vehicle type recognition model. The upper left and lower right pixel coordinates of vehicle type a and its bounding box are output. The satellite map resolution is obtained from the satellite remote sensing map open platform. The average length of vehicles and the average distance between different vehicles are calculated using the above data.

[0137] Specifically, suppose that on the nth highway, the total number of samples of type a vehicle in the dataset is V, where the top-left pixel coordinates of the bounding box in the vth sample are (X(v,lu),Y(v,lu)), and the bottom-right pixel coordinates of the bounding box are (X(v,rd),Y(v,rd)). The resolution of the acquired satellite remote sensing map is e meters / pixel. If the body length of type a vehicle in the vth sample is l(a,v), and the average body length of type a vehicle on the nth highway is l(a), the calculation formula is as follows:

[0138]

[0139]

[0140] Assuming the time type is T, the total number of samples of type a vehicle in the dataset obtained on the nth highway is V. The top-left pixel coordinates of the bounding box in the v-th sample are (X(v,lu),Y(v,lu)), and the bottom-right pixel coordinates are (X(v,rd),Y(v,rd)). The top-left pixel coordinates of the bounding box in the (v+1)-th sample are (X(v+1,lu),Y(v+1,lu)), and the bottom-right pixel coordinates are (X(v+1,rd),Y(v+1,rd)). The resolution of the obtained satellite remote sensing map is e meters / pixel. If the vehicle spacing between the v-th sample and the (v+1)-th sample is D(v), and the average spacing of type a vehicle on the nth highway in the T-th time type is D(T,a), then the calculation formula is as follows:

[0141]

[0142] .

[0143] S304. The average driving speed of various types of vehicles on highways under different time-type scenarios can be estimated by using expert experience, or the driving speed of different types of vehicles under different time-type scenarios can be obtained by using historical data from the highway section speed measurement system of the transportation department, and the average driving speed of various types of vehicles under different time-type scenarios can be calculated.

[0144] Specifically, if expert experience is used for estimation, the average speed of vehicle type a on the j-th risk unit of the n-th highway in time type T can be directly determined as V(T,j,a) based on the expert's estimate. If historical data from the transportation department is used to determine the average speed, it can be assumed that in time type T, a total of I(a) data points on the speed of vehicle type a on the j-th risk unit of the n-th highway are obtained, where the speed of the vehicle in the i(a)-th data point is V(i(a)). The average speed of vehicle type a on the j-th risk unit of the n-th highway in time type T, calculated using the above method, is defined as V(T,j,a). The calculation formula is as follows:

[0145]

[0146] If a risk unit on a highway does not belong to a section speed measurement segment, then the average driving speed of the nearest section speed measurement segment of the same type is selected as the average driving speed of the risk unit, according to the principle of proximity.

[0147] The daily average traffic flow on S305 expressway is related to time, average vehicle speed, average vehicle length, and average vehicle spacing. By substituting the obtained data into the traffic flow formula, the traffic flow of different vehicle types on different expressways and their respective risk units can be calculated.

[0148] Specifically, under the T-th time type, the average daily traffic flow of the a-th type of vehicle on the j-th risk unit of the n-th highway is defined as Q(T,j,a).

[0149]

[0150] Where 24 represents 24h / d, V(T,j,a) is in km / h, D(T,a) and l(a) need to be converted to km, and the unit of average daily traffic flow is veh / d.

[0151] Define the average annual traffic flow Q(n,j,a) of type a vehicle on the j-th risk unit of the n-th highway as follows:

[0152]

[0153] Where D(1), D(2) and D(3) represent the total number of working days, weekends and statutory holidays within the entire statistical period of the historical images of the highway satellite remote sensing map, respectively, and Y(n) represents the number of years in the entire statistical period of the historical images of the highway satellite remote sensing map.

[0154] Suppose that there are J risk units on the nth highway. Under the Tth time type, the average daily traffic flow of the ath type of vehicle on the nth highway is defined as Q(T,n,a), and its calculation formula is as follows:

[0155]

[0156] Let Q(n,a) be the average annual traffic volume of type a vehicle on the nth highway, and its calculation formula is as follows:

[0157] .

[0158] S40. Calculate the average number of fires with different consequences and risk levels occurring annually for different vehicle types on different road sections, the average mileage of different vehicle types on the highway network, the average number of fires with different consequences and risk levels occurring annually in different risk units of the highway network, and the average loss per fire with different consequences and risk levels on the highway network.

[0159] Specifically, in this embodiment, step S40 consists of the following steps:

[0160] S401. Obtain historical data on vehicle fires on each expressway in the study area over the past five years or more from the fire and rescue authorities and transportation departments. The data includes the time, location, vehicle type, casualties, and property damage for each fire. The type of road segment where the accident occurred is determined based on the location of the fire. Based on casualties and property damage, the consequence level of expressway fire accidents is determined according to the "Fire Statistics Management Regulations." Fires of the same type and vehicle on the same road segment are filtered and integrated according to different risk levels of fire consequences, thereby statistically obtaining the number of times different types of vehicles on different road segments have fires with different risk levels of consequences.

[0161] Specifically, assuming the period for collecting historical fire data is q years, the number of times a type a fire of level c occurs on the b-th type of road segment of the nth highway is obtained by statistics is F(b,a,c). The average annual number of fires of level c occurring on the b-th type of road segment of the nth highway, where c ∈ {1,2,3,4,5}, represents {minor fire, general fire, major fire, serious fire, and extremely serious fire}, respectively. The calculation formula is as follows:

[0162]

[0163] Let YAF(n,a,c) be the average annual number of fires of type a and class c occurring on the nth highway, and its calculation formula is as follows:

[0164]

[0165] Where B represents the total number of road segment types on the nth highway.

[0166] S402. Treat the section between two toll stations on each expressway to be measured as a whole. If the starting point and ending point of the expressway are not toll stations, then treat them as one toll station. Assume that there are N expressways and m toll stations (including the starting point and ending point) in the expressway network to be measured. Number the toll stations of each expressway in ascending order of their chainage. Then, the section between the i-th toll station and the (i+1)-th toll station on the n-th expressway is numbered i, and the length of this section is L(n,i). Use the intelligent monitoring equipment set up on this section to count the average daily number of vehicles of type a passing through this section, which is Q(n,i,a). Define the average mileage of vehicle type a on the n-th expressway as L(n,a), and its calculation formula is as follows:

[0167] .

[0168] S403. The highway fire accident rate can characterize the frequency of fire accidents on highways to assess the degree of fire risk on highways. According to the accident rate calculation formula, the annual average accident rate of different vehicle types on highways with different risk levels of fire consequences can be obtained.

[0169] Specifically, the annual accident rate of Class c fire involving vehicle type a on highway n is defined as P(n,a,c), and its calculation formula is as follows:

[0170]

[0171] Wherein, P(n,a,c) is in units of times per million vehicle-kilometers.

[0172] S404. Based on the average annual traffic volume of different risk units in the highway network, the length of the risk unit, and the average annual fire accident rate of the highway, the average number of fires of different risk levels occurring annually in different risk units in the highway network can be calculated.

[0173] Specifically, in the j-th risk unit of the n-th highway, the average number of Class c fires occurring annually for vehicle type a is defined as T(j,a,c), and its calculation formula is as follows:

[0174]

[0175] Wherein, P(n,a,c) is the annual accident rate of type a fire of type c on type a vehicle on type n highway, Q(n,j,a) is the annual traffic volume of type a vehicle on type j risk unit of type n highway, and l(n,j) represents the length of type j risk unit of type n highway.

[0176] The main types of losses from fire accidents on S405 expressways include casualties and direct property damage. To convert these different types of losses into a unified standard measurement, it is proposed to convert the different types of losses caused by fire accidents into fire accident loss scores according to their risk levels. The risk level classification of expressway fire accident consequences and corresponding loss scores is shown in Table 1 below.

[0177] Table 1

[0178] Fire rating Number of seriously injured Death toll Direct property loss (yuan) Loss fraction Minor fire 0 0 [0,1000) [0,3) General fire [1,10) [1,3) [1000,30000000) [3,6) Major fire [10,50) [3,10) [30000000,100000000) [6,8) Major fire [50,100) [10,30) [100000000,300000000) [8,10) Extremely serious fire [100,+∞) [30,+∞) [300000000,+∞) 10

[0179] The specific method for converting fire accident loss scores is as follows: Based on the fire levels and corresponding loss score standards in Table 1, linear interpolation is used to convert the losses caused by fire accidents into loss scores. Assuming the period for statistically analyzing historical fire data is q years, fire accidents within this period are numbered chronologically. If the number of seriously injured persons in the h-th fire accident is Z(h), the number of deaths is S(h), and the direct property loss is C(h), then the casualties and direct property losses are converted into loss scores using the following formula:

[0180]

[0181]

[0182]

[0183] Wherein, DZ(h), DS(h), and DC(h) are the values ​​of the number of seriously injured, the number of dead, and the direct property loss caused by the fire accident after being converted into loss fractions, respectively.

[0184] When determining the level of a fire accident, if any one of the following—the number of seriously injured, the number of fatalities, or the property damage—reaches the corresponding fire accident level, the fire accident is determined to belong to that level. Therefore, the loss fraction of the h-th fire accident is defined as D(h), and its calculation formula is as follows:

[0185]

[0186] Using the above formula, the loss fraction of all fire accidents within the statistical period of fire history data is calculated. Fire accidents are classified according to different highways, road section types, vehicle types, and fire accident levels, and each category is numbered sequentially. Assume that within the entire statistical period, the total number of times vehicle type a experiences a level c fire on the nth highway is G(a,c), and the loss fraction of the g-th occurrence of a level c fire involving vehicle type a is D(a,g,c). Define the average loss fraction per occurrence of a level c fire involving vehicle type a on the nth highway as AD(a,c), then its calculation formula is as follows:

[0187] .

[0188] S50, calculate the accessibility of fire and rescue services in different risk units of the expressway network.

[0189] Specifically, in this embodiment, when calculating the loss from a fire accident, the impact of differences in fire rescue accessibility across different risk units needs to be considered. Samples for different time scenarios are determined, and these time scenarios must be representative and reflect the dynamic changes in highway road conditions. The time scenario samples are categorized according to time type. Using the online map API interface service, request parameters are constructed using the latitude and longitude coordinates of the fire and rescue stations in the area to be measured and the latitude and longitude coordinates of each risk unit in the highway network. The online map API batch route calculation service is used to obtain the fire and rescue travel time from the nearest fire and rescue station to different risk units in the highway network, thereby calculating the fire and rescue accessibility of different risk units under each time scenario. A time-weighted method is applied to calculate the average fire and rescue accessibility of different risk units in the highway network under different time scenarios.

[0190] Specifically, all time scenario samples are classified according to different time types, and time samples of the same type are numbered in chronological order. Assuming that the samples of the Tth time type consist of z time scenarios, and the reachability score corresponding to the jth risk unit in the T(t)th time scenario is A(T(t)), then the average reachability score corresponding to the jth risk unit in the Tth time type is A(T,j), and its calculation formula is as follows:

[0191]

[0192] Assuming that the Tth time type has a total of T(d) days during the entire measurement period, the average accessibility score of the jth risk unit can be obtained as A(j), and its calculation formula is as follows:

[0193]

[0194] Suppose that in a highway network, the total number of risk units on the nth highway is N(j), and its overall average fire and rescue accessibility is TA(N), calculated as follows:

[0195] .

[0196] S60, Calculate the score adjustment value of highway fire accident loss taking into account fire rescue accessibility, the annual average loss value of fire accidents in different risk units of the highway network, and the fire risk value of different risk units of the highway network.

[0197] Specifically, in this embodiment, step S60 consists of the following steps:

[0198] S601. An adjustment coefficient f(j) is introduced to reflect the percentage adjustment of fire accident losses caused by the difference between the accessibility of fire rescue in the j-th risk unit of the highway network and the average accessibility of fire rescue. The adjustment coefficient f(j) can be determined by comparing the impact of the difference in accessibility of fire rescue on the amount of fire loss in historical fire rescue data, or by simulating the impact of fire accident losses under different accessibility scenarios. The adjustment value of the highway fire accident loss score can be calculated by the average loss per incident and the adjustment coefficient.

[0199] Specifically, suppose that in the j-th risk unit of the n-th highway, the adjustment factor for fire rescue accessibility to fire accident losses is f(j). In this risk unit, the average loss fraction adjustment value for a type a fire of class c is TD(a,c), which is calculated as follows:

[0200] .

[0201] S601. Using the annual average number of fires of different consequences and risk levels on highways and the average loss per fire of different consequences and risk levels on highways, adjust the score to obtain the annual average loss value of different vehicle types and risk levels of fires in different highway risk units to be measured.

[0202] Specifically, the average annual loss value for a Class c fire involving vehicle type a on the j-th risk unit of the n-th highway is defined as S(j,a,c), and its calculation formula is as follows:

[0203]

[0204] Let S(n,,j) be the average annual loss from a fire accident in the j-th risk unit of the n-th highway, and its calculation formula be:

[0205]

[0206] Here, 5 represents the risk level of a highway fire accident, corresponding to minor fire, general fire, major fire, serious fire and extremely serious fire, respectively, and A represents the number of vehicle types traveling on the nth highway.

[0207] S603. Based on the average annual loss value of fire consequences for each risk unit in the highway network, calculate the risk value of fire consequences for each risk unit in the highway network to be measured.

[0208] Specifically, the average annual risk value for a fire accident occurring in the j-th risk unit of the n-th highway is defined as R(n,j), and its calculation formula is as follows:

[0209] When S(n,j)≤β(1), the fire risk level of this risk unit is Level I, and the corresponding risk value is:

[0210]

[0211] When β(1)<S(n,j)≤β(2), the fire risk level of this risk unit is Level II, and the corresponding risk value is:

[0212]

[0213] When β(2)<S(n,j)≤β(3), the fire risk level of this risk unit is Level III, and the corresponding risk value is:

[0214]

[0215] When S(n,j)>β(3), the fire risk level of this risk unit is Level IV, and the corresponding risk value is:

[0216]

[0217] The values ​​of β(1), β(2) and β(3) are determined according to the loss fractions corresponding to different fire risk levels of highways in Table 1, with β(1)=3, β(2)=6, and β(3)=8.

[0218] Similarly, the fire risk value of vehicle type a in the j-th risk unit of the n-th highway can be obtained using the above formula.

[0219] S70. Determine the fire risk level of different risk units in the expressway network and their corresponding RGB values, as well as the latitude and longitude coordinate translation direction and location of different risk units in the expressway network, and visualize the risk values ​​of the risk units in the expressway network.

[0220] Specifically, in this embodiment, step S70 involves the following steps:

[0221] S701. Based on the risk values ​​of different risk units in the highway network, risk units are divided into major fire risk units, relatively large fire risk units, general fire risk units, and low fire risk units, with their corresponding risk value ranges shown in Table 2. To present the dynamic changes between the risk values ​​of different risk units, a correspondence between risk values ​​and RGB color values ​​is established, using different colors to represent the changes in risk values ​​of different risk units.

[0222] Table 2

[0223] Fire risk level Corresponding risk value Risk Unit Type Level I [0,40) Low fire risk unit Level II [40,60) General fire risk unit Level III [60,80) Larger fire risk unit Level IV [80,100] Major fire risk unit

[0224] The risk unit is defined as follows: a risk value of 0 corresponds to blue (RGB coordinates: (0, 0, 255)); a risk value of 40 corresponds to green (RGB coordinates: (0, 255, 0)); a risk value of 60 corresponds to yellow (RGB coordinates: (255, 255, 0)); a risk value of 80 corresponds to orange (RGB coordinates: (255, 128, 0)); and a risk value of 100 corresponds to red (RGB coordinates: (255, 0, 0)). These three RGB coordinates correspond to the brightness values ​​of the red, green, and blue channels, respectively. The RGB coordinates for other risk units can be calculated using nonlinear interpolation based on the known correspondence between risk values ​​and RGB coordinates. The colors corresponding to different risk values ​​are then determined based on the calculated RGB coordinates.

[0225] S702, such as Figure 3 As shown, by using the latitude and longitude coordinate translation method, the position of the risk units in the up and down directions of the highway in the satellite remote sensing map is changed, so that when the highway fire risk is visualized, the difference in risk level of the risk units in the up and down directions of the highway can be presented.

[0226] Specifically, the proposed latitude and longitude coordinate translation method assumes that the j-th risk unit of the n-th highway corresponds to point E on the map. j Its latitude and longitude coordinates are (x j ,y j Point E represents the risk unit in the upbound direction of the highway, while point E represents the corresponding risk unit in the downbound direction. p Its latitude and longitude coordinates are (x p ,y p In order to clearly show the difference in risk levels between uplink and downlink risk units in satellite remote sensing maps, it is proposed to divide point E... j Along vector Translation distance in the direction Point E p Along vector Translation distance in the direction , The map scale is determined to ensure that the changing trends of risk levels in both upward and downward directions are clearly visible during risk visualization. If the vector... Let θ be the azimuth angle of the direction in which the vector is located. Then, when θ ∈ [0°, 180°), the vector... The azimuth angle of the direction is (θ+180°). When θ∈[180°,360°), then the vector... The azimuth angle of the direction is (θ-180°). The formula for calculating the azimuth angle θ is as follows:

[0227]

[0228] Assume point E j Along vector Translation distance in the direction After that, it became a dot. Its latitude and longitude coordinates are Point E p Along vector Translation distance in the direction After that, it became a dot. Its latitude and longitude coordinates are The formula for calculating the translated latitude and longitude coordinates is as follows:

[0229]

[0230]

[0231]

[0232] .

[0233] S703. Based on the latitude and longitude coordinates of each risk unit after coordinate translation, mark the position of each risk unit in the satellite remote sensing map, mark the corresponding position with color according to the RGB value coordinates of the risk value of the risk unit, and save it, so as to realize the visualization of the risk values ​​of different risk units of the highway.

[0234] In this embodiment, a dataset containing the latitude and longitude coordinates and corresponding RGB value coordinates of different risk units is prepared. A DataFrame can be constructed using the pandas library. Then, a basic map object is created using the `folium.Map()` function. Each row in the data is iterated through, and markers are added to the map using the `folium.CircleMarker()` or `folium.Marker()` functions. The RGB values ​​in the data are converted to a color format recognizable by folium and applied to the `color`, `fillColor`, and other attributes of the markers. Finally, the map is saved as an HTML file using the `save()` method.

[0235] This invention also provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements any of the methods in this embodiment.

[0236] This invention also provides an electronic terminal, including: a processor and a memory; the memory is used to store a computer program, and the processor is used to execute the computer program stored in the memory, so that the terminal performs any of the methods in this embodiment.

[0237] As will be understood by those skilled in the art, the computer-readable storage medium described in this embodiment allows for the implementation of all or part of the steps in the above method embodiments by computer program-related hardware. The aforementioned computer program can be stored in a computer-readable storage medium. When executed, the program performs the steps of the above method embodiments; and the aforementioned storage medium includes various media capable of storing program code, such as ROM, RAM, magnetic disks, or optical disks.

[0238] The electronic terminal provided in this embodiment includes a processor, a memory, a transceiver, and a communication interface. The memory and the communication interface are connected to the processor and the transceiver and complete communication between them. The memory is used to store computer programs, the communication interface is used to perform communication, and the processor and the transceiver are used to run the computer programs, so that the electronic terminal performs the steps of the above method.

[0239] Compared with existing technologies, most studies in the highway network fire risk assessment method, medium, and terminal described in the above embodiments only calculate the fire risk level of the research object or risk unit, but do not visualize it. This makes it difficult for decision-makers and drivers to understand the fire risk distribution in a clear and intuitive way, hindering them from taking targeted prevention and control measures. This invention integrates various dynamic and static factors to construct a multi-dimensional, multi-scenario risk assessment model, significantly improving the scientific rigor and accuracy of fire risk assessment. By combining complex calculation results with Geographic Information Systems (GIS) and satellite remote sensing maps, and utilizing latitude and longitude coordinate translation and RGB color mapping technology, this invention achieves a visualized and intuitive presentation of the fire risk level of different risk units along the highway, greatly facilitating decision-makers, managers, and drivers in understanding the risk distribution. The calculation results of this invention not only provide the overall risk level of the road network but also refine the risk values ​​for each highway, each specific road segment (risk unit), and different vehicle types and consequences, helping traffic management departments and fire and rescue agencies to conduct refined management and scientific decision-making, effectively improving the overall fire prevention and control capabilities and emergency response efficiency of the highway network.

[0240] Obviously, the embodiments described above are merely preferred embodiments of the present invention, and not all embodiments. The accompanying drawings illustrate preferred embodiments of the present invention, but do not limit the scope of the patent. The present invention can be implemented in many different forms; rather, these embodiments are provided to provide a more thorough and complete understanding of the disclosure of the present invention. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art can still modify the technical solutions described in the foregoing specific embodiments, or make equivalent substitutions for some of the technical features. Any equivalent structures made using the content of this specification and drawings, directly or indirectly applied to other related technical fields, are similarly within the scope of patent protection of this invention.

Claims

1. A method for measuring and calculating the risk of highway network fire, characterized in that, The method comprises the following steps: S10, determining the composition and geographical position information of the expressway network, obtaining the road network map and satellite remote sensing image of the expressway network, determining the terrain type and road section type of each expressway, and obtaining the position information of the boundaries thereof; S20, dividing the risk units of the expressway network and obtaining the latitude and longitude coordinates thereof; S30, identifying the satellite remote sensing images of different risk units by using a target detection method, counting the types and quantities of vehicles in different risk units, and measuring the average total quantity of different vehicle types in different risk units under different time type scenarios, the average driving speed of different vehicle types on the expressway under different time type scenarios, and the annual average vehicle flow of different risk units of the expressway network; S40, measuring the number of times of different consequence risk grade fires occurring in different road sections and different vehicle types per year, the average driving mileage of different vehicle types on the expressway network, the number of times of different consequence risk grade fires occurring in different risk units of the expressway network per year, and the average loss amount score of different consequence risk grade fires on the expressway network; S50, measuring the fire rescue accessibility of different risk units of the expressway network; S60, measuring the loss amount score adjustment value of the expressway fire accident considering the fire rescue accessibility, the loss value of the expressway fire accident occurring in different risk units of the expressway network per year, and the fire risk value of different risk units of the expressway network; S70, determining the fire risk grade of different risk units of the expressway network and the corresponding RGB value, the translation direction and position of the latitude and longitude coordinates of different risk units of the expressway network, and visualizing the risk value of the risk units of the expressway network.

2. The method according to claim 1, wherein, In S10, the specific steps are as follows: S101, determining the composition and geographical position information of the expressway network: defining that the expressway network is composed of N expressways, obtaining the length L(n) of the nth expressway, regarding the uplink and downlink of the expressway as different roads, the latitude and longitude coordinates of the uplink starting point position are (x(n, ls), y(n, ls)), the latitude and longitude coordinates of the uplink terminal position are (x(n, lf), y(n, lf)), the latitude and longitude coordinates of the downlink starting point position are (x(n, rs), y(n, rs)), and the latitude and longitude coordinates of the downlink terminal position are (x(n, rf), y(n, rf)); S102, obtaining the road network map and satellite remote sensing map historical image: based on the geographical position information, obtaining the road network data of the expressway network, and obtaining high-resolution satellite remote sensing map historical images at different times in the past three years or more, screening and integrating the historical images according to the time types that can reflect the vehicle flow changes, and defining the number of satellite remote sensing map historical images of the Tth time type after integration as T(z); S103, determining the terrain type, road section type and its boundary information: based on the satellite remote sensing image, road network map and field investigation, the terrain type and road section type of each expressway are determined, and the boundary line where the type changes is determined, and the longitude and latitude coordinates of the center point of each boundary line are obtained; the same road section type road with the same uplink and downlink direction, terrain type and road section type in the same expressway is defined, the boundary lines of different road section types of the nth expressway are numbered in a predetermined order, and the longitude and latitude coordinates (x(b,k),y(b,k)) of the center point of the kth boundary line of the bth road section type are obtained.

3. The method according to claim 2, wherein, In the S20, the nth expressway has B different road section types, the different road section types of the expressway are equally divided according to the interval l, and a plurality of risk units are divided, wherein the remaining part with a length less than l is separately used as a risk unit; each service area and each toll station are respectively used as an independent risk unit; all risk units on the same expressway are numbered in the order of uplink expressway first and downlink expressway second, and according to the stake number from small to large; the longitude and latitude coordinates of each risk unit are obtained, wherein the longitude and latitude coordinates of the jth risk unit of the nth expressway are (x(n,j),y(n,j)).

4. The method according to claim 3, wherein, In the S30, the specific steps are as follows: S301, the satellite remote sensing map historical image of the expressway is segmented according to the risk unit boundary, a data set of labeled vehicle type and boundary box is constructed, and the vehicle type and quantity in each risk unit are identified by a target detection algorithm; For the unit that cannot be identified due to shielding or tunnel, the linear interpolation method is used to calculate the vehicle type quantity, and the formula is: ； Wherein, q(z,α,a) and q(z,β,a) respectively represent the total number of the ath vehicle type on the front and rear two completely identifiable vehicle risk units adjacent to the jth risk unit on the nth expressway at the zth moment of the Tth time type; S302, the data is classified according to the time type, and the average number of vehicle types in each risk unit is calculated, and the formula is: ； Wherein, T(z) represents the number of historical images of the Tth time type in all expressway network satellite remote sensing map historical images, and q(z,j,a) represents the total number of the ath vehicle type in the jth risk unit of the nth expressway at the zth moment of the Tth time type; S303, through orthographic correction and image rotation, based on the vehicle boundary box pixel coordinates and map resolution, the average vehicle body length l(a) is calculated, and the formula is as follows: ； ； Assuming the Tth time type, the total number of samples of the ath vehicle type in the data set is V on the nth highway, wherein the upper left corner pixel coordinates of the bounding box in the vth sample are (X(v,lu), Y(v,lu)), and the right lower corner pixel coordinates are (X(v,rd), Y(v,rd)); the upper left corner pixel coordinates of the bounding box in the (v+1)th sample are (X(v+1,lu), Y(v+1,lu)), and the right lower corner pixel coordinates are (X(v+1,rd), Y(v+1,rd)); the resolution of the obtained satellite remote sensing map is e meters / pixel; if the vehicle distance between the vth sample and the (v+1)th sample is D(v), the average distance of the ath vehicle type on the nth highway in the Tth time type is D(T,a), and the calculation formula is as follows: ； ； S304, obtain the average driving speed of each vehicle type on the different risk units of the highway through expert experience or interval speed measurement historical data, and assign values to the non-speed measurement units according to the nearest principle: Assuming that in the Tth time type, the total number of data of the ath vehicle speed on the jth risk unit of the nth highway is I(a), wherein the driving speed of the vehicle in the ith(a) data is V(i(a)), and the average driving speed of the ath vehicle type on the jth risk unit of the nth highway in the Tth time type is V(T,j,a) which is calculated by the above method, and the calculation formula is as follows: ； S305, combine the time type, average speed, vehicle length and vehicle distance to calculate the daily vehicle flow Q(T,j,a) and annual vehicle flow Q(n,j,a), and the formulas are as follows: ； ； Wherein, 24 represents 24h / d, V(T,j,a) is in unit of km / h, D(T,a) and l(a) are in unit of km, the unit of daily vehicle flow is veh / d, D(1), D(2) and D(3) respectively represent the total number of working days, weekends and statutory holidays in the historical satellite remote sensing map image statistical period of the entire highway, and Y(n) represents the number of years in the historical satellite remote sensing map image statistical period of the entire highway.

5. The method according to claim 4, wherein, In the S40, the specific steps are as follows: S401, based on historical fire data, count the number of fires F(b,a,c) of the ath vehicle type with the cth consequence level on the nth highway in the bth road section type in the q-year period, and calculate the annual average number of fires AF(b,a,c), and the formula is as follows: ； S402, sum all road section types of the nth highway to calculate the annual average number of fires YAF(n,a,c) of the ath vehicle type with the cth level on the highway, and the formula is as follows: ； Wherein, B represents the total number of road section types on the nth highway; S403, according to the interval length L(n,i) of each toll station of the nth highway and the daily traffic volume Q(n,i,a) of the vehicle type, the average driving distance L(n,a) of the vehicle type is calculated, and the formula is as follows: ； Wherein, m is the number of toll stations, including the start and end points; S404. The annual accident rate P(n,a,c) for a type a vehicle causing a fire of class c on the nth highway is defined by the following formula: ； Where P(n,a,c) is in units of times per million vehicle-kilometers; S404. Define the number of times a type a vehicle experiences a fire of class c per year in the j-th risk unit of the n-th highway as T(j,a,c), as follows: ； Wherein, P(n,a,c) is the annual accident rate of type a fire of type c on type a vehicle on type n highway, Q(n,j,a) is the annual traffic volume of type a vehicle on type j risk unit of type n highway, and l(n,j) represents the length of type j risk unit of type n highway. S405. Using linear interpolation, convert the losses caused by fire accidents into loss fractions. Assume the period for statistically analyzing historical fire data is q years. Number the fire accidents within this period in chronological order. If the number of seriously injured people in the h-th fire accident is Z(h), the number of deaths is S(h), and the direct property loss is C(h), then convert the casualties and direct property losses into loss fractions according to the following formula: ； ； ； Where DZ(h), DS(h), and DC(h) are the values ​​of the number of seriously injured, the number of dead, and the direct property loss caused by the fire accident after being converted into loss fractions, respectively. The loss fraction D(h) for the h-th fire accident is defined by the following formula: ； S406. Assume that during the entire statistical period, the total number of times a type a fire of class c occurs on the nth highway is G(a,c), and the loss fraction of the gth occurrence of a type a fire of class c is D(a,g,c). Define the average loss fraction per occurrence of a type a fire of class c on the nth highway as AD(a,c), as follows: 。 6. The method according to claim 5, wherein, In step S50, all time scenario samples are classified according to time type, and time samples of the same type are numbered according to their chronological order. Assuming that the samples of the Tth time type consist of z time scenarios, and the reachability score corresponding to the jth risk unit in the T(t)th time scenario is A(T(t)), then the average reachability score corresponding to the jth risk unit in the Tth time type is A(T,j), as shown in the following formula: ； Assuming that the Tth time type has a total of T(d) days during the entire measurement period, the average accessibility score of the jth risk unit can be obtained as A(j), as shown in the following formula: ； Suppose that in a highway network, the total number of risk units on the nth highway is N(j), and its overall average fire and rescue accessibility is TA(N), as shown in the following formula: 。 7. The method according to claim 6, wherein, In S60, the specific steps are as follows: S601. Assume that in the j-th risk unit of the n-th highway, the adjustment coefficient for fire rescue accessibility to the amount of fire accident loss is f(j). In this risk unit, the average loss fraction adjustment value for a type a fire of class c is TD(a,c), as shown in the following formula: ； S602. Define the average annual loss value of a Class c fire occurring on the a-type vehicle in the j-th risk unit of the n-th highway as S(j,a,c), as follows: ； The average annual loss value of a fire accident in the j-th risk unit of the n-th highway is defined as S(n,j), and is expressed as follows: ； Here, 5 represents the risk level of a highway fire accident, including minor fire, general fire, major fire, serious fire and extremely serious fire, and A represents the number of vehicle types traveling on the nth highway. S603. Define the average annual risk value of a fire accident occurring in the j-th risk unit of the n-th highway as R(n,j), as follows: When S(n,j)≤β(1), the fire risk level of this risk unit is Level I, and the corresponding risk value is: ； When β(1)<S(n,j)≤β(2), the fire risk level of this risk unit is Level II, and the corresponding risk value is: ； When β(2)<S(n,j)≤β(3), the fire risk level of this risk unit is Level III, and the corresponding risk value is: ； When S(n,j)>β(3), the fire risk level of this risk unit is Level IV, and the corresponding risk value is: ； The values ​​of β(1), β(2) and β(3) are determined according to the loss fractions corresponding to different fire risk levels of highways, β(1)=3, β(2)=6, β(3)=8.

8. The method according to claim 7, wherein, In S70, the specific steps are as follows: S701. Based on the preset risk value range, the highway network risk units are divided into low, general, relatively large, and major fire risk units. Based on the RGB color coordinates (blue, green, yellow, orange, and red) corresponding to the key risk values ​​0, 40, 60, 80, and 100, the RGB values ​​of any risk value are calculated using a nonlinear interpolation method to achieve dynamic mapping between risk values ​​and colors. S702, using the latitude and longitude coordinate translation method, assuming that the jth risk unit of the nth highway corresponds to point E on the map j , whose latitude and longitude coordinates are (x j , y j ), which is a risk unit in the uplink direction of the highway, and the corresponding risk unit in the downlink direction is point E p , whose latitude and longitude coordinates are (x p , y p ), point E j is translated by a distance of in the direction of vector , and point E p is translated by a distance of in the direction of vector , According to the scale of the map, if the azimuth of vector is θ, then when θ ∈ [0°, 180°), the azimuth of vector is (θ + 180°), and when θ ∈ [180°, 360°), the azimuth of vector is (θ - 180°), and the azimuth θ is calculated as follows: ； Assume point E j along the vector in the direction of the vector and a distance of later, becomes point with latitude and longitude coordinates , point E p along the vector in the direction of the vector and a distance of later, becomes point with latitude and longitude coordinates The formula for the latitude and longitude coordinates after translation is as follows: ； ； ； ； S703. Based on the latitude and longitude coordinates of each risk unit after coordinate translation, mark the position of each risk unit in the satellite remote sensing map, mark the corresponding position with color according to the RGB value coordinates of the risk value of the risk unit, and save it, so as to realize the visualization of the risk values ​​of different risk units of the highway.

9. A computer-readable storage medium, characterized in that, The storage medium stores a computer program that, when executed by a processor, implements the method as described in any one of claims 1 to 8.

10. An electronic terminal, characterized in that include: Processor and memory; The memory is used to store a computer program, and the processor is used to execute the computer program stored in the memory to cause the terminal to perform the method as described in any one of claims 1 to 8.

Images