CN113553754A - Memory, fire risk prediction model construction method, system and device - Google Patents
Memory, fire risk prediction model construction method, system and device Download PDFInfo
- Publication number
- CN113553754A CN113553754A CN202010328197.7A CN202010328197A CN113553754A CN 113553754 A CN113553754 A CN 113553754A CN 202010328197 A CN202010328197 A CN 202010328197A CN 113553754 A CN113553754 A CN 113553754A
- Authority
- CN
- China
- Prior art keywords
- variable
- data set
- fire
- generating
- prediction model
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000013058 risk prediction model Methods 0.000 title claims abstract description 61
- 238000010276 construction Methods 0.000 title claims description 21
- 238000012549 training Methods 0.000 claims abstract description 46
- 238000000034 method Methods 0.000 claims abstract description 41
- 230000001419 dependent effect Effects 0.000 claims abstract description 34
- 238000012360 testing method Methods 0.000 claims abstract description 27
- 230000002265 prevention Effects 0.000 claims abstract description 18
- 238000012216 screening Methods 0.000 claims abstract description 17
- 238000001556 precipitation Methods 0.000 claims description 9
- 238000010586 diagram Methods 0.000 claims description 8
- 238000012545 processing Methods 0.000 claims description 8
- 238000012163 sequencing technique Methods 0.000 claims description 6
- 238000004422 calculation algorithm Methods 0.000 claims description 5
- 238000002790 cross-validation Methods 0.000 claims description 5
- 238000012937 correction Methods 0.000 claims description 3
- 230000004044 response Effects 0.000 abstract description 5
- 238000007637 random forest analysis Methods 0.000 description 7
- 238000004458 analytical method Methods 0.000 description 5
- 230000000694 effects Effects 0.000 description 5
- 230000009286 beneficial effect Effects 0.000 description 4
- 238000004590 computer program Methods 0.000 description 4
- 230000008878 coupling Effects 0.000 description 3
- 238000010168 coupling process Methods 0.000 description 3
- 238000005859 coupling reaction Methods 0.000 description 3
- 230000007547 defect Effects 0.000 description 3
- 238000004364 calculation method Methods 0.000 description 2
- 238000004891 communication Methods 0.000 description 2
- 238000010219 correlation analysis Methods 0.000 description 2
- 238000011161 development Methods 0.000 description 2
- 238000012417 linear regression Methods 0.000 description 2
- 230000008569 process Effects 0.000 description 2
- 238000012795 verification Methods 0.000 description 2
- 230000008030 elimination Effects 0.000 description 1
- 238000003379 elimination reaction Methods 0.000 description 1
- 238000011156 evaluation Methods 0.000 description 1
- 230000007717 exclusion Effects 0.000 description 1
- 230000006870 function Effects 0.000 description 1
- 230000001788 irregular Effects 0.000 description 1
- 238000010295 mobile communication Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000012544 monitoring process Methods 0.000 description 1
- 238000010606 normalization Methods 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 238000003672 processing method Methods 0.000 description 1
- 239000007787 solid Substances 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
- G06F30/20—Design optimisation, verification or simulation
- G06F30/27—Design optimisation, verification or simulation using machine learning, e.g. artificial intelligence, neural networks, support vector machines [SVM] or training a model
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
- G06Q10/063—Operations research, analysis or management
- G06Q10/0635—Risk analysis of enterprise or organisation activities
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2111/00—Details relating to CAD techniques
- G06F2111/10—Numerical modelling
Landscapes
- Engineering & Computer Science (AREA)
- Business, Economics & Management (AREA)
- Human Resources & Organizations (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Strategic Management (AREA)
- Economics (AREA)
- Entrepreneurship & Innovation (AREA)
- Evolutionary Computation (AREA)
- Operations Research (AREA)
- Data Mining & Analysis (AREA)
- General Business, Economics & Management (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Development Economics (AREA)
- Tourism & Hospitality (AREA)
- Game Theory and Decision Science (AREA)
- Quality & Reliability (AREA)
- Artificial Intelligence (AREA)
- General Engineering & Computer Science (AREA)
- Marketing (AREA)
- Life Sciences & Earth Sciences (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Bioinformatics & Computational Biology (AREA)
- Evolutionary Biology (AREA)
- Educational Administration (AREA)
- Medical Informatics (AREA)
- Software Systems (AREA)
- Computer Hardware Design (AREA)
- Geometry (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention discloses a memory, a method, a system and a device for constructing a fire risk prediction model, wherein the method comprises the following steps: generating a variable data set according to a variable item related to the fire occurrence of the target area and generating an auto-variable data set; taking the fire density of the target area as a dependent variable item and generating a dependent variable data set; screening the variable items in the variable data set through feature selection to obtain preferred variable items, and generating importance degree ranking of the preferred variable items; determining a final variable term from the preferred variable terms according to the importance ranking; forming a sample data set according to the final auto-variable term and the dependent variable term; and randomly dividing a preset number of data subsets according to the sample data set, determining a training set and an internal test set from the data subsets, and generating a plurality of sub-models through training to construct a fire risk prediction model. The method can provide more accurate model reference for predicting the fire risk of the industrial park and developing more effective fire response prevention measures.
Description
Technical Field
The invention relates to the field of process industrial safety, in particular to a memory, a method, a system and a device for constructing a fire risk prediction model.
Background
The frequency and the harm of fire in the area are obviously improved due to the dense buildings and crowds in the industrial park and the abundant vegetation resources at the periphery. The occurrence of fire is closely related to factors such as terrain, climate and human activities.
The prior art mainly applies a multiple linear regression model and correlation analysis to analyze the relationship between the fire risk of the industrial park and the influence factor; the inventor finds that the method for analyzing the relationship between the fire risk and the influence factor in the industrial park by using the multiple linear regression model and the correlation analysis in the prior art at least has the following defects:
only linear relation among variables is reflected, complex nonlinear relation among fire risks and influence factors of the industrial park is ignored, and co-linearity and independence among independent variables are obvious, so that accuracy of a prediction model is low.
The information disclosed in this background section is only for enhancement of understanding of the general background of the invention and should not be taken as an acknowledgement or any form of suggestion that this information forms the prior art already known to a person skilled in the art.
Disclosure of Invention
The present invention aims to provide a more accurate model reference for predicting fire risk in an industrial park and developing more effective fire response prevention measures.
The invention provides a method for constructing a fire risk prediction model, which comprises the following steps:
s11, generating a variable data set according to the variable item related to the fire occurrence of the target area and generating a self-variable data set; taking the fire density of the target area as a dependent variable item and generating a dependent variable data set;
s12, screening the variable items in the variable data set through feature selection to obtain preferred variable items, and generating importance degree sequence of the preferred variable items; determining a final variable term from the preferred variable terms according to the importance ranking;
s13, forming a sample data set according to the final auto-variable item and the dependent variable item;
and S14, randomly dividing a preset number of data subsets according to the sample data set, determining a training set and an internal test set from the data subsets, and generating a plurality of sub-models through training to construct a fire risk prediction model.
In the present invention, the variable term related to the occurrence of a fire in the target area includes:
elevation, slope direction, normalized vegetation index, population density, road line density, railway line density, monthly average precipitation, monthly average minimum temperature, monthly average temperature and monthly average maximum temperature, and any combination thereof.
In the present invention, the generating the autovariate data set includes:
carrying out 0-1 standardization processing on specific numerical values in the variable items to generate variable data of each variable item in the independent variable data set;
in the present invention, the normalizing the specific values in the variable terms by 0-1 to generate the variable data of each variable term in the independent variable data set includes:
calculating the gradient and the numerical value type gradient value of the preset area by adopting an ArcGIS 'grid surface-gradient' tool on the basis of a digital elevation model DEM with the resolution of 30M; the numerical slope value is obtained according to the slope, and is a numerical variable between-1 and 1;
calculating a normalized vegetation index based on a 500MNDVI month synthetic product;
calculating the population density of the people in the preset area by using a grid distribution algorithm based on the population density distribution diagram of the corresponding year;
calculating the road line density and the railway line density by using an ArcMap line density tool on the basis of a road network and a railway network;
and calculating climate variables such as the monthly mean highest temperature, the monthly mean lowest temperature, the monthly mean precipitation and the like based on the monthly mean highest temperature, the monthly mean lowest temperature data set and the ground precipitation monthly data set which are observed by original remote sensing.
In the invention, the numerical slope value is obtained according to the slope, and comprises the steps of calculating the numerical slope value according to a formula (1);
aspect ═ cos (θ × 2 × pi/360), formula (1); wherein, theta is a gradient value, and Aspect is a numerical gradient value.
In the present invention, the generating a dependent variable data set by using the fire density of the target area as a dependent variable term includes:
a factor data set is generated by performing 0-1 standardization processing on the fire occurrence frequency in each square kilometer of each year in an industrial park.
In the present invention, the screening variable terms in the variable data set through feature selection to obtain preferred variable terms, and generating an importance ranking of the preferred variable terms, includes:
screening the variable items in the variable data set according to a Root Mean Square Error (RMSE) to obtain the preferred variable items;
and carrying out model training through Rstudio software to obtain the importance degree sequence of the preferred variable items.
In the present invention, the randomly dividing a preset number of data subsets according to the sample data set, determining a training set and an internal test set from the data subsets, and generating a plurality of sub-models through training to construct a fire risk prediction model, includes:
randomly dividing the sample data set into five data subsets; and taking four data subsets as a training set and the other data subset as a test set, and generating five sub-models through training according to a reference cross validation mode to construct a fire risk prediction model.
In the present invention, the method further comprises:
and judging the accuracy of the importance ranking of each variable factor according to the simple correlation coefficient of the predicted value and the actual value in the training set and the test set.
In the present invention, the variable term related to the occurrence of a fire in the target area further includes a fire prevention and control index;
the fire prevention and control index is generated according to the level of fire prevention measures of the target area.
In another aspect of the present invention, there is also provided a fire risk prediction model construction apparatus, including:
a data acquisition unit for generating a variable data set from a variable data set composed of variable items related to the occurrence of a fire in a target area and generating a self-variable data set; taking the fire density of the target area as a dependent variable item and generating a dependent variable data set;
the sorting unit is used for screening the variable items in the variable data set through feature selection to obtain preferred variable items and generating importance degree sorting of the preferred variable items; determining a final variable term from the preferred variable terms according to the importance ranking;
a sample generating unit, configured to form a sample data set according to the final auto-variable term and the dependent variable term;
and the model generation unit is used for randomly dividing a preset number of data subsets according to the sample data set, determining a training set and an internal test set from the data subsets, and generating a plurality of sub-models through training to construct a fire risk prediction model.
In the present invention, the method further comprises:
and the sequencing correction unit is used for judging the accuracy of the importance sequencing of each variable factor according to the simple correlation coefficient of the predicted value and the actual value in the training set and the test set.
In another aspect of the present invention, there is also provided a memory including a software program adapted to be executed by a processor for the steps of the above-described fire risk prediction model construction method.
In another aspect of the embodiments of the present invention, there is also provided a fire risk prediction model construction device, including a computer program stored on a memory, the computer program including program instructions, which, when executed by a computer, cause the computer to perform the method of the above aspects and achieve the same technical effects.
Compared with the prior art, the invention has the following beneficial effects:
according to the method, a random forest model is used for modeling and analyzing the fire risk of the industrial park, and the variable items and the variable numbers which are suitable for the prediction model are determined by screening independent variables and sequencing importance degrees, so that a fire risk prediction model with a high prediction effect can be constructed; the invention overcomes a series of defects when the traditional method analyzes the relationship between the fire risk and the influence factor of the industrial park, and comprises the following steps: only the linear relation among the variables is reflected, the complex nonlinear relation among the fire risks and the influence factors of the industrial park is ignored, and the co-linearity and the independence among the independent variables are obvious. The method uses a random forest model to carry out modeling analysis on the fire risk of the industrial park; the importance analysis of each influence factor of the fire risk is obtained and verified through a feature selection algorithm (feature selection) and a simple correlation coefficient, so that the fire risk and the development trend thereof can be more effectively predicted, and more accurate model reference can be provided for effective fire response prevention measures.
The foregoing description is only an overview of the technical solutions of the present invention, and in order to make the technical means of the present invention more clearly understood and to make the technical means implementable in accordance with the contents of the description, and to make the above and other objects, technical features, and advantages of the present invention more comprehensible, one or more preferred embodiments are described below in detail with reference to the accompanying drawings.
Drawings
FIG. 1 is a diagram of the steps of a method of constructing a fire risk prediction model according to the present invention;
FIG. 2 is a schematic of population density shown as Thiessen polygons in the present invention;
FIG. 3 is a graphical illustration of population density in grid values in the present invention;
FIG. 4 is a graph showing the results of screening according to the root mean square error in the present invention;
FIG. 5 is a schematic structural diagram of a fire risk prediction model construction apparatus according to the present invention;
fig. 6 is a schematic structural diagram of a fire risk prediction model construction apparatus according to the present invention.
Detailed Description
The following detailed description of the present invention is provided in conjunction with the accompanying drawings, but it should be understood that the scope of the present invention is not limited to the specific embodiments.
Throughout the specification and claims, unless explicitly stated otherwise, the word "comprise", or variations such as "comprises" or "comprising", will be understood to imply the inclusion of a stated element or component but not the exclusion of any other element or component.
In this document, the terms "first", "second", etc. are used to distinguish two different elements or portions, and are not used to define a particular position or relative relationship. In other words, the terms "first," "second," and the like may also be interchanged with one another in some embodiments.
Example one
In order to provide a more accurate model reference for predicting fire risk of an industrial park and developing more effective fire response prevention measures, as shown in fig. 1, a method for constructing a fire risk prediction model is provided in an embodiment of the present invention, and includes the steps of:
s11, generating a variable data set according to the variable item related to the fire occurrence of the target area and generating a self-variable data set; taking the fire density of the target area as a dependent variable item and generating a dependent variable data set;
the target area in the embodiment of the invention refers to an area range for predicting the fire risk of the industrial park; first, the influence factors related to the occurrence of fire are performed within the target area as variable terms, and specifically, the variable terms may include altitude, gradient, slope direction, normalized vegetation index, population density, road line density, railway line density, monthly precipitation amount, monthly minimum temperature, monthly average temperature, monthly maximum temperature, and the like.
In order to adapt the data to the model construction, in the embodiment of the present invention, a 0-1 normalization process may be further performed on specific values in the variable terms to generate specific variable data of each variable term in the independent variable data set; the specific mode can be as follows:
1) calculating the gradient and the numerical value type gradient value of the preset area by adopting an ArcGIS 'grid surface-gradient' tool on the basis of a digital elevation model DEM with the resolution of 30M; the numerical slope value is obtained according to the slope, and is a numerical variable between-1 and 1; preferably, when obtaining the numerical slope value according to the slope value, the numerical slope value may be calculated according to formula (1);
aspect ═ cos (θ × 2 × pi/360), formula (1); wherein theta is a gradient value, and Aspect is a numerical gradient value
2) Calculating a normalized vegetation index based on a 500MNDVI month synthetic product;
3) calculating the population density of people in the preset area by using a grid distribution algorithm on the basis of the population density distribution diagram of the corresponding year;
specifically, referring to fig. 2 and 3, a specific way of calculating the oral density of a person in a preset area according to an embodiment of the present invention may be:
in fig. 2, the central value of each of the tesson polygons represents the population density within the polygon, wherein the population count in the region where the irregular figure (corresponding to the target region in the embodiment of the present invention) delineated by the thicker line intersects with the tesson polygon can be indirectly obtained through the central value; in fig. 3, the proportion of each grid in the illustrated area is formed by the environment and other factors, and the population density in the administrative unit is estimated by the intersection of the areas; in this way, the grid is used to estimate the population density, that is, the area is gridded and then divided and counted, and the value of each pixel (grid) in fig. 3 can represent the average population per square kilometer in a certain period;
4) calculating the road line density and the railway line density by using an ArcMap 'line density' tool on the basis of a road network and a railway network;
5) and calculating climate variables such as the monthly mean maximum temperature, the monthly mean minimum temperature, the monthly mean precipitation and the like based on the monthly mean maximum temperature, the monthly mean average temperature, the monthly mean minimum temperature data set and the ground precipitation monthly data set which are originally observed by remote sensing.
Preferably, in the embodiment of the present invention, the variable item (belonging to the independent variable data set) related to the fire occurrence of the target area further includes a fire prevention and control index; the fire prevention and control index is generated according to the level of fire prevention measures of the target area. The inventor finds that the risk of fire occurrence has a great correlation with the implementation and management of the current prevention and control measures of the target area, so that the prediction accuracy of the prediction model can be effectively improved by incorporating the fire prevention and control index into the variable term related to the fire occurrence of the target area.
In practical application, the effectiveness of the prevention and control measures can be respectively endowed with different weights and quantized according to the two aspects of the existing monitoring equipment and management system, and further 0-1 standardization treatment is realized to generate the fire prevention and control index. The specific numerical values can be obtained by those skilled in the art through empirical or experimental verification, and are not specifically limited herein.
In practical applications, in the embodiment of the present invention, the fire density of the target area is taken as a dependent variable term, and a dependent variable data set is generated by the following specific method: a factor data set is generated by performing 0-1 standardization processing on the fire occurrence frequency in each square kilometer of each year in an industrial park.
The specific calculation method may include: the method comprises the steps of selecting a fixed window width of 60km as a radius and 1km as a grid resolution, calculating and extracting fire density by adopting a nuclear density estimation method, and taking the fire density as a dependent variable (in the embodiment of the invention, historical data of fire striking in 2002-2011 Anhui province 70820 can be referred to for generation of a dependent variable data set during fire risk modeling of an industrial park).
S12, screening the variable items in the variable data set through feature selection to obtain preferred variable items, and generating importance degree sequence of the preferred variable items; determining a final variable term from the preferred variable terms according to the importance ranking;
in practical application, the variable terms in the variable data set can be screened according to the root mean square error RMSE to obtain the preferred variable terms; then, model training is carried out through Rstudio software to obtain the importance ranking of the preferred variable items, specifically:
during feature selection, Recursive Feature Elimination (RFE) is used for screening independent variables, root-mean-square-errors (RMSE) are used as criteria for independent variable screening, so that the number of corresponding preferred variable items after model training is obtained, then, Rstudio is used for model training, the importance ranking of each preferred variable item is obtained, and further, the final variable item participating in random forest model training is obtained.
The idea of using feature selection to screen for independent variables comes from cross-validation, and specifically may be to use one of the subsets of data that has been equally divided into four as a test set, the other three parts are used as training sets, so that the four generated internal models are evaluated and tested, and after the evaluation and the test are repeated five times, the importance of different independent variables in the models can be obtained (as shown in table 1), then, finally selecting the independent variable (i.e. the final variable item) entering the final model according to the number of the independent variables corresponding to the optimal expression of the model, in one example, as shown in fig. 4, the optimal number of arguments corresponds to 6 when RMSE is minimal, that is, the independent variables with the top 6 ranks can be selected as final variable terms according to the ranking of the importance of the independent variables in table 1, and are used as the input of the fire risk prediction model (random forest optimal model) in the embodiment of the invention.
From table 1 and fig. 4, it can be determined that the 6 final variable terms are altitude, normalized vegetation index, population density, road line density, railway line density, and monthly minimum temperature, respectively.
Table 1:
s13, forming a sample data set according to the final auto-variable item and the dependent variable item;
after the final variable term and the dependent variable term are determined, a sample data set required by a fire risk prediction model can be constructed according to historical data corresponding to the final variable term and the dependent variable term.
And S14, randomly dividing a preset number of data subsets according to the sample data set, determining a training set and an internal test set from the data subsets, and generating a plurality of sub-models through training to construct a fire risk prediction model.
In practical application, the sample data set can be randomly divided into five data subsets; taking four data subsets as a training set and the other data subset as a test set, and generating five sub-models through training according to a reference cross validation mode to construct a fire risk prediction model, specifically:
firstly, randomly dividing data in a sample data set into five subsets, then taking every four data subsets as a training set, and taking the other data subsets as an internal test set; and (4) evaluating and testing the generated five internal models by referring to a cross validation idea, and generating five sub models by training so as to construct a fire risk prediction model.
Further, in the embodiment of the present invention, in order to make the prediction result of the final fire risk prediction model more accurate, the method may further include: judging the accuracy of the importance ranking of each variable factor according to the simple correlation coefficient of the predicted value and the actual value in the training set and the test set; therefore, more reasonable variable factor importance ranking can be obtained through the accuracy of the variable factor importance ranking, and further the ranking and selection of the final variable items of the fire risk prediction model are more reasonable.
Specifically, after training and calculation, the performance of the random forest model on a training set and a test set is integrated, and the validity of the importance ranking of each variable factor is determined by taking a simple correlation coefficient of a predicted value and an actual value as a verification standard, namely, the accuracy of the importance ranking of each variable factor is judged by the simple correlation coefficient of the predicted value and the actual value in the training set and the test set, wherein the more the simple correlation coefficient is close to 1, the more accurate the importance ranking of each variable factor is.
Table 2 shows measured values of simple correlation coefficients of predicted values and actual values in the training set and test set of five submodels in the embodiment of the present invention.
Table 2:
as can be seen from Table 2, the importance ranking of each variable factor in the embodiment of the present invention has high accuracy.
In summary, the embodiment of the invention uses the random forest model to perform modeling analysis on the fire risk of the industrial park, and determines the variable terms and the variable numbers suitable for the prediction model by screening the independent variables and sequencing the importance, so that a fire risk prediction model with higher prediction effect can be constructed; the invention overcomes a series of defects when the traditional method analyzes the relationship between the fire risk and the influence factor of the industrial park, and comprises the following steps: only the linear relation among the variables is reflected, the complex nonlinear relation among the fire risks and the influence factors of the industrial park is ignored, and the co-linearity and the independence among the independent variables are obvious. The method uses a random forest model to carry out modeling analysis on the fire risk of the industrial park; the importance analysis of each influence factor of the fire risk is obtained and verified through a feature selection algorithm and a simple correlation coefficient, so that the fire risk and the development trend thereof can be more effectively predicted, and more accurate model reference can be provided for effective fire response prevention measures.
Example two
In another aspect of the embodiment of the present invention, a fire risk prediction model construction device is further provided, and fig. 5 shows a schematic structural diagram of the fire risk prediction model construction device provided in the embodiment of the present invention, where the fire risk prediction model construction device is a device corresponding to the fire risk prediction model construction method in the embodiment corresponding to fig. 1, that is, the fire risk prediction model construction method in the embodiment corresponding to fig. 1 is implemented by using a virtual device, and each virtual module constituting the fire risk prediction model construction device may be executed by an electronic device, such as a network device, a terminal device, or a server. Specifically, the device for constructing a fire risk prediction model in the embodiment of the present invention includes:
a data acquisition unit 01 for generating a variable data set based on a variable data set composed of variable items related to the occurrence of a fire in a target area and generating a self-variable data set; taking the fire density of the target area as a dependent variable item and generating a dependent variable data set;
the sorting unit 02 is used for screening the variable items in the variable data set through feature selection to obtain preferred variable items and generating importance degree sorting of the preferred variable items;
the sample generating unit 03 is configured to determine a final variable item from the preferred variable items according to the importance ranking, and form a sample data set according to the final auto-variable item and the dependent variable item;
and the model generating unit 04 is used for randomly dividing a preset number of data subsets according to the sample data set, determining a training set and an internal test set from the data subsets, and generating a plurality of sub-models through training to construct a fire risk prediction model.
Preferably, in the embodiments of the present invention,
since the working principle and the beneficial effects of the device for constructing the fire risk prediction model in the embodiment of the present invention have been described and explained in the method for constructing the fire risk prediction model corresponding to fig. 1, they may be referred to each other and are not described herein again.
EXAMPLE III
On the basis of the second embodiment, the embodiment of the present invention may further include a middle ranking correction unit 05, configured to judge accuracy of importance ranking of each variable factor according to a simple correlation coefficient between a predicted value and an actual value in the training set and the test set.
Similarly, the working principle and the beneficial effects of the fire risk prediction model construction device in the embodiment of the present invention are also described and explained in the fire risk prediction model construction method corresponding to fig. 1, and therefore, they may be referred to each other and are not described herein again.
Example four
In an embodiment of the present invention, a memory is further provided, where the memory includes a software program, and the software program is adapted to enable the processor to execute each step of the method for constructing a fire risk prediction model corresponding to fig. 1.
The embodiment of the present invention may be implemented by a software program, that is, by writing a software program (and an instruction set) for implementing each step in the method for constructing a fire risk prediction model corresponding to fig. 1, the software program is stored in a storage device, and the storage device is disposed in a computer device, so that the software program can be called by a processor of the computer device to implement the purpose of the embodiment of the present invention.
EXAMPLE five
In an embodiment of the present invention, a fire risk prediction model building device is further provided, where a memory included in the fire risk prediction model building device includes a corresponding computer program product, and program instructions included in the computer program product, when executed by a computer, enable the computer to execute the fire risk prediction model building method described in the above aspects, and achieve the same technical effects.
Fig. 6 is a schematic diagram of a hardware structure of a fire risk prediction model building device as an electronic device according to an embodiment of the present invention, and as shown in fig. 6, the device includes one or more processors 610, a bus 630, and a memory 620. Taking one processor 610 as an example, the apparatus may further include: input device 640, output device 650.
The processor 610, the memory 620, the input device 640, and the output device 650 may be connected by a bus or other means, such as the bus connection in fig. 6.
The memory 620, which is a non-transitory computer readable storage medium, may be used to store non-transitory software programs, non-transitory computer executable programs, and modules. The processor 610 executes various functional applications and data processing of the electronic device, i.e., the processing method of the above-described method embodiment, by executing the non-transitory software programs, instructions and modules stored in the memory 620.
The memory 620 may include a storage program area and a storage data area, wherein the storage program area may store an operating system, an application program required for at least one function; the storage data area may store data and the like. Further, the memory 620 may include high speed random access memory, and may also include non-transitory memory, such as at least one magnetic disk storage device, flash memory device, or other non-transitory solid state storage device. In some embodiments, the memory 620 optionally includes memory located remotely from the processor 610, which may be connected to the processing device via a network. Examples of such networks include, but are not limited to, the internet, intranets, local area networks, mobile communication networks, and combinations thereof.
The input device 640 may receive input numeric or character information and generate a signal input. The output device 650 may include a display device such as a display screen.
The one or more modules are stored in the memory 620 and, when executed by the one or more processors 610, perform:
s11, generating a variable data set according to the variable item related to the fire occurrence of the target area and generating a self-variable data set; taking the fire density of the target area as a dependent variable item and generating a dependent variable data set;
s12, screening the variable items in the variable data set through feature selection to obtain preferred variable items, and generating importance degree sequence of the preferred variable items; determining a final variable term from the preferred variable terms according to the importance ranking;
s13, forming a sample data set according to the final auto-variable item and the dependent variable item;
and S14, randomly dividing a preset number of data subsets according to the sample data set, determining a training set and an internal test set from the data subsets, and generating a plurality of sub-models through training to construct a fire risk prediction model.
Preferably, in the embodiment of the present invention, the method may further include:
and S15, judging the importance ranking accuracy of each variable factor according to the simple correlation coefficient of the predicted value and the actual value in the training set and the test set.
The product can execute the method provided by the embodiment of the invention, and has corresponding functional modules and beneficial effects of the execution method. For technical details that are not described in detail in this embodiment, reference may be made to the method provided by the embodiment of the present invention.
In the embodiments provided in the present invention, it should be understood that the disclosed system, apparatus and method may be implemented in other ways. For example, the above-described apparatus embodiments are merely illustrative, and for example, the division of the units is only one logical division, and other divisions may be realized in practice, for example, a plurality of units or components may be combined or integrated into another system, or some features may be omitted, or not executed. In addition, the shown or discussed mutual coupling or direct coupling or communication connection may be an indirect coupling or communication connection through some interfaces, devices or units, and may be in an electrical, mechanical or other form.
The units described as separate parts may or may not be physically separate, and parts displayed as units may or may not be physical units, may be located in one place, or may be distributed on a plurality of network units. Some or all of the units can be selected according to actual needs to achieve the purpose of the solution of the embodiment.
In addition, functional units in the embodiments of the present invention may be integrated into one processing unit, or each unit may exist alone physically, or two or more units are integrated into one unit. The integrated unit can be realized in a form of hardware, and can also be realized in a form of a software functional unit.
The integrated unit, if implemented in the form of a software functional unit and sold or used as a stand-alone product, may be stored in a computer readable storage medium. Based on such understanding, the technical solution of the present invention may be embodied in the form of a software product, which is stored in a storage device and includes several instructions for causing a computer device (which may be a personal computer, a server, or a network device) to execute all or part of the steps of the method according to the embodiments of the present invention. And the aforementioned storage device includes: a usb disk, a removable hard disk, a Read-Only Memory (ROM), a Random Access Memory (RAM), a ReRAM, an MRAM, a PCM, a NAND Flash, a NOR Flash, a Memory, a magnetic disk, an optical disk, or other various media that can store program codes.
The above-mentioned embodiments are only used for illustrating the technical solutions of the present invention, and not for limiting the same; although the invention has been described in detail with reference to the foregoing embodiments, it will be understood by those skilled in the art that: the technical solutions described in the foregoing embodiments may still be modified, or some technical features may be equivalently replaced; and such modifications or substitutions do not depart from the spirit and scope of the corresponding technical solutions of the embodiments of the present invention.
Claims (14)
1. A fire risk prediction model construction method is characterized by comprising the following steps:
s11, generating a variable data set according to the variable item related to the fire occurrence of the target area and generating a self-variable data set; taking the fire density of the target area as a dependent variable item and generating a dependent variable data set;
s12, screening the variable items in the variable data set through feature selection to obtain preferred variable items, and generating importance degree sequence of the preferred variable items; determining a final variable term from the preferred variable terms according to the importance ranking;
s13, forming a sample data set according to the final auto-variable item and the dependent variable item;
and S14, randomly dividing a preset number of data subsets according to the sample data set, determining a training set and an internal test set from the data subsets, and generating a plurality of sub-models through training to construct a fire risk prediction model.
2. A method of constructing a fire risk prediction model according to claim 1, wherein the variable term relating to fire occurrence in the target area includes:
elevation, slope direction, normalized vegetation index, population density, road line density, railway line density, monthly average precipitation, monthly average minimum temperature, monthly average temperature and monthly average maximum temperature, and any combination thereof.
3. A method of constructing a fire risk prediction model according to claim 2, wherein the generating the autovariate data set comprises:
and carrying out 0-1 standardization processing on specific numerical values in the variable terms to generate variable data of each variable term in the independent variable data set.
4. A method as claimed in claim 3, wherein the normalizing the specific values in the variable terms by 0-1 to generate variable data of each variable term in the independent variable data set comprises:
calculating the gradient and the numerical value type gradient value of the preset area by adopting an ArcGIS 'grid surface-gradient' tool on the basis of a digital elevation model DEM with the resolution of 30M; the numerical slope value is obtained according to the slope, and is a numerical variable between-1 and 1;
calculating a normalized vegetation index based on a 500MNDVI month synthetic product;
calculating the population density of the people in the preset area by using a grid distribution algorithm based on the population density distribution diagram of the corresponding year;
calculating the road line density and the railway line density by using an ArcMap line density tool on the basis of a road network and a railway network;
and calculating climate variables such as the monthly mean highest temperature, the monthly mean lowest temperature, the monthly mean precipitation and the like based on the monthly mean highest temperature, the monthly mean lowest temperature data set and the ground precipitation monthly data set which are observed by original remote sensing.
5. A fire risk prediction model building method according to claim 4, characterized in that the numerical slope value is obtained from the slope, comprising calculating the numerical slope value according to formula (1);
aspect ═ cos (θ × 2 × pi/360), formula (1); wherein, theta is a gradient value, and Aspect is a numerical gradient value.
6. The method for constructing a fire risk prediction model according to claim 1, wherein the taking the fire density of the target area as a dependent variable term and generating a dependent variable data set comprises:
a factor data set is generated by performing 0-1 standardization processing on the fire occurrence frequency in each square kilometer of each year in an industrial park.
7. The method for constructing a fire risk prediction model according to claim 1, wherein the screening variable terms in the variable data set through feature selection to obtain preferred variable terms and generating the importance ranking of the preferred variable terms comprises:
screening the variable items in the variable data set according to a Root Mean Square Error (RMSE) to obtain the preferred variable items;
and carrying out model training through Rstudio software to obtain the importance degree sequence of the preferred variable items.
8. The method according to claim 1, wherein the randomly dividing a preset number of data subsets according to the sample data set, determining a training set and an internal test set from the data subsets, and generating a plurality of sub-models through training to construct the fire risk prediction model comprises:
randomly dividing the sample data set into five data subsets; and taking four data subsets as a training set and the other data subset as a test set, and generating five sub-models through training according to a reference cross validation mode to construct a fire risk prediction model.
9. A fire risk prediction model construction method according to claim 8, further comprising:
and judging the accuracy of the importance ranking of each variable factor according to the simple correlation coefficient of the predicted value and the actual value in the training set and the test set.
10. A fire risk prediction model construction method according to claim 2, wherein the variable term related to fire occurrence of the target area further includes a fire prevention and control index;
the fire prevention and control index is generated according to the level of fire prevention measures of the target area.
11. A fire risk prediction model construction device is characterized by comprising:
a data acquisition unit for generating a variable data set from a variable data set composed of variable items related to the occurrence of a fire in a target area and generating a self-variable data set; taking the fire density of the target area as a dependent variable item and generating a dependent variable data set;
the sorting unit is used for screening the variable items in the variable data set through feature selection to obtain preferred variable items and generating importance degree sorting of the preferred variable items; determining a final variable term from the preferred variable terms according to the importance ranking;
a sample generating unit, configured to form a sample data set according to the final auto-variable term and the dependent variable term;
and the model generation unit is used for randomly dividing a preset number of data subsets according to the sample data set, determining a training set and an internal test set from the data subsets, and generating a plurality of sub-models through training to construct a fire risk prediction model.
12. A fire risk prediction model construction apparatus according to claim 11, further comprising:
and the sequencing correction unit is used for judging the accuracy of the importance sequencing of each variable factor according to the simple correlation coefficient of the predicted value and the actual value in the training set and the test set.
13. A memory comprising a software program adapted to be executed by a processor for performing the steps of the method of constructing a fire risk prediction model according to any one of claims 1 to 10.
14. A fire risk prediction model building apparatus comprising a bus, a processor and a memory as claimed in claim 13;
the bus is used for connecting the memory and the processor;
the processor is configured to execute a set of instructions in the memory.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010328197.7A CN113553754A (en) | 2020-04-23 | 2020-04-23 | Memory, fire risk prediction model construction method, system and device |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010328197.7A CN113553754A (en) | 2020-04-23 | 2020-04-23 | Memory, fire risk prediction model construction method, system and device |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN113553754A true CN113553754A (en) | 2021-10-26 |
Family
ID=78101058
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202010328197.7A Pending CN113553754A (en) | 2020-04-23 | 2020-04-23 | Memory, fire risk prediction model construction method, system and device |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113553754A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116824785A (en) * | 2023-05-16 | 2023-09-29 | 中国消防救援学院 | Early warning method and system for forest fire hazard degree |
Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103886386A (en) * | 2014-02-20 | 2014-06-25 | 中国林业科学研究院森林生态环境与保护研究所 | Method for predicting manual fire day occurrence probability based on space grid |
| WO2015175150A1 (en) * | 2014-05-13 | 2015-11-19 | Wildfire Defense Systems, Inc. | System and method for collecting and assessing wildfire hazard data and generating wildfire risk valuations and mitigation recommendations |
| US20170161614A1 (en) * | 2015-12-07 | 2017-06-08 | Rapidsos, Inc. | Systems and methods for predicting emergency situations |
| CN107331132A (en) * | 2017-08-04 | 2017-11-07 | 深圳航天智慧城市系统技术研究院有限公司 | A kind of method and system of Urban Fires hidden danger dynamic prediction monitoring |
| CN109242361A (en) * | 2018-10-31 | 2019-01-18 | 深圳市中电数通智慧安全科技股份有限公司 | A kind of fire-fighting methods of risk assessment, device and terminal device |
| CN109242223A (en) * | 2018-11-26 | 2019-01-18 | 武汉理工光科股份有限公司 | The quantum support vector machines of city Public Buildings Fire Risk is assessed and prediction technique |
| JP2019128904A (en) * | 2018-01-26 | 2019-08-01 | Ｎｅｃソリューションイノベータ株式会社 | Prediction system, simulation system, method and program |
| CN110866624A (en) * | 2018-08-27 | 2020-03-06 | 中国石油化工股份有限公司 | Chemical accident prediction method and system |
-
2020
- 2020-04-23 CN CN202010328197.7A patent/CN113553754A/en active Pending
Patent Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103886386A (en) * | 2014-02-20 | 2014-06-25 | 中国林业科学研究院森林生态环境与保护研究所 | Method for predicting manual fire day occurrence probability based on space grid |
| WO2015175150A1 (en) * | 2014-05-13 | 2015-11-19 | Wildfire Defense Systems, Inc. | System and method for collecting and assessing wildfire hazard data and generating wildfire risk valuations and mitigation recommendations |
| US20170161614A1 (en) * | 2015-12-07 | 2017-06-08 | Rapidsos, Inc. | Systems and methods for predicting emergency situations |
| CN107331132A (en) * | 2017-08-04 | 2017-11-07 | 深圳航天智慧城市系统技术研究院有限公司 | A kind of method and system of Urban Fires hidden danger dynamic prediction monitoring |
| JP2019128904A (en) * | 2018-01-26 | 2019-08-01 | Ｎｅｃソリューションイノベータ株式会社 | Prediction system, simulation system, method and program |
| CN110866624A (en) * | 2018-08-27 | 2020-03-06 | 中国石油化工股份有限公司 | Chemical accident prediction method and system |
| CN109242361A (en) * | 2018-10-31 | 2019-01-18 | 深圳市中电数通智慧安全科技股份有限公司 | A kind of fire-fighting methods of risk assessment, device and terminal device |
| CN109242223A (en) * | 2018-11-26 | 2019-01-18 | 武汉理工光科股份有限公司 | The quantum support vector machines of city Public Buildings Fire Risk is assessed and prediction technique |
Non-Patent Citations (2)
| Title |
|---|
| 梁慧玲;林玉蕊;杨光;苏漳文;王文辉;郭福涛;: "基于气象因子的随机森林算法在塔河地区林火预测中的应用", 林业科学, no. 01, pages 89 - 98 * |
| 石雪怀等: "基于深度集成模型的火灾事故预警方法", 北京林业大学学报, vol. 39, no. 1, pages 8 - 12 * |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116824785A (en) * | 2023-05-16 | 2023-09-29 | 中国消防救援学院 | Early warning method and system for forest fire hazard degree |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Wang et al. | Seismic fragility analysis with artificial neural networks: Application to nuclear power plant equipment | |
| Ghil et al. | Extreme events: dynamics, statistics and prediction | |
| CN108399248A (en) | A kind of time series data prediction technique, device and equipment | |
| CN112258093A (en) | Risk level data processing method and device, storage medium and electronic equipment | |
| CN112070357A (en) | Radar radiation source threat assessment method based on improved BP neural network | |
| CN111079999A (en) | Flood disaster susceptibility prediction method based on CNN and SVM | |
| CN112700324A (en) | User loan default prediction method based on combination of Catboost and restricted Boltzmann machine | |
| CN110706213A (en) | Bridge cluster structure damage judgment method based on strain response cumulative distribution function difference | |
| CN114266289A (en) | Complex equipment health state assessment method | |
| CN114509162A (en) | Sound environment data monitoring method and system | |
| CN116341901A (en) | Integrated evaluation method for landslide surface domain-monomer hazard early warning | |
| CN113553754A (en) | Memory, fire risk prediction model construction method, system and device | |
| DE112014006799T5 (en) | Method for measuring a user behavior consistency level based on a complex correspondence system | |
| CN117272145A (en) | Health state evaluation method and device of switch machine and electronic equipment | |
| CN114911788B (en) | Data interpolation method and device and storage medium | |
| CN112034139A (en) | Method and device for judging rock burst tendency grade and electronic equipment | |
| CN114546841B (en) | Software quality assessment method based on cloud computing | |
| CN115130887B (en) | Reservoir dam environmental impact evaluation method and device, electronic equipment and storage medium | |
| CN115983329A (en) | Method, device, equipment and storage medium for predicting air quality and meteorological conditions | |
| US10878329B1 (en) | Probabilistic accumulation approach to assess primary uncertainty in catastrophe models | |
| CN111368837A (en) | Image quality evaluation method and device, electronic equipment and storage medium | |
| CN109933921B (en) | Rolling rock disaster risk assessment method, device and system and storage medium | |
| CN113902187A (en) | Time-of-use electricity price prediction method and device and terminal equipment | |
| CN105528657A (en) | Building earthquake damage prediction method based on Beidou and vector machine | |
| Olmedo et al. | Validation of Soft Maps Produced by a Land Use Cover Change Model |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |