Lightweight traditional village fire spread prediction method, system, device and medium

Abstract

The application discloses a kind of lightweight traditional village fire spread prediction method, system, equipment and medium, steps are as follows: making traditional village fire spread dataset and verification set, building combustible feature attribute and burning environment feature attribute are given;Using the above dataset, based on two classification logistic regression analysis model, the main influencing factor of fire spread with significant correlation is filtered out, and the formula for calculating the probability of fire spread is determined;Fire spread simulation mode based on neighborhood is constructed, fire spread judgment standard is formulated, and then lightweight fire spread prediction model is formed;The probability of building fire spread in the neighborhood of the building on fire is calculated step by step, the path direction of fire spread is judged, and the calculation and visualization of fire spread sequence and key building are carried out.This method can be used for fire spread simulation between traditional village building groups, predict fire spread sequence, detect key buildings affecting fire spread, and provide scientific basis for deployment of traditional village fire fighting facilities.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of traditional village protection based on geographic information data, and particularly relates to a lightweight traditional village fire spread prediction method and system, an electronic device and a storage medium. BACKGROUND

[0002] Due to the dense layout of traditional village buildings, there is a large amount of wood in the building structure, and flammable objects are stored indoors, resulting in a high fire risk. Once a fire breaks out, the fire will spread rapidly, especially when there is wind outside, which can easily cause a "campfire" disaster, causing irreparable damage to the historical and cultural value of the village, the safety of residents' lives and economic development. The fire between traditional village buildings occurs in a complex burning environment, and is affected by external factors such as human beings, village characteristics, geographical environment and other intervention factors, resulting in the coexistence of certainty and randomness in the burning spread process. Predicting the spread of fire between buildings is of great significance to the prevention and control of traditional village fires.

[0003] At present, domestic and foreign scholars have carried out a large number of scientific researches on the prediction of building fire spread, mainly in two directions. One direction is to measure the changes of various related numerical values during fire spread according to actual experiments, and to restore the physical characteristics of the research object to the greatest extent; the other direction is to calculate and simulate according to the principle of fire dynamics. The current traditional village fire spread prediction still has the following problems:

[0004] (1) No overall fire spread prediction method for traditional village building groups has been proposed, and the existing technology can only predict whether a fire spread path will occur between two buildings;

[0005] (2) No fire spread simulation method considering the spread conditions of traditional villages has been proposed, and the existing technology mainly simulates the change of a single influencing factor, simplifies the pattern characteristics and terrain environmental factors of traditional villages, and cannot comprehensively consider the complex burning properties of traditional villages;

[0006] (3) No lightweight fire spread probability calculation and spread simulation method has been proposed, and the existing technology considers redundant factors, and the calculation method is complex, requiring a long time and a large amount of manpower;

[0007] Therefore, there is an urgent need to propose a lightweight fire spread prediction method that considers the complex burning properties of traditional villages and captures key influencing factors to build a probability calculation formula, to efficiently evaluate the fire spread path of traditional villages, determine key building nodes, and guide the deployment of later fire prevention strategies. SUMMARY

[0008] The purpose of the present application is to solve the above-mentioned defects in the prior art, and provide a lightweight traditional village fire spread prediction method, system, electronic device and storage medium. The prediction method collects traditional village fire spread related data, uses a binary classification logistic regression analysis model to screen the main influencing factors of fire spread with significant correlation, determines the fire spread probability calculation formula, constructs a neighborhood-based fire spread simulation mode, forms a lightweight fire spread prediction model, judges the path direction of fire spread after simulating and predicting the target traditional village, and calculates and visualizes the fire spread sequence and key buildings. The method can be used for fire spread simulation of traditional villages in various regions, simplifies the calculation steps, and improves the accuracy and automation degree of the overall fire spread situation evaluation of traditional villages.

[0009] In order to achieve the above-mentioned purpose, the present application adopts the following technical scheme:

[0010] In a first aspect, the present application provides a lightweight traditional village fire spread prediction method, which comprises the following steps:

[0011] S1, making a traditional village fire spread dataset and a verification set: importing the surveyed traditional village CAD building block plan into a geographic information processing software GIS platform to form a dataset T and a verification set Y, wherein the buildings in the dataset T are assigned actual disaster conditions, building combustible feature attributes and burning environment feature attributes, and the buildings in the verification set Y are assigned building combustible feature attributes and burning environment feature attributes;

[0012] S2, screening the main influencing factors of fire spread: putting the dataset T into a binary classification logistic regression analysis model to screen the main influencing factors of fire spread with significant correlation;

[0013] S3, determining the fire spread probability calculation formula: putting the main influencing factors of fire spread with significant correlation into the binary classification logistic regression analysis model, calculating the correlation coefficients corresponding to the influencing factors, and constructing the fire spread probability calculation formula;

[0014] S4, constructing a neighborhood-based fire spread simulation mode: determining the building neighborhood range, gradually calculating the fire spread probability in stages, defining the buildings with a probability P greater than or equal to 0.5 as the next calculation stage of the burning buildings, forming the building neighborhood of the new calculation stage, calculating the probability of the buildings that have not caught fire in the building neighborhood of the new calculation stage, and stopping the simulation until there is no burning building in the building neighborhood of the new calculation stage or the fire spread probability P of the building is less than 0.5;

[0015] S5, formulating the fire spread judgment standard: in the same calculation stage, the buildings that have not caught fire and have a probability P greater than or equal to 0.5 in the same burning building neighborhood determine the burning order according to the probability size, and the buildings with a larger probability have a higher sequence in the fire spread sequence, and the buildings with a smaller probability have a lower sequence in the fire spread sequence.

[0016] S6. Data Visualization: Visualize the fire spread sequence, mark the path direction of fire spread and the sequence of building fires, count the number of fire spread paths in traditional village buildings, and obtain the fire spread sequence map and the distribution map of key buildings with multiple spread paths in the target traditional village.

[0017] Furthermore, the index layer factors included in the traditional village fire spread dataset T and validation set Y in step S1 are determined based on the principles of fire dynamics, referencing existing research on traditional village fire spread, and considering the influencing factors of numerous traditional villages that have experienced fire spread disasters. Fire dynamics is the science that studies the basic principles and laws of fire occurrence, development, and control. Fire dynamics indicates that three elements are required for a fire to occur: combustible material, oxidizer, and ignition source, all of which must be present simultaneously for combustion to occur. Since fire spread between buildings in traditional villages occurs in open spaces, where oxygen, as an oxidizer, is always abundant, the characteristics of the combustible material in the buildings and the characteristics of the combustion environment are the main factors affecting fire spread in traditional villages. Fire dynamics describes fire spread as divided into two processes: fire spread within buildings and fire spread between buildings. Fire spread between buildings is a complex process dominated by thermal radiation, with multiple pathways such as thermal conduction, thermal convection, and flying embers coexisting.

[0018] This invention aims to predict the fire spread sequence in traditional villages and explore the probability of fire spreading between buildings. Therefore, the constructed traditional village fire spread dataset T and validation set Y contain two types of numerical values: building combustible material characteristics and combustion environment characteristics. Dataset T mainly contains one type of numerical value: actual disaster situation. The actual disaster situation is a binary variable: a target building ignited by a burning building is marked as 1, and a target building not ignited by a burning building is marked as 0. The building combustible material characteristics consist of four index layer factors: a fixed fire load density... Qi Mobile fire load density Qm Building surface area S f Building opening area S 0 The combustion environment characteristics are composed of two index layer factors: building proximity (D) and building relative wind direction position (R). The building relative wind direction position (R) is a categorical variable, while the other five index layer factors are continuous variables, with values ​​obtained from actual observation and measurement. The specific statistical standards are as follows:

[0019] ① Fixed fire load density QiDefinition: The fire load density of building structure and component, determined by the material and structure of the building, with reference to the type value: reinforced concrete structure: 0 MJ / ㎡, masonry structure: 200 MJ / ㎡, rammed earth structure: 300 MJ / ㎡, brick-wood structure / stone-wood structure: 1000 MJ / ㎡, all-wood structure: 1500 MJ / ㎡, grass-wood structure: 2000 MJ / ㎡;

[0020] ②Mobile fire load density Qm Definition: The total fire load density of building interior items, determined by the function of the building, with reference to the type value: office building: 400 MJ / ㎡, public exhibition building: 500 MJ / ㎡, residential building: 780 MJ / ㎡, leisure and entertainment building: 1000 MJ / ㎡, commercial building: 1500 MJ / ㎡, educational building: 2500 MJ / ㎡;

[0021] ③Building surface area S f Definition: The sum of the facade and roof area of the building, measured actually;

[0022] ④Building opening area S 0 Definition: The sum of the area of doors, windows and openings on the facade of the building, measured actually;

[0023] ⑤Building proximity D Definition: The nearest spatial distance between the facade of the target building and the facade of the burning building, measured actually, which is the square root of the sum of the square of the horizontal distance value and the square of the vertical distance value;

[0024] ⑥Building relative wind direction position R Definition: The position relationship between the target building and the burning building under certain wind direction conditions, with the upwind position recorded as 1 and the downwind position recorded as 0; In the data set T and the verification set Y, the wind direction on the day of the fire event is taken as the basis for statistics, and when simulating the prediction of traditional villages not affected by fire, the annual average wind direction of the target village is taken as the basis for statistical calculation.

[0025] Further, in step S1, to avoid accidental results when using binary logistic regression analysis model to screen the main influencing factors of fire spread and determine the fire spread probability calculation formula, samples with 20 times the number of independent variables should be selected for regression analysis. The present application has a total of 6 independent variables, so it is required to collect data set T containing 2 or more traditional villages, and the total number of traditional village buildings is 120 or more.

[0026] Further, in the step S2, the binary classification logistic regression analysis model used is a multivariate statistical analysis model, which is suitable for binary response variables, i.e. the dependent variable can only be represented by the numbers 0 and 1. The binary classification logistic regression analysis model can construct a scientific mathematical analysis approach between the independent variables and the dependent variables, judge the correlation between the independent variables and the dependent variables, and the calculation formula is as follows:

[0027] Formula (1)

[0028] In the formula, y =1 represents that an event occurs, P y =1) represents the probability value of the occurrence of an event; x i is the ith independent variable, i=1, 2, 3, …n; n is the total number of independent variables, is the regression calculation-based x i corresponding logistic regression coefficient;

[0029] Therefore, the main influencing factors of fire spread are screened in the process, which includes:

[0030] In the data set T formed in the step S1, the actual disaster situation of the traditional village building is taken as the dependent variable, the fixed fire load density Qi , the moving fire load density Qm , the building surface area S f , the building opening area S 0 and the building proximity D in the 6 index layers are taken as continuous variables, the building relative wind direction position R with a classification value is taken as a virtual variable, and they are jointly included in the calculation formula for analysis to calculate the fire spread probability of the target building when it is ignited by the building on fire.

[0031] It is specified that in the coefficient comprehensive test of the binary classification logistic regression model, when the significance value of the model is <0.005, the binary classification logistic regression model has statistical significance, and when the significance value of the independent variable is ≤0.05, it has significant correlation, at this time, it indicates that this independent variable can be used as the main influencing factor of fire spread, and is included in the construction of the subsequent fire spread probability calculation formula, otherwise, the independent variable needs to be removed.

[0032] ​Further, in the step S3, the binary logistic regression analysis model can calculate the contribution degree of each independent variable to the change of the dependent variable and the significant correlation, and integrate the relationship between the prediction variable and the prediction result into the interval of 0 to 1 by the maximum likelihood method, and then construct the corresponding probability formula. The probability formula threshold is usually set to 0.5, that is, the samples with P≥0.5 are classified as "1" corresponding to the event occurrence, and the samples with P<0.5 are classified as "0" corresponding to the event non-occurrence. The present application calculates the correlation coefficient corresponding to the influence factor, and constructs the fire spread probability calculation formula, the process is as follows:

[0033] The main fire spread influencing factors screened out in S2 are put into the binary classification logistic regression model as independent variables, and the actual disaster situation is taken as the dependent variable to calculate the fire spread probability of the target building when it is ignited by the building on fire, wherein the calculation formula is the same as the above formula (1), wherein the independent variable x i The main fire spread influencing factors are screened out, the fire spread probability calculation formula is constructed according to formula (1), and the fire spread probability formula threshold is defined as 0.5, and the formula is constructed as follows:

[0034] Formula (2)

[0035] In the formula, y =1 indicates that a building is ignited by a building on fire, P y =1) indicates the probability value that a building is ignited by a building on fire. is the logistic regression coefficient corresponding to the main fire spread influencing factor based on regression calculation, i=1, 2, 3, …n, n is the total number of independent variables, n=6.

[0036] Further, in the step S4, the fire spread between buildings is a process of heat transfer in multiple ways, in which thermal radiation is the main heat transfer way. The characteristic of thermal radiation heat transfer is that the heat energy generated by fire is transmitted in the form of electromagnetic wave from the building on fire to the outside, and it is weakened with the increase of distance, so the thermal radiation heat transfer has a certain influence range. According to a large number of research results and actual disaster analysis, when the building spacing is more than 10m, the fire is basically difficult to spread between buildings. In order to ensure the rigor of the research, the concept of neighborhood is introduced in the present application, and the neighborhood is defined as the buffer zone formed by expanding the outer contour of the building on fire by 10m. The process of spread is divided into multiple stages of development by using the concept of neighborhood, which can ensure the extraction of time sequence of spread, and avoid the increase of calculation amount due to the difference of complete combustion time of buildings. The specific process is as follows:

[0037] ​The fire spreading between buildings is simplified as gradually expanding the fire disaster area in a phased manner. The building neighborhood range is defined as the farthest range when the fire directly spreads outwards from each building in each phase. A buffer zone formed by expanding the outer contour of the building by 10m is the building neighborhood. When a building is on fire, the fire spreading neighborhood caused by the on-fire building is formed. It is further determined whether the buildings located in the on-fire building neighborhood meet the un-fired condition. If so, the fire spreading probability from the on-fire building to the un-fired building is calculated according to formula (2). Figure 2

[0038] If a building is only partially located in the on-fire building neighborhood, the building is also regarded as a potential object of fire spreading from the on-fire building. It is determined whether the building is on fire. If so, the fire spreading probability of the building in the neighborhood is calculated.

[0039] If there are multiple un-fired buildings with a fire spreading probability P≥0.5 in a neighborhood, they are defined as being ignited in the same calculation phase. When the buildings in the on-fire building neighborhood are ignited, the fire spreading neighborhood calculation of the next phase is already generated at the same time, forming the on-fire building neighborhood of the new calculation phase.

[0040] If an un-fired building is located in the neighborhood of two on-fire buildings in a calculation phase, the fire spreading probability of the un-fired building in the neighborhood of the two on-fire buildings is calculated respectively. It is determined whether the un-fired building will be ignited by the two on-fire buildings.

[0041] In the building neighborhood of the new calculation phase, if a building has been spread in the previous calculation phase, the repeated fire spreading probability calculation is not performed. If the building does not meet the un-fired condition, it is determined.

[0042] If there is no on-fire building in the building neighborhood of the new calculation phase or the fire spreading probability P<0.5 of the building, the simulation is stopped.

[0043] Further, since heat radiation is the main heat transfer path during the outward spreading of the fire from the on-fire building, for multiple buildings that will be ignited in the same calculation phase and the same building neighborhood, there is a sequence due to different fire spreading probabilities. The building with a larger probability is in the front of the sequence. Therefore, in order to distinguish the ignition sequence of the buildings ignited in the same calculation phase, the buildings that meet both the un-fired and fire spreading probability P≥0.5 conditions in the same calculation phase are compared in size in step S5. The building with a larger probability is in the front of the sequence, and the building with a smaller probability is in the back of the sequence.

[0044] ​If there are multiple adjacent buildings of the same calculation stage, the adjacent buildings of the first fire building in the previous stage are arranged in the sequence of the spread, and the buildings in the adjacent buildings of the first fire building that meet the conditions of not being on fire and the fire spread probability P being greater than or equal to 0.5 are arranged in the sequence of the spread.

[0045] Further, the fire spread between buildings is not a process in which a single building affects another building, one fire building can cause multiple eligible buildings to catch fire, and one non-fire building can receive heat energy from multiple fire buildings and be ignited by multiple buildings at the same time, so that more buildings with more fire spread paths play a key role in the fire spread in the traditional village. In step S6, when the non-fire building has a fire spread probability P greater than or equal to 0.5 in the adjacent building of a certain fire building, one spread path is marked between the non-fire building and the fire building, and after the fire spread prediction simulation is stopped, the number of fire spread paths generated by the buildings in the traditional village is counted, and a fire spread sequence diagram and a key building distribution diagram of multiple spread paths of the target traditional village are drawn;

[0046] If there are n non-fire buildings with a probability P greater than or equal to 0.5 in the adjacent building of the same fire building, n outwardly diffusing spread paths are formed.

[0047] If a non-fire building is in the adjacent building of n fire buildings at the same calculation stage, and the corresponding spread probability P is greater than or equal to 0.5, the non-fire building has n spread paths pointing to the non-fire building.

[0048] In a second aspect, the present application provides a lightweight traditional village fire spread prediction system for executing the lightweight traditional village fire spread prediction method described above, and the high-speed communication system comprises:

[0049] A data set making module is configured to import the surveyed traditional village CAD building block plan into a geographic information processing software GIS platform to form a data set T and a verification set Y, wherein the buildings in the data set T are assigned actual disaster conditions, building combustible material characteristic attributes, and burning environment characteristic attributes, and the buildings in the verification set Y are assigned building combustible material characteristic attributes and burning environment characteristic attributes.

[0050] An influence factor screening module is configured to put the data set T into a binary classification logistic regression analysis model to screen out main fire spread influence factors with significant correlation.

[0051] A spread probability calculation module is configured to put the main fire spread influence factors with significant correlation into a binary classification logistic regression analysis model, calculate the correlation coefficients corresponding to the influence factors, and construct a fire spread probability calculation formula.

[0052] A fire spread simulation mode construction module is configured to determine a building neighborhood range, calculate a fire spread probability in stages, define a building with a probability P greater than or equal to 0.5 as a burning building in a next calculation stage, form a building neighborhood of the next calculation stage, calculate a probability of a building not burning in the building neighborhood of the next calculation stage, and stop the simulation when there is no burning building in the building neighborhood of the next calculation stage or the fire spread probability P of the building is less than 0.5.

[0053] A fire spread judgment standard formulation module is configured to determine a burning sequence of a building not burning and with a probability P greater than or equal to 0.5 in a same calculation stage and a same burning building neighborhood according to a probability size, and arrange a sequence in front of a building with a large probability and arrange a sequence in back of a building with a small probability.

[0054] A data visualization module is configured to visualize a fire spread sequence, mark a path direction of the fire spread and a sequence of the building burning, count a number of fire spread paths of a traditional village building, and obtain a fire spread sequence diagram of a target traditional village and a key building distribution diagram of multiple spread paths.

[0055] In a third aspect, the present application provides an electronic device including a processor and a memory for storing a program executable by the processor, and the processor implements the lightweight traditional village fire spread prediction method when executing the program stored in the memory.

[0056] In a fourth aspect, the present application provides a storage medium storing a program, and the program is executed by a processor to implement the lightweight traditional village fire spread prediction method.

[0057] Compared with the prior art, the present application has the following advantages and beneficial effects:

[0058] (1) The method of the present application proposes a whole fire spread prediction method for a traditional village building group, and comprehensively predicts a space and a time of a fire occurrence. The method can judge whether a fire spread path from a burning building to a target building not burning by calculating a fire spread probability of the target building not burning and the burning building in stages. The method can realize a prediction of a fire spread sequence of the whole traditional village by calculating the fire spread probability of different calculation stages in turn. The method can determine a key building affecting the fire spread by counting a number of spread paths of each building, and lay a theoretical foundation for realizing accurate fire prevention and control of the traditional village.

[0059] (2) The method comprehensively considers the complex combustion properties of the traditional village to determine the main influencing factors. According to the complex combustion properties of the traditional village, the index layer factors are sorted out based on the principle of fire dynamics, so as to ensure that the selected factors comprehensively cover the potential disaster factors of the fire spread between the buildings in the traditional village. Based on the binary logistic regression analysis model, the main influencing factors of the fire spread in the traditional village are screened out, the fire spread probability calculation formula retains the main influencing factors of the fire spread with significant correlation, and the non-key factors are removed, so as to increase the accuracy and scientificity of the fire spread probability prediction.

[0060] (3) The present application proposes a light-weight fire spread probability calculation formula and a fire spread simulation mode. The fire spread probability calculation formula based on the binary logistic regression analysis model proposed by the present application condenses the main influencing factors of the fire spread probability calculation. The fire spread simulation mode constructed by the present application takes the neighborhood as the minimum spread unit of each calculation stage, simplifies the calculation steps of the fire spread sequence, and ensures the efficiency and effect of the fire spread model. The present application can reduce redundant calculation steps, save manpower and material resources, realize efficient evaluation of the fire spread sequence in the traditional village, and improve the efficiency of fire protection deployment. BRIEF DESCRIPTION OF DRAWINGS

[0061] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings needed in the embodiment description will be briefly introduced. Obviously, the drawings in the following description are only some embodiments of the present application, and other drawings can be obtained by those skilled in the art without creative labor.

[0062] Figure 1 is a flowchart of a light-weight traditional village fire spread prediction method disclosed in the present application;

[0063] Figure 2 is a schematic diagram of a fire spread mode based on neighborhood disclosed in the present application;

[0064] Figure 3 is a schematic diagram of six calculation stages of fire spread prediction in embodiment 1 of the present application;

[0065] Figure 4 is a schematic diagram of the fire spread prediction result and actual situation in embodiment 1 of the present application;

[0066] Figure 5 is a distribution diagram of key buildings of multiple spread paths in embodiment 1 of the present application;

[0067] Figure 6 is a schematic diagram of six calculation stages of fire spread prediction in embodiment 2 of the present application;

[0068] Figure 7 is a schematic diagram of fire spread prediction results in Embodiment 2 of the present application;

[0069] Figure 8 is a key building distribution diagram of multiple spread paths in Embodiment 2 of the present application;

[0070] Figure 9a is a schematic diagram of actual fire spread sequence in Jinzhu Zhuang Village, Figure 9b is a schematic diagram of actual fire spread sequence prediction results in Jinzhu Zhuang Village in Embodiment 1 of the present application, Figure 9c is a schematic diagram of actual fire spread sequence prediction results in Jinzhu Zhuang Village in Embodiment 2 of the present application;

[0071] Figure 10 is a structural block diagram of a lightweight traditional village fire spread prediction system in Embodiment 3 of the present application;

[0072] Figure 11 is a structural block diagram of an electronic device in Embodiment 4 of the present application. DETAILED DESCRIPTION

[0073] In order to enable persons skilled in the art to better understand the present application, the technical solutions in the embodiments of the present application will be described clearly and completely below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only some of the embodiments of the present application, not all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by persons skilled in the art without creative work fall within the scope of protection of the present application.

[0074] In the present application, the phrase “embodiment” means that the specific features, structures or characteristics described in combination with the embodiment can be included in at least one embodiment of the present application. The appearance of this phrase in various places in the specification does not necessarily mean the same embodiment, nor is it an independent or alternative embodiment to other embodiments. It is explicitly and implicitly understood by those skilled in the art that the embodiments described in the present application can be combined with other embodiments.

[0075] Embodiment 1

[0076] Figure 1 is a flowchart of a lightweight traditional village fire spread prediction method disclosed in the present application. The present embodiment discloses a lightweight traditional village fire spread prediction method, which comprises the following steps:

[0077] S1, making a traditional village fire spread dataset and a verification set: importing the surveyed traditional village CAD building block plan into a geographic information processing software GIS platform to form a dataset T and a verification set Y, wherein the buildings in the dataset T are assigned actual disaster conditions, building combustible material characteristic attributes, and burning environment characteristic attributes, and the buildings in the verification set Y are assigned building combustible material characteristic attributes and burning environment characteristic attributes;

[0078] In the above step S1, the present embodiment first surveys the building blocks and village terrain of the traditional villages Wending Village, Xiaozhai Village, and Jinzhu Zhuang Village in the mountainous area of southern China that have been greatly affected by fires using CAD software, imports them into a geographic information processing software GIS platform, and assigns building fire spread related attributes. The dataset T contains two traditional village samples, Wending Village and Xiaozhai Village, a total of 557 buildings, and the building assignments include six index factors of actual disaster conditions, building combustible material characteristic attributes, and burning environment characteristic attributes, specifically: actual disaster conditions, fixed fire load density Qi , mobile fire load density Qm , building surface area S f , building opening area S 0 , and building proximity D and building relative wind direction position R. The verification set Y contains one traditional village sample, Jinzhu Zhuang Village, a total of 61 buildings, and includes six index factors of building combustible material characteristic attributes and burning environment characteristic attributes, specifically: fixed fire load density Qi , mobile fire load density Qm , building surface area S f , building opening area S 0 , building proximity D and building relative wind direction position R. Part of the fire spread dataset T is shown in Table 1, and part of the fire spread verification set Y is shown in Table 2.

[0079] Table 1. Part of the fire spread dataset T data table of Example 1

[0080]

[0081] Table 2. Part of the fire spread verification set Y data table of Example 1

[0082]

[0083] S2, screening main fire spread influencing factors: placing the dataset T into a binary classification logistic regression analysis model to screen out main fire spread influencing factors with significant correlation;

[0084] In step S2, the data set T in step S1 is put into a binary logistic regression analysis model, the actual disaster situation is taken as a dependent variable, and six index layer factors are taken as independent variables to calculate the fire spread probability when the target building is ignited by the building on fire, and the calculation formula is as follows:

[0085] Formula (1)

[0086] In the formula, y =1 indicates that an event occurs, P y =1) indicates a probability value when an event occurs; x i is the ith independent variable, i=1, 2, 3, …n; n is the total number of independent variables, is the regression calculation-based x i corresponding logistic regression coefficient;

[0087] In this embodiment, χ 2 =131.241, the significance value of the model is <0.005, the binary logistic regression analysis model has statistical significance, and the significance values of the independent variable factors, the moving fire load density Qm , and the building surface area S f are >0.05, which do not have significant correlation in the analysis (Table 3), so these two index layer factors are excluded and not included in the construction of the subsequent fire spread probability calculation formula.

[0088] Table 3. Binary logistic regression analysis results of traditional village fire spread data

[0089]

[0090] S3, determining the fire spread probability calculation formula: putting the fire spread main influencing factors with significant correlation into a binary logistic regression analysis model, calculating the corresponding correlation coefficients of the influencing factors, and constructing the fire spread probability calculation formula;

[0091] In step S3, the fire spread probability calculation formula is determined, and the process includes:

[0092] 1. The fire spread main influencing factors screened out in step S2 are put into a binary logistic regression analysis model as independent variables, the actual disaster situation is taken as a dependent variable, the fire spread probability when the target building is ignited by the building on fire is calculated, and the calculation formula is the same as the above formula (1), in which, is the regression calculation-based x i corresponding logistic regression coefficient; the independent variable x ​i The logistic regression coefficients of each independent variable calculated according to the main factors of fire spread screened out are shown in Table 4;

[0093] Table 4. Results of binary logistic regression analysis of main factors of fire spread

[0094]

[0095] 2. A fire spread probability calculation formula is constructed according to formula (1) as follows:

[0096] Formula (3)

[0097] In the formula, y =1 indicates that a building is ignited by the building on fire, P y =1) indicates the probability value of a building being ignited by the building on fire.

[0098] S4. Constructing a fire spread simulation model based on neighborhood: determining the building neighborhood range, calculating the fire spread probability step by step in stages, defining the building with a probability P≥0.5 as the building on fire in the next calculation stage to form the building neighborhood of the new calculation stage, calculating the probability of the building not on fire in the building neighborhood of the new calculation stage, and stopping the simulation until there is no building on fire in the building neighborhood of the new calculation stage or the fire spread probability P<0.5 of the building.

[0099] ​In step S4, the fire spread simulation mode is a phased spread, and the spread process of the fire between buildings is simplified as gradually expanding outward in stages. The building neighborhood range is defined as the farthest range when the fire directly spreads outward from each burning building in each stage. A buffer zone of 10 m outward from the outer contour of the burning building is formed as the building neighborhood of the burning building. When a building is on fire, the fire spread neighborhood caused by the burning building is formed, and it is further determined whether the buildings located in the neighborhood of the burning building meet the condition of not being on fire. If not, the fire spread probability from the burning building to the unburned building is calculated according to formula (3). If only a part of a building is located in the neighborhood of the burning building, the building is also regarded as a potential object of fire spread of the burning building, and it is necessary to determine whether the building is on fire. If the condition is met, the fire spread probability of the building in the neighborhood is calculated. If there are multiple unburned buildings with a spread probability P≥0.5 in a neighborhood, they are defined as being ignited in the same calculation stage. When the buildings in the neighborhood of the burning building are ignited, the calculation of the fire spread neighborhood of the next stage has already been generated at the same time, forming the neighborhood of the burning building of the new calculation stage. If an unburned building is located in the neighborhoods of two burning buildings in a calculation stage, the spread probabilities of the unburned building in the two neighborhoods are calculated respectively, and it is determined whether the unburned building will be spread by fire from the two buildings at the same time. In the neighborhood of the building in the new calculation stage, if a building has been spread in the previous calculation stage, the repeated spread probability calculation is not performed, that is, it has been determined that the condition of not being on fire is not met. If there is no burning building or the fire spread probability P<0.5 of the building in the neighborhood of the building in the new calculation stage, the simulation is stopped. In this embodiment, the actual fire buildings of Jinzhu Zhuang Village in 2017 in the verification set Y are set as the fire buildings in the prediction model to perform fire spread prediction, and six calculation stages are formed, and a total of 25 buildings are spread by fire Figure 3 ). Through statistical analysis of the predicted values and the actual values of the model, the accuracy rate of the prediction of the buildings being ignited is 75.7%, the accuracy rate of the prediction of the buildings not being ignited is 83.8%, and the accuracy rate of the prediction of all building samples is 79.4%.

[0100] S5, a fire spread judgment standard is formulated: in the same calculation stage, the unburned buildings with a probability ≥0.5 in the same neighborhood of the burning building are determined in the ignition order according to the probability size, and the one with a larger probability is in the front of the spread sequence, and the one with a smaller probability is in the back of the spread sequence.

[0101] In step S5, the buildings in the same calculation stage and the same building neighborhood that meet the two conditions of not being on fire and fire spread probability P≥0.5 are compared in size, and the building with the larger probability spreads first, and the building with the smaller probability spreads later. If there are multiple building neighborhoods generated by the buildings on fire in the same calculation stage, the spread sequence of the building that meets the above two conditions in the building neighborhood of the building on fire that ranks first in the spread sequence in the previous stage is arranged first. After the fire spread prediction in this embodiment, the 25 buildings affected by the fire spread are sorted, the predicted spread sequence is basically the same as the actual spread sequence, the number of buildings affected by the fire spread is consistent, and only the local spread sequence is different.

[0102] S6, data visualization: the fire spread sequence is visualized, the path direction of the fire spread and the sequence of the buildings on fire are marked, the number of fire spread paths of the traditional village buildings is counted, and the fire spread sequence diagram and the key building distribution diagram of the target traditional village are obtained;

[0103] In step S6, when the fire spread probability P≥0.5 of the building not on fire in the neighborhood of a building on fire, a spread path is marked between the building not on fire and the building on fire. After the fire spread prediction simulation is stopped, the number of fire spread paths generated by the buildings in the traditional village is counted, and the fire spread sequence diagram and the key building distribution diagram of the target traditional village are drawn. If there are n buildings not on fire with a probability P≥0.5 in the neighborhood of the same building on fire, n spread paths that spread outward are formed. If a building not on fire is in the neighborhood of n buildings on fire in a calculation stage, and the corresponding spread probability P≥0.5, the building has n spread paths pointing to the building. In this embodiment, the path direction of the fire spread and the sequence of the buildings on fire are marked, the number of fire spread paths of the traditional village buildings is counted, and the fire spread sequence diagram ( Figure 4 ) and the key building distribution diagram ( Figure 5 ) of the target traditional village are obtained. It is judged that there are 2 key buildings with more than 3 spread paths.

[0104] Embodiment 2

[0105] Based on steps S1 to S6 of the above embodiment 1, this embodiment further discloses a method for predicting the fire spread in a light-weight traditional village, comprising the following steps:

[0106] S1, this step refers to step S1 in embodiment 1;

[0107] S2, screening the main influencing factors of fire spread: putting the data set T into a binary classification logistic regression analysis model to screen the main influencing factors of fire spread with significant correlation;

[0108] In step S2, the calculation results of this embodiment are the same as those of embodiment 1, and χ2 = 131.241, the significance value of the model <0.005, the logistic regression model is statistically significant. Although the independent variable factor mobile fire load density Qm , building surface area S f The significance value is >0.05, which is not significantly related in the analysis, but the correlation coefficient value of mobile fire load density Qm , building surface area S f The correlation coefficient value is small (Table 5), which has little effect on probability calculation, so it is also included in the construction of the subsequent probability calculation formula.

[0109] Table 5. Binary logistic regression analysis results of 6 index layer factors

[0110]

[0111] S3, determine the fire spread probability calculation formula: put the main influencing factors of fire spread with significant correlation into the binary logistic regression analysis model, calculate the correlation coefficient corresponding to the influencing factor, and construct the fire spread probability calculation formula;

[0112] In the above step S3, the fire spread probability calculation formula is determined, and the process includes: constructing the fire spread probability model according to formula (1) according to the 6 index layer factors and their correlation coefficients in step S2, and the model is constructed as follows:

[0113] Formula (4)

[0114] In the formula, y =1 indicates that a building is ignited by a burning building, P ( y =1) indicates the probability value of a building being ignited by a burning building.

[0115] S4, construct a neighborhood-based fire spread simulation model: determine the building neighborhood range, calculate the fire spread probability in stages according to formula (4), define the buildings with probability P≥0.5 in the next calculation stage as the burning buildings, form the building neighborhood of the new calculation stage, and calculate the probability of the buildings in the new calculation stage that have not been ignited until there is no burning building in the building neighborhood of the new calculation stage or the fire spread probability P<0.5 of the building, and the simulation stops;

[0116] In the above step S4, the present embodiment sets the actual fire buildings of Jinzhuangzhai in 2017 in the verification set Y as the fire buildings in the prediction model, performs fire spread prediction, forms 6 calculation stages, and a total of 18 buildings are spread by fire Figure 6). Through statistical analysis of the predicted values and actual values of the model, the accuracy rate of the prediction of the buildings being ignited was 65.2%, the accuracy rate of the prediction of the buildings not being ignited was 91.9%, and the accuracy rate of the prediction of all the building samples was 81.7%.

[0117] S5, formulating a fire spread judgment standard: in the same calculation stage, the buildings in the same fire-ignited building neighborhood that are not on fire and have a probability of ≥0.5 are arranged in a fire-ignited sequence according to the probability, and the building with a larger probability is arranged in a front position in the sequence, and the building with a smaller probability is arranged in a rear position in the sequence;

[0118] In the above step S5, the fire spread of the 18 buildings is sequenced after the fire spread prediction in the embodiment, and the predicted spread sequence and the actual spread sequence have a large difference, and the fire spread prediction of the 8 buildings in the north region of Jinzhuzhuang Village is opposite to the actual situation.

[0119] S6, data visualization: the fire spread sequence is visualized, the path direction of the fire spread and the sequence of the building being ignited are marked, the number of the fire spread paths of the buildings in the traditional village is counted, and the fire spread sequence diagram of the target traditional village and the key building distribution diagram of the multiple spread paths are obtained;

[0120] In the above step S6, the path direction of the fire spread and the sequence of the building being ignited are marked, the number of the fire spread paths of the buildings in the traditional village is counted, and the fire spread sequence diagram of the target traditional village and the key building distribution diagram of the multiple spread paths are obtained in the embodiment. Figure 7 Figure 8 There is one key building with a number of spread paths > 3.

[0121] The embodiment is basically the same as the method of the provided embodiment 1, and the main difference is that the index layer factors without significant correlation are not removed when the fire spread probability calculation formula is determined, and the six index layer factors and the correlation coefficients are all included in the formula. The calculation results of embodiment 1 and embodiment 2 are listed in Table 6, the accuracy rate of the prediction of the buildings being ignited in embodiment 1 is 75.7%, the accuracy rate of the prediction of the buildings not being ignited in embodiment 2 is 65.2%. From the data, although the accuracy rate of the prediction of all types in embodiment 2 is higher than that in embodiment 1, the predicted probability of the building fire spread is generally low, which leads to the low accuracy rate of the prediction of the building being ignited, and the result will lead to the fact that the fire fighting intervention means cannot completely cover the buildings with the risk of fire spread, so the operation of embodiment 1 is better than that of embodiment 2.

[0122] Table 6. Comparison table of data of different embodiments

[0123]

[0124] ​According to the correlation with the actual fire disaster situation, the index layer factor is screened, which will affect the simplicity and accuracy of the fire spread probability calculation formula constructed. If the index layer factor without significant correlation is not removed and directly included in the formula construction, although the information is not lost, there may be a large amount of redundant or irrelevant features in the data set, the model has noise interference, which leads to overfitting of the probability model and cannot accurately predict the probability. By comparing Example 1 and the actual fire spread prediction graph (the actual result and the prediction result comparison graph is shown in Figure 9a 、 Figure 9b and Figure 9c ), it is found that in Example 2, when determining the fire spread probability calculation formula, the index layer factor without significant correlation is not removed, which overestimates the possibility of fire spread in actual prediction, and is far from the real fire spread situation. The probability of the building being spread is too small, and a large number of spread paths and spread buildings are missing, which leads to inaccurate fire spread prediction results.

[0125] Example 3

[0126] With reference to Figure 10 , the present embodiment provides a lightweight traditional village fire spread prediction system, which comprises a data set making module 301, an influence factor screening module 302, a spread probability calculation module 303, a fire spread simulation mode construction module 304, a fire spread judgment standard making module 305 and a data visualization module 306 connected in sequence, wherein:

[0127] The data set making module 301 is used for importing the surveyed traditional village CAD building block plan into a geographic information processing software GIS platform to form a data set T and a verification set Y, wherein the buildings in the data set T are assigned with actual disaster situation, building combustible feature attribute and burning environment feature attribute, and the buildings in the verification set Y are assigned with building combustible feature attribute and burning environment feature attribute;

[0128] The influence factor screening module 302 is used for putting the data set T into a binary classification logistic regression analysis model to screen out the main influence factors of fire spread with significant correlation;

[0129] The spread probability calculation module 303 is used for putting the main influence factors of fire spread with significant correlation into a binary classification logistic regression analysis model, calculating the correlation coefficients corresponding to the influence factors, and constructing a fire spread probability calculation formula;

[0130] The fire spread simulation mode construction module 304 is configured to determine the building neighborhood range, calculate the fire spread probability in stages, define the building with a probability P greater than or equal to 0.5 as the next calculation stage of the building on fire, form a new calculation stage of the building neighborhood, calculate the probability of the building on fire in the new calculation stage of the building neighborhood, and stop the simulation until there is no building on fire in the new calculation stage of the building neighborhood or the fire spread probability P of the building is less than 0.5.

[0131] The fire spread judgment standard formulation module 305 is configured to determine the fire sequence of the building on fire in the same calculation stage and the same building neighborhood according to the probability, and the building on fire with a larger probability has a higher sequence in the fire spread sequence.

[0132] The data visualization module 306 is configured to perform the visualization processing on the fire spread sequence, mark the path direction of the fire spread and the sequence of the building on fire, count the number of the fire spread paths of the building in the traditional village, and obtain the fire spread sequence diagram of the target traditional village and the key building distribution diagram of the multiple spread paths.

[0133] Embodiment 4

[0134] The embodiment provides an electronic device, which can be a computer, such as Figure 11 As shown in the figure, the processor 402, the memory, the input device 403, the display 404 and the network interface 405 are connected through the system bus 401, the processor is used to provide calculation and control capability, the memory includes a non-volatile storage medium 406 and an internal memory 407, the non-volatile storage medium 406 stores an operating system, a computer program and a database, the internal memory 407 provides an environment for the running of the operating system and the computer program in the non-volatile storage medium, and when the processor 402 executes the computer program stored in the memory, the above-mentioned embodiment 1 provides a light-weight traditional village fire spread prediction method, and the light-weight traditional village fire spread prediction method includes the following steps:

[0135] S1, making a traditional village fire spread dataset and a verification set: importing the surveyed traditional village CAD building block plan into a geographic information processing software GIS platform to form a dataset T and a verification set Y, wherein the buildings in the dataset T are assigned actual disaster conditions, building combustible feature attributes and burning environment feature attributes, and the buildings in the verification set Y are assigned building combustible feature attributes and burning environment feature attributes;

[0136] S2, screening the main influence factors of the fire spread: putting the dataset T into a binary classification logistic regression analysis model to screen the main influence factors of the fire spread with significant correlation;

[0137] S3, determining a fire spread probability calculation formula: putting the main influencing factors of fire spread with significant correlation into a binary logistic regression analysis model, calculating the correlation coefficients corresponding to the influencing factors, and constructing a fire spread probability calculation formula;

[0138] S4, constructing a fire spread simulation model based on neighborhood: determining the building neighborhood range, calculating the fire spread probability step by step in stages, defining the buildings with a probability P≥0.5 as the burning buildings in the next calculation stage, forming the building neighborhood of the new calculation stage, and calculating the probability of the buildings that have not burned in the new calculation stage until there is no burning building in the new calculation stage or the fire spread probability P<0.5 of the building, and the simulation stops;

[0139] S5, formulating a fire spread judgment standard: in the same calculation stage, the buildings that have not burned and have a probability P≥0.5 in the same burning building neighborhood are determined in the burning order according to the probability, and the one with a larger probability is in the front of the burning sequence, and the one with a smaller probability is in the back of the burning sequence;

[0140] S6, data visualization: visualizing the fire spread sequence, marking the path direction of the fire spread and the sequence of the building burning, and counting the number of fire spread paths of the traditional village buildings to obtain the fire spread sequence diagram of the target traditional village and the key building distribution diagram of multiple spread paths.

[0141] Example 5

[0142] The embodiment provides a storage medium, which is a computer readable storage medium, and stores a computer program, the computer program is executed by a processor to realize a light traditional village fire spread prediction method provided in the above embodiment 1, the light traditional village fire spread prediction method comprises the following steps:

[0143] S1, making a traditional village fire spread dataset and a verification set: importing the surveyed traditional village CAD building block plan into a geographic information processing software GIS platform to form a dataset T and a verification set Y, wherein the buildings in the dataset T are assigned actual disaster conditions, building combustible feature attributes and burning environment feature attributes, and the buildings in the verification set Y are assigned building combustible feature attributes and burning environment feature attributes;

[0144] S2, screening main influencing factors of fire spread: putting the dataset T into a binary logistic regression analysis model to screen out main influencing factors of fire spread with significant correlation;

[0145] S3, determining a fire spread probability calculation formula: putting the main influencing factors of fire spread with significant correlation into a binary logistic regression analysis model, calculating the correlation coefficients corresponding to the influencing factors, and constructing a fire spread probability calculation formula;

[0146] S4, constructing a neighborhood-based fire spread simulation model: determining the building neighborhood range, calculating the fire spread probability in stages, defining the building with a probability P>0.5 as the next calculation stage of the burning building, forming the building neighborhood of the new calculation stage, calculating the probability of the unburned buildings in the new calculation stage of the building neighborhood, until there is no burning building in the new calculation stage of the building neighborhood or the fire spread probability P<0.5 of the building, and the simulation stops;

[0147] S5, formulating the fire spread judgment standard: in the same calculation stage, the unburned buildings with a probability P>0.5 in the same burning building neighborhood, determine the burning sequence according to the probability, the larger the probability, the earlier the spread sequence, and the smaller the probability, the later the spread sequence;

[0148] S6, data visualization: visualizing the fire spread sequence, marking the path direction of the fire spread and the sequence of building burning, and counting the number of fire spread paths of the traditional village buildings, to obtain the fire spread sequence diagram and the key building distribution diagram of multiple spread paths of the target traditional village.

[0149] The technical features of the above embodiments can be combined in any way. To make the description concise, not all possible combinations of the technical features in the above embodiments are described, but as long as the combinations of the technical features do not exist, they should be considered as within the scope of the present disclosure.

[0150] The above embodiments are the preferred embodiments of the present application, but the embodiments of the present application are not limited to the above embodiments, and any changes, modifications, substitutions, combinations, simplifications made without departing from the spirit and principles of the present application shall be equivalent replacement methods, and shall be within the scope of protection of the present application.

Claims

1. A lightweight traditional village fire spread prediction method, characterized by, The prediction method comprises the following steps: S1, making a traditional village fire spread dataset and a verification set: importing the surveyed traditional village CAD building block plan into a geographic information processing software GIS platform to form a dataset T and a verification set Y, wherein the buildings in the dataset T are assigned actual disaster conditions, building combustible material characteristic attributes and burning environment characteristic attributes, and the buildings in the verification set Y are assigned building combustible material characteristic attributes and burning environment characteristic attributes; S2, screening main fire spread influencing factors: putting the dataset T into a binary classification logistic regression analysis model to screen main fire spread influencing factors with significant correlation; S3, determining a fire spread probability calculation formula: putting the main fire spread influencing factors with significant correlation into a binary classification logistic regression analysis model, calculating the correlation coefficients corresponding to the influencing factors, and constructing a fire spread probability calculation formula; wherein the process of determining the fire spread probability calculation formula is as follows: The main fire spread influencing factors screened in step S2 are put into a binary classification logistic regression analysis model as independent variables The actual disaster situation is taken as the dependent variable, and the fire spread probability of the target building when ignited by the building is calculated, and the formula is as follows: Equation (2) In the formula, y =1 indicates that a building is ignited by a fire building, P y =1) indicates the probability value of a building being ignited by a fire building; is the logistic regression coefficient corresponding to the main influencing factor of fire spread based on regression calculation, i=1, 2, 3, …n, n is the total number of independent variables, n=6;​ S4, constructing a neighborhood-based fire spread simulation mode: determining a building neighborhood range, gradually calculating fire spread probabilities in stages, defining buildings with a probability P greater than or equal to 0.5 as ignited buildings in the next calculation stage, forming a building neighborhood in the new calculation stage, calculating the probabilities of buildings not ignited in the building neighborhood in the new calculation stage, and stopping the simulation until there are no ignited buildings in the building neighborhood in the new calculation stage or the fire spread probability P of the buildings is less than 0.5; defining a neighborhood as a buffer zone formed by expanding the outline of an ignited building outward by 10 m; the fire spread simulation mode in the step S4 is a stage-based spread, which simplifies the process of fire spread between buildings into a form of gradually expanding outward in stages, and defines the building neighborhood range as the farthest range of direct outward spread of fire of each ignited building in each stage, and the buffer zone formed by expanding the outline of the ignited building outward by 10 m as the building neighborhood; When a building is ignited, a fire spread neighborhood caused by the ignited building is formed, and it is further determined whether the buildings located in the neighborhood of the ignited building meet the condition of not being ignited; if not, the fire spread probability from the ignited building to the unignited building is calculated according to formula (2); wherein if a building only has a part in the neighborhood of the ignited building, the building is also regarded as a potential object of fire spread of the ignited building, and it is necessary to determine whether the building is ignited, and the fire spread probability of the building in the neighborhood is calculated if the condition is met; If there are multiple unignited buildings with a fire spread probability P greater than or equal to 0.5 in a neighborhood, they are defined as being ignited in the same calculation stage, that is, when the buildings in the neighborhood of the ignited building are ignited, the calculation of the fire spread neighborhood in the next stage has already been produced at the same time, forming the neighborhood of the ignited building in the new calculation stage; If an unignited building is in the neighborhoods of two ignited buildings in a calculation stage, the fire spread probabilities of the unignited building in the neighborhoods of the two ignited buildings are calculated respectively, and it is determined whether the unignited building will be spread by fire from the two ignited buildings. If a building in the building neighborhood of the new calculation stage has been spread in the previous calculation stage, the repeated spread probability calculation is not performed, that is, it has been determined that it does not meet the condition of not yet being on fire; If there is no building on fire or the fire spread probability P of the building in the building neighborhood of the new calculation stage is less than 0.5, the simulation stops; S5, a fire spread judgment standard is formulated: in the same calculation stage, the buildings in the same on-fire building neighborhood that are not on fire and have a probability P greater than or equal to 0.5 are arranged in a fire spread sequence according to the probability, and the building with a larger probability is arranged in the front of the spread sequence, and the building with a smaller probability is arranged in the back of the spread sequence; S6, data visualization: the fire spread sequence is visualized, the path direction of the fire spread and the sequence of the building on fire are marked, the number of the fire spread paths generated by the buildings in the traditional village is counted, and the fire spread sequence diagram of the target traditional village and the key building distribution diagram of the multiple spread paths are obtained.

2. The light-weight traditional village fire spread prediction method according to claim 1, characterized in that, In the step S1, the buildings in the data set T are assigned with actual disaster conditions, building combustible material characteristic attributes and burning environment characteristic attributes, and the buildings in the verification set Y are assigned with building combustible material characteristic attributes and burning environment characteristic attributes, wherein the actual disaster condition is a binary variable, and the target building is marked as 1 if it is ignited by the on-fire building, and the target building is marked as 0 if it is not ignited by the on-fire building; The building combustible characteristic attribute is composed of four index layer factors: fixed fire load density Qi , mobile fire load density Qm , building surface area S f , building opening area S 0 ; the burning environment characteristic attribute is composed of two index layer factors: building proximity D , building relative wind direction position R ; wherein, the building relative wind direction position R is a classification variable, and the remaining five index layer factors are continuous variables, taking values obtained by actual observation and measurement, and the specific statistical standards are as follows: ① Fixed fire load density Qi Definition: The fire load density of building structure and components, which is determined by the materials and structure of the building. The reference type values are: reinforced concrete structure: 0 MJ / ㎡, masonry structure: 200 MJ / ㎡, rammed earth structure: 300 MJ / ㎡, brick-wood structure / stone-wood structure: 1000 MJ / ㎡, all-wood structure: 1500 MJ / ㎡, and grass-wood structure: 2000 MJ / ㎡. ② Mobile fire load density Qm Defined as: the total fire load density of the building's indoor items, determined by the function of the building, wherein the reference type takes the value: office building: 400 MJ / ㎡, public exhibition building: 500 MJ / ㎡, residential building: 780 MJ / ㎡, leisure and entertainment building: 1000 MJ / ㎡, commercial building: 1500 MJ / ㎡, educational building: 2500 MJ / ㎡; iii. building surface area S f defined as: the sum of the facade and roof area of the building, as actually measured; (4) The building opening area S 0 The sum of the areas of the doors, windows and openings of the building facade, as actually measured. (5) Building proximity D Defined as: the nearest spatial distance between the facade of the target building and the facade of the building on fire, measured in actual distance, which is the square root of the sum of the square of the horizontal distance value and the square of the vertical distance value between the facades; Relative wind direction position of building R Definition: the position relationship between the target building and the building on fire under certain wind direction conditions, 1 for upwind position and 0 for downwind position; both in the data set T and the verification set Y, the wind direction on the day of the fire event is taken as the basis for statistics, and when simulating the prediction of traditional villages not affected by fire, the annual average wind direction of the target village is taken as the basis for statistical calculation. The data set T contains a number of traditional villages greater than or equal to 2, and a total number of buildings greater than or equal to 120.

3. The light-weight traditional village fire spread prediction method according to claim 2, characterized in that, In the step S2, the calculation formula of the binary logistic regression analysis model is as follows: Formula (1) In the formula, y =1 indicates that an event occurs, P y =1) indicates the probability value when an event occurs; x i is the ith independent variable, i=1, 2, 3, …n; n is the total number of independent variables, is the regression calculation-based x i corresponding to the logistic regression coefficient;​ In the step S1, the data set T formed is put into formula (1), the actual disaster condition is taken as the dependent variable, and the six index layer factors are taken as the independent variables, and the fire spread probability calculation when the target building is ignited by the on-fire building is performed; It is specified that in the coefficient comprehensive test of the binary logistic regression analysis model, when the significance value of the model is less than 0.005, the binary logistic regression model has statistical significance, and when the significance value of the independent variable is less than or equal to 0.05, it has significant correlation, at this time, it indicates that this independent variable can be used as a main influencing factor of the fire spread, and is included in the construction of the subsequent fire spread probability calculation formula, otherwise, the independent variable is excluded.

4. The light-weight traditional village fire spread prediction method according to claim 1, characterized in that, In the step S5, the buildings in the same calculation stage and the same building neighborhood that meet the two conditions of not being on fire and having a fire spread probability P greater than or equal to 0.5 are compared in size, the building with a larger probability is arranged in the front of the spread sequence, and the building with a smaller probability is arranged in the back of the spread sequence; If there are multiple on-fire building neighborhoods in the same calculation stage, the spread sequence of the buildings in the on-fire building neighborhood of the building that is arranged in the front of the spread sequence in the previous stage and meets the two conditions of not being on fire and having a fire spread probability P greater than or equal to 0.5 is arranged in the front.

5. The light-weight traditional village fire spread prediction method according to claim 1, characterized in that, In the step S6, when the fire spread probability P of the building not on fire in the neighborhood of a on-fire building is greater than or equal to 0.5, one spread path is marked between the building not on fire and the on-fire building, after the fire spread prediction simulation stops, the number of the fire spread paths generated by the buildings in the traditional village is counted, and the fire spread sequence diagram of the target traditional village and the key building distribution diagram of the multiple spread paths are drawn; If there are n buildings not on fire in the neighborhood of the same on-fire building and having a probability P greater than or equal to 0.5, n spread paths outwardly diffusing are formed. If an unburning building is in the neighborhood of n burning buildings at the same calculation stage, and the corresponding spread probability P is greater than or equal to 0.5, then there are n spread paths pointing to the building.

6. A light-weighted traditional village fire spread prediction system for executing the light-weighted traditional village fire spread prediction method according to any one of claims 1 to 5, characterized by The lightweight traditional village fire spread prediction comprises: A dataset making module is configured to import the surveyed traditional village CAD building block plan into a geographic information processing software GIS platform to form a dataset T and a verification set Y, wherein the buildings in the dataset T are assigned actual disaster conditions, building combustible material characteristic attributes and burning environment characteristic attributes, and the buildings in the verification set Y are assigned building combustible material characteristic attributes and burning environment characteristic attributes. An influence factor screening module is configured to put the dataset T into a binary classification logistic regression analysis model to screen out main fire spread influence factors with significant correlation. A spread probability calculation module is configured to put the main fire spread influence factors with significant correlation into a binary classification logistic regression analysis model, calculate the correlation coefficients of the influence factors, and construct a fire spread probability calculation formula. A fire spread simulation mode construction module is configured to determine the building neighborhood range, calculate the fire spread probability in stages, define the buildings with a probability P greater than or equal to 0.5 as burning buildings in the next calculation stage, form the building neighborhood of the new calculation stage, and calculate the probability of the unburning buildings in the building neighborhood of the new calculation stage until there is no burning building in the building neighborhood of the new calculation stage or the fire spread probability P of the buildings is less than 0.5, and the simulation stops. A fire spread judgment standard formulation module is configured to determine the burning order of the unburning buildings with a probability P greater than or equal to 0.5 in the same calculation stage and in the same burning building neighborhood according to the probability, and the building with a larger probability has a higher spread sequence, and the building with a smaller probability has a lower spread sequence. A data visualization module is configured to visually process the fire spread sequence, mark the path direction of the fire spread and the burning sequence of the buildings, count the number of fire spread paths generated by the buildings in the traditional village, and obtain a fire spread sequence diagram of the target traditional village and a key building distribution diagram of multiple spread paths.

7. An electronic device comprising a processor and a memory for storing a processor-executable program, characterized in that The processor executes the program stored in the memory to implement the lightweight traditional village fire spread prediction method of any one of claims 1 to 5.

8. A storage medium storing a program, characterized by comprising: The program is executed by the processor to implement the lightweight traditional village fire spread prediction method of any one of claims 1 to 5.

Images