JP6006902B1 - Fire simulation apparatus, fire simulation method, and fire simulation program - Google Patents

Abstract

A fire point suitable for simulation is accurately estimated. A fire simulation apparatus according to the present invention uses, as input variables, fire point information including position information, topographic information, and meteorological information about a fire point, and information indicating the scale of a fire at an evaluation target point as an output variable. The simulation unit 22 that is a mathematical model and a plurality of candidates for the firing point are received, the firing point information is received for each of the received plurality of firing point candidates, and the received firing point information is input to the simulation unit 22 and output. A fire point candidate determination unit 21 that receives a variable from the simulation unit 22 and orders a plurality of received fire point candidates according to the degree of influence of the input fire point information on the received output variable. And [Selection] Figure 5

Description

  The present invention relates to a fire simulation apparatus, a fire simulation method, and a fire simulation program.

  In Japan, although urbanization has progressed in recent years, the ratio of the area covered by forest to the total land area is still high. And once a fire breaks out at a certain point in the forest, the fire spreads from the starting point and not only burns down the forest, but also reaches the city area, causing enormous human and economic losses. The cause of a fire in a forest can be an artificial carelessness such as a misfire of a cigarette on a forest road, or it can be a natural phenomenon such as a lightning strike, a volcanic eruption, or a spark from a rock collapse. The scale of forest fires depends greatly on natural conditions such as weather conditions, topography, and vegetation. Therefore, in particular, companies having facilities such as buildings in the vicinity of forests often simulate many cases of forest fires for the purpose of drafting a fire prevention plan against forest fires.

The fire detection system of Patent Literature 1 divides a monitoring target area into a plurality of areas, and causes an infrared camera to monitor each area to detect a fire. When the infrared camera detects a fire, the fire detection system causes the visible camera to image the area. On the other hand, the fire detection system obtains the wind direction and wind speed of the fire point from a weather satellite or the like, predicts the fire spread state after a predetermined time, and displays the predicted fire spread state on the screen together with a visible image (FIG. 2 of Patent Document 1). FIG. 3).
The fire management system of Patent Document 2 shares the detection of fire (imaging), fire fighting (dropping of water, chemicals, etc.), and confirmation of fire suppression for manned and unmanned aircraft flying over the vast forest. (FIGS. 5 to 8 of Patent Document 2).

JP 2013-196655 A JP 2012-227920 A

  When simulating the impact of a forest fire on a facility, first, map information (terrain, land use, vegetation, etc.) around the facility is prepared. Second, a certain point in the forest near the facility is set as a fire point. Third, weather information (temperature, humidity, wind direction, wind speed, etc.) at the fire point is set. Fourth, map information, fire point, and weather information are input to the fire simulation application. Then, the application calculates the time until the fire reaches the facility, the fire line intensity at the time when the fire reaches, and the like.

  However, among the input values for the application, the position of the fire point is the simplest, but in practice it is the most difficult to select. The map information is originally fixed, and the weather information has a realistic value for the region from past experience. However, the fire point is not limited to a point along the forest road, and can be set to the entire range covered by the forest near the facility, assuming the case of spontaneous combustion.

  For example, suppose that five candidate fire points are set based on personal experience and knowledge. Among the five fire point candidates, there may be included those that have little effect on the facility (the weather information has an advantageous effect and the time to reach the facility is long). On the other hand, in addition to the five candidates for fire point, it has a great impact on the facility. (When clouds are trapped in the sky, the fire spreads along the valley. Something that reaches (hidden) may be hidden. Even in such a case, it is possible to perform a simulation by slightly changing the weather information. However, in the first place, it is not meaningful to perform a simulation assuming such five candidates for fire points, and the disaster prevention plan created as a result of such a simulation is meaningless.

The fire detection system of Patent Document 1 and the fire management system of Patent Document 2 are useful for detecting a fire that actually occurs. However, Patent Literature 1 and Patent Literature 2 do not have a viewpoint of stocking candidates for a fire point suitable for the simulation in order to simulate a forest fire that may occur in the future. Therefore, a separate measure is required to estimate the fire point suitable for simulation.
Therefore, an object of the present invention is to accurately estimate a fire point suitable for simulation.

The fire simulation apparatus of the present invention is a simulation that is a mathematical model that uses as input variables fire point information including position information, topographic information, and meteorological information about the fire point, and uses as an output variable information indicating the scale of the fire at the evaluation target point And receiving a plurality of ignition point candidates, receiving the ignition point information for each of the received plurality of ignition point candidates, inputting the received ignition point information to the simulation unit, receiving the output variable from the simulation unit, A fire point candidate determining unit that orders a plurality of received fire point candidates according to the degree of influence of the input fire point information on the received output variable.
Other means will be described in an embodiment for carrying out the invention described later.

  According to the present invention, it is possible to accurately estimate a fire point suitable for simulation.

It is a figure (case 1, case 2) explaining expansion of a forest fire. It is a figure (case 3, case 4) explaining expansion of a forest fire. It is a figure (case 5, case 6) explaining expansion of a forest fire. It is a figure (case 7) explaining expansion of a forest fire. It is a figure explaining the structure of a fire simulation apparatus. It is a figure which shows an example of correlation information. It is a flowchart of a processing procedure.

  Hereinafter, a mode for carrying out the present invention (referred to as “the present embodiment”) will be described in detail with reference to the drawings. First, in order to facilitate understanding of the present embodiment, an actual forest fire will be described. However, the fire simulation apparatus of the present invention is not limited to the forest area, but can be applied to other areas such as a pasture area, a rural area, a factory area, and an urban area.

(Actual forest fire)
The expansion of forest fire will be described with reference to FIGS. 1, 2, 3 and 4. FIG. Cases 1 to 6 in FIGS. 1, 2, and 3 are diagrams showing to which point the fire that occurred at the fire point in the center of the figure expands 1 hour and 2 hours after the fire occurred. is there. In all cases 1 to 6, the land is divided into four regions, northeast, southeast, southwest, and northwest, by two straight lines with the fire point as an intersection. In all cases 1 to 6, the evaluation target point is located in the southeast (lower right) of the fire point. The evaluation target is, for example, the location of a facility used by a company or the like that makes a fire prevention plan.

  In case 1 of FIG. 1, there is no difference in elevation of land in all areas, and the area is flat, and the same kind of forest spreads. In all areas, there is no wind. The circumference of the circle 51 indicates a fire front (hereinafter also referred to as “fire line”) one hour after the occurrence of the fire. The circumference of the circle 52 indicates a fire line 2 hours after the occurrence of the fire. The evaluation target point is on the circumference of the circle 52. The circle 51 and the circle 52 are concentric circles centered on the fire point.

  In case 2 of FIG. 1, in all areas, there is no difference in elevation of land and it is flat, and the same kind of forest spreads. And wind is blowing from the northwest direction in all areas. The circumference of the ellipse 61 is a fire line one hour after the occurrence of the fire. The evaluation target point is inside the ellipse 61. The circumference of the ellipse 62 is a fire line 2 hours after the occurrence of the fire. The ellipse 61 and the ellipse 62 are significantly biased and spread in the southeast direction from the other directions as seen from the fire point.

  In case 3 of FIG. 2, in all areas, there is no difference in elevation of land and it is flat, and the same kind of forest spreads. In all areas, there is no wind. However, it rains in most of the northeast and northwest areas (about 3/4 from the north), and the humidity is relatively high. Other places are clear and the humidity is relatively low. The circumference of the bowl-shaped figure 71 is a fire line 1 hour after the occurrence of the fire. The circumference of the bowl-shaped figure 72 is a fire line 2 hours after the occurrence of the fire. The evaluation target point is on the circumference of the graphic 72.

  In case 4 of FIG. 2, in all areas, there is no difference in elevation of the land, it is flat, and there is no wind. And there is a factory almost halfway between the fire point and the evaluation target point. The factory stores a large amount of flammable dangerous goods. When the dangerous goods are ignited, the fire spreads at a speed that is not comparable to the natural movement speed of the fire line. The circumference of the bowl-shaped figure 81 is a fire line 1 hour after the occurrence of the fire. The evaluation target point is inside the figure 81. The circumference of the bowl-shaped figure 82 is a fire line 2 hours after the occurrence of the fire.

  In case 5 of FIG. 3, in all areas, there is no difference in elevation of land and it is flat and no wind. However, the forests in the northeast and southwest regions are of a different type than the forests in the northwest and southeast regions. The forest in the northwest and southeast regions (vegetation B) has a higher plant density and contains more oil and fat than the forest in the northeast and southwest regions (vegetation A). The circumference of the ellipse 91 is a fire line one hour after the occurrence of the fire. The evaluation target point is on the circumference of the ellipse 91. The circumference of the ellipse 92 is a fire line 2 hours after the occurrence of the fire. The ellipse 91 and the ellipse 92 are centered on the fire point, but as viewed from the fire point, the ellipse 91 and the ellipse 92 are spread more significantly in the southeast and northwest directions than the other directions.

  In case 6 of FIG. 3, there is no wind in all regions. However, the fire point is located in a valley extending from northwest to southeast. This valley is sandwiched between a mountain with a summit in the northeast and a mountain with a summit in the southwest. And although it cannot be confirmed in the figure, clouds spread in the space above a certain altitude line in all areas. The circumference of the ellipse 101 is a fire line one hour after the occurrence of the fire. The evaluation target point is inside the ellipse 101. The ellipse 101 is centered around the fire point, but as viewed from the fire point, the ellipse 101 spreads more significantly in the southeast and northwest directions (along the valley) than the other directions. The circumference of the figure 102 is a fire line 2 hours after the occurrence of the fire. The figure 102 has a shape along the shape of the valley, which is no longer an ellipse. By the way, if there are no clouds in the sky, the fire line runs up the mountain slope. Therefore, in this case, the ellipse 101 and the graphic 102 spread significantly biased in the northeast and southwest directions.

  In case 7 of FIG. 4, there is a sea C in the southwest, and mountains E, F, G, and H approach the sea. River A and river B are pouring toward the sea C, and each river forms a valley. The evaluation target point is located at a location near the mouth of River A along the coast. In the land shown in the figure, forests are sparse overall. However, the ridgeline formed by the mountain G and the mountain H located between the valley of the river A and the valley of the river B has a portion (a saddle point) whose altitude is slightly lower than the surrounding area. Vegetation J spreads locally so as to straddle two valleys around the saddle point. Vegetation J has a high density and a high fat content. Island D is floating in the sea C.

Now, assuming that the user has fired from each of the candidates 1 to 5 using the existing technology, the influence on the evaluation target point is examined. At this time, the user can obtain the following temporary simulation result.
(1) In the case of a fire from candidate 1 of the fire point, it is only necessary to pay attention to the north wind (wind blowing from the north).
(2) In the case of a fire from Candidate 2 of the fire point, attention should be paid only to the fire spread along the valley of River A.
(3) There is no need to pay particular attention when a fire breaks out from candidate 3 of the fire point. This is because the mountain G and the mountain H between the river A and the river B block the spread of fire.
(4) In the case of a fire from the fire point candidate 4, there is no need to pay particular attention. Similarly, the mountain G and the mountain H between the river A and the river B block the spread of fire.
(5) In the case of a fire from the fire point candidate 5, no special attention is required. This is because the sea C blocks the spread of fire.

  What should be truly noted here is whether (3) and (4) are unconditionally correct. For example, a fire may spread to vegetation J if a fire breaks out from candidate fire point 3 and an east wind blows and the cloud position is low. The fire then goes over the saddle point and reaches the valley of river A. Furthermore, it is quite possible that the fire crosses river A and reaches the evaluation target point. In addition, even if a fire starts from the candidate 4 of the fire point, the south wind blows, and eventually changes to the east wind, the fire can sufficiently reach the evaluation target point through the vegetation J. It is difficult to find a candidate 3 for a fire point and a candidate 4 for a fire point as fire points that a user with little experience should be aware of.

If the simulation is repeated by changing the vegetation and weather information in various ways after fixing the fire point and the evaluation target point, it becomes clear that the candidates for the fire point are classified into two types.
The first type is a type in which the influence on the evaluation target point does not change much even if weather information or the like is changed. The fire point candidate 5 belongs to the first type. This is because the fire point candidate 5 is surrounded by the sea C, and no matter how the wind direction, wind speed, temperature, humidity, forest type, etc. change, it will not eventually spread to the evaluation target point. . The fire point candidate 1 also belongs to the first type. This is because the fire point candidate 1 is sufficiently close to the evaluation target point, and no matter how the weather conditions change, it spreads to the evaluation target point in the same time after the fire.

  The second type is a type in which the influence on the evaluation target point varies greatly depending on changes in weather information and the like. As apparent from the above, the candidate 3 for the fire point and the candidate 4 for the fire point belong to the second type. This is because, depending on the wind direction, it may not be necessary to pay attention for the time being, and conversely, special attention may be required. The fire point candidate 2 is subtle and can enter either type.

(Configuration of fire simulation equipment)
The structure of the fire simulation apparatus 1 is demonstrated along FIG. The fire simulation apparatus 1 is a general computer. The fire simulation device 1 includes a central control device 11, an input device 12, an output device 13, a main storage device 14, and an auxiliary storage device 15. These are connected by a bus. The auxiliary storage device 15 stores correlation information 31 (details will be described later). The fire point candidate determination unit 21 and the simulation unit 22 in the main storage device 14 are programs. In the following description, when the operation subject is described as “XX part”, it means that the central control unit 11 reads the XX part from the auxiliary storage device 15 and loads it into the main storage device 14 and then loads the XX part. It means to realize the function (detailed later).

  Many forest fire simulation applications are currently in widespread use. For example, FARSITE (Fire Area Simulator), which is an application developed by the US Department of Agriculture, is one example. The simulation unit 22 (see FIG. 5) of the present embodiment is also a kind of application as described above. The simulation unit 22 includes a mathematical model that is a collection of many mathematical expressions. It can be said that a mathematical model is a function that outputs an output variable when an input variable is input.

(Input variable)
The input variables of the mathematical model include wind direction, wind speed, temperature, humidity, vegetation, altitude, tilt direction, tilt angle, difference between wind direction and tilt direction, etc. (The meaning of each input variable will be described later). The simulation unit 22 reads a topographic map, a vegetation map, or a land use map published by the public office, thereby vegetation, altitude, and slope for each point on the map (more precisely, each area of the mesh having a small area). Set the direction and tilt angle. The simulation unit 22 accepts that the user selects a specific area of the map on the screen and inputs a numerical value with a keyboard or the like, and the wind direction for each point on the map including the fire point position and the evaluation target point position, Set wind speed, temperature, humidity, etc.

  In addition to the above, the input variables of the mathematical model include the fire point position and the evaluation target point position. The simulation unit 22 accepts that the user selects a certain point on the map on the screen, and sets the fire point position and the evaluation target point position on the map.

(Output variable)
Examples of output variables of the mathematical model include information on the scale of fire (variables), such as fire line intensity, arrival time, and the like. The strength of the fire wire is a value obtained by dividing the energy generated by the fire by the length of the fire wire. That is, the fire wire intensity is logically the energy amount of fire occurring in a square area whose perimeter is a unit length (for example, 1 m). The caustic intensity is defined for each time point and for each point (a square with a unit length around it). Of the output variables, the most important is the fire line intensity at the evaluation target point when the fire line reaches the evaluation target point. The user determines, for example, the width for cutting the forest around the facility based on the fire line intensity, and determines the thickness of the outer wall of the storage facility. The arrival time is the time until the fire line reaches the evaluation target point after the fire has occurred. The user makes an evacuation plan or the like that can be executed within the range of the arrival time.

Now, it is assumed that the fire line intensity among the output variables is y ₁ and the arrival time is y ₂ . Of the input variables, the fire point position and _{x 1,} the evaluation point position and _{x 2,} the other input _{variables, x 3, x 4, x} 5, and,. Then, the mathematical model of the simulation unit 22 can be simplified and expressed in the following implicit function format.
F (y ₁ , y ₂ , x ₁ , x ₂ , x ₃ , x ₄ , x ₅ ,...) = 0
As a matter of course, the implicit function expresses the relationship between a plurality of variables by discarding the relationship between the input and the output. The word “input variable” here means that the variable is generally a given variable in many cases, and the term “output variable” may be that the variable is generally unknown. It just means a lot. In other words, what F = 0 says is that any _one of y ₁ , y ₂ , x ₁ , x ₂ , x ₃ , x ₄ , x ₅ ,. If all else is given, one is specified (the equation can be solved).

For example, the simulation unit 22 of the present embodiment considers variables other than y ₁ and x ₃ as constants, changes the value of x ₃ to various values, and determines the degree of influence on the value of y _1. Investigate. For example, it is examined how much the fire line intensity y _{1 at a} specific evaluation target point changes when the wind speed x ₃ changes. Obtaining this influence degree mathematically includes obtaining the slope of the partial derivative δy ₁ / δx ₃ or the correlation coefficient between y ₁ and x ₃ . Similarly, the simulation unit 22 obtains the slope of δy ₁ / δx ₄ or the correlation coefficient between y ₁ and x ₄ (hereinafter, the same applies to x ₅ ,...). The simulation unit 22 summarizes these slopes or correlation coefficients into one value, and evaluates the collected value (details will be described later).

(Correlation information)
The correlation information 31 will be described with reference to FIG. In the correlation information 31, in association with the fire point ID stored in the fire point ID column 201, the position column 202 has the fire point position, the wind direction column 203 has the wind direction, the wind speed column 204 has the wind speed, and the temperature. The column 205 is the temperature, the humidity column 206 is the humidity, the vegetation column 207 is the vegetation, the altitude column 208 is the altitude, the tilt direction column 209 is the tilt direction, and the tilt angle column 210 is the tilt angle. The difference between the wind direction and the inclination direction is stored in the difference column 211 between the wind direction and the inclination direction. These are collectively called “fire point information”.

  Further, in association with the fire point ID stored in the fire point ID column 201, the correlation coefficient is in the correlation coefficient column 212 and the effect coefficient is in the effect coefficient column 213 on the right side of the double line of the correlation information 31. Coefficients are stored.

The fire point ID in the fire point ID column 201 is an identifier that uniquely identifies a fire point candidate. Hereinafter, unless there is a misunderstanding, a candidate for a fire point may be simply referred to as a “fire point”.
The fire point position in the position column 202 is the position (latitude and longitude) of the fire point. Note that “# 1, ♭ 1”, “# 2, ♭ 2”, and “# 3, ♭ 3” shown in FIG. 6 are simple examples of latitude and longitude values. Among the variables of the implicit function F, two components of the fire point position “x ₁ ” which is a vector.
The wind direction in the wind direction column 203 is the wind direction (upstream direction) at the fire point at the time of the fire, and is a numerical value of 0 to 360 degrees indicating the direction of east, west, south, and north.
The wind speed in the wind speed column 204 is the wind speed at the fire point at the time of fire.
The temperature in the temperature column 205 is the temperature at the fire point at the time of the fire.
The humidity in the humidity column 206 is the humidity at the fire point at the time of the fire.

The vegetation in the vegetation column 207 is an index indicating the ease of ignition and the magnitude of energy when burning, such as the density of the plant of the forest at the fire point, the component ratio of fats and oils, and the like. Consider the interval between buildings in the city area and the material of the buildings in the broad vegetation. In FIG. 5, character strings such as “grassland” and “matsubayashi” are described for ease of understanding, but it is assumed that an index (numerical value) is associated with these character strings.
The altitude of the altitude column 208 is the altitude (elevation above sea level) of the fire point.
The inclination direction of the inclination direction column 209 is a numerical value of 0 to 360 degrees indicating the direction in which the sphere moves according to gravity when the sphere is placed at the fire point.
The inclination angle in the inclination angle column 210 is an angle formed by the horizontal plane at the fire point and the locus of movement of the sphere.
The difference between the wind direction and the tilt direction in the difference column 211 between the wind direction and the tilt direction is a value obtained by dividing the tilt direction from the wind direction.

The correlation coefficient column 212 includes small columns 212a to 212i for each of the fire point information (wind direction, wind speed,..., Difference between the wind direction and the tilt direction). The wind direction column 212a stores a correlation coefficient between the fire line intensity and the wind direction at the evaluation target point. The wind speed column 212b stores a correlation coefficient between the fire line intensity and the wind speed at the evaluation target point. The same applies to the temperature column 212c to the difference column 212i between the wind direction and the inclination direction.
The effect coefficient in the effect coefficient column 213 is an absolute value of the product of nine types of correlation coefficients stored in each of the small columns 212a to 212i.

  As described above, these pieces of information called “fire point information” are not limited to fire points, and can be set for all points on the map. However, for simplification of explanation, the correlation information 31 here representatively describes only information on the fire point. Furthermore, as in the case where the clouds are located in an altitude range of 300 m to 1000 m, the humidity is originally set for each altitude above each point (the wind direction, wind speed, and temperature are the same). However, for simplification of explanation, the correlation information 31 here assumes that the humidity and the like are the same at any altitude above the sky at a certain point.

  As described above, the correlation information 31 is the evaluation target point for each of the wind direction, the wind speed, the temperature, the humidity, the vegetation, the altitude, the tilt direction, the tilt angle, and the difference between the wind direction and the tilt direction in the fire point information. The correlation coefficient with the caustic intensity at is stored. There are nine correlation coefficients in total. Now, records (for example, records 251 to 260) regarding a certain specific fire point candidate among the fire point information records (rows) are collectively referred to as a simulation group. Each of the nine types of correlation coefficients is associated with one simulation group. The correlation information 31 also stores an effect coefficient. The effect factor is also associated with one simulation group. The number of simulation groups is equal to the number of fire point candidates.

(Outline of processing)
Now, the fire simulation apparatus 1 creates one simulation group (for example, records 251 to 260 of the correlation information 31). This simulation group has 10 records. The fire simulation apparatus 1 fixes the fire point position to a certain value. This is because, in the following, an effect coefficient is finally calculated, but this effect coefficient is defined for each fire point.
The fire simulation apparatus 1 sets the same value for the fire point information other than the wind speed in all ten records ("↑" means "same as above"). The fire simulation apparatus 1 sets a different value for each of the ten individual records as the wind speed.

  The fire simulation apparatus 1 inputs the fire point information of the record 251 to the mathematical model. As a matter of course, the position of the evaluation target point is also input. Then, the mathematical model outputs the fire ray intensity (at the evaluation target point). A coordinate plane 271 in FIG. 6 is a coordinate plane having the horizontal axis as the wind speed and the vertical axis as the fire line intensity. Then, the fire simulation apparatus 1 can dot the combination of the wind speed of the input variables and the fire ray intensity that is the output variable as a point (●) on the coordinate plane 271.

  Similarly, the fire simulation apparatus 1 inputs the fire point information of the record 252 to the mathematical model. Then, the fire simulation apparatus 1 can dot another point (●) on the coordinate plane 271. If such a process is repeated, ten dots are doted on the coordinate plane. The fire simulation apparatus 1 calculates the correlation coefficient between the wind speed and the fire line intensity based on the coordinate values of the ten points. The correlation coefficient calculated here is “0.8” (reference numeral 272).

  Next, the fire simulation device 1 resets the value of the fire point information of the simulation group (for example, the records 251 to 260 of the correlation information 31) while the fire point position is continuously fixed to a certain value. For example, the fire simulation apparatus 1 sets the same value for the flash point information other than the temperature in all ten records. The fire simulation apparatus 1 sets a different value for each of the ten individual records as the temperature. Then, similarly to the above, the fire simulation apparatus 1 can dot 10 points (●) on the coordinate plane 273 having the horizontal axis as the temperature and the vertical axis as the fire line intensity. Similarly, the fire simulation apparatus 1 calculates a correlation coefficient between the air temperature and the fire line intensity based on the coordinate values of the ten points. The correlation coefficient calculated here is “0.4” (reference numeral 274).

  Note that the fire simulation apparatus 1 may calculate the slopes of the regression lines of the two variables instead of calculating the correlation coefficient.

  When such processing is repeated, the “correlation coefficient between the fire line intensity and the wind speed at the evaluation target point”, “the correlation coefficient between the fire line intensity and the air temperature at the evaluation target point”, “the fire line intensity at the evaluation target point and The correlation coefficient of humidity ”,... Is calculated. As is well known, the larger the absolute value of the correlation coefficient, the greater the relevance of the two variables. In the example of FIG. 6, the correlation coefficient “0.8” between the caustic intensity and the wind speed has the largest absolute value on the positive side. The largest absolute value on the negative side is the correlation coefficient “−0.7” between caustic intensity and humidity. This indicates that the fire line intensity increases as the wind speed increases (the oxygen supply amount per unit time increases), and the fire line intensity decreases as the humidity increases (the fire extinguishing effect due to moisture works).

Here, the effect coefficient is an absolute value of the product of nine types of correlation coefficients. Of course, the effect coefficient may be a sum, not a product, or a value other than that. That is, the effect coefficient may be any value as long as nine types of correlation coefficients are reflected. When the effect coefficient is a product, in order to prevent the effect coefficient from becoming “0” due to the presence of “0” in the correlation coefficient, a product obtained by adding a constant (for example, “2”) to the correlation coefficient. And a value obtained by subtracting another constant (for example, 2 ⁹ = 512) from the product may be used as the effect coefficient.

  In the above, an example was described in which the wind speed (or humidity) was changed and the other fire point information was set as a constant. At this time, the fire simulation apparatus 1 may use a past observation value or a simulated value as the value to be changed. The same applies to the constants. Furthermore, the fire simulation apparatus 1 performs the above-described processing for the next simulation group after changing the fire point position. Then, the correlation information 31 stores the effect coefficient in association with each of many ignition point candidates.

  The larger the effect factor of a fire point candidate, the more suitable the fire point candidate is as a fire point candidate when a forest fire simulation is performed. For example, the user selects a predetermined number of fire point candidates in descending order of the effect coefficient, and uses them for actual simulation. The user may select a predetermined number of firing point candidates in descending order of a specific correlation coefficient instead of the effect coefficient. For example, when the change in the wind speed near the fire point is large throughout the year, it is effective to perform a simulation for comparing the fire line intensity by changing the wind speed in various ways. In this case, the user may select a predetermined number of ignition point candidates in descending order of the correlation coefficient related to the wind speed, and use it for the actual simulation.

  Furthermore, when the change in humidity is large in addition to the wind speed, it is effective to perform a simulation in which the combination of the wind speed and humidity is changed in various ways to compare the fire line intensities. In this case, the user selects a predetermined number of firing point candidates in descending order of the correlation coefficient related to the wind speed, selects a predetermined number of firing point candidates in descending order of the correlation coefficient related to humidity, and Candidates may be used for actual simulations.

  When the user uses the fire simulation apparatus 1 of the present embodiment for the first time, for example, 20 candidate fire points are randomly input. The fire simulation apparatus 1 executes the above-described processing procedure using data of public offices and the like that are available at that stage, and calculates an effect coefficient for 20 fire point candidates. The user performs an actual simulation using the “best 10” fire point candidates. For example, after a period of one year or the like, the data of the public office is updated. The fire simulation apparatus 1 recalculates the effect coefficient using the data updated for the previous 20 locations (the effect coefficient is different from the previous one). The user performs an actual simulation using the new “best 10” fire point candidates.

(Processing procedure)
A processing procedure will be described with reference to FIG.
In step S301, the fire point candidate determination unit 21 of the fire simulation apparatus 1 accepts a fire point candidate. Specifically, first, the fire point candidate determination unit 21 displays a map (not shown) including evaluation target points on the output device 13. It is assumed that such a map is stored in advance in the auxiliary storage device 15.

  Second, the fire point candidate determination unit 21 accepts that the user selects a predetermined number of fire point candidates via the input device 12. At this time, the user designates, for example, 20 points on the map. Here, it is assumed that the input fire point candidates are “P01, P02, P03,..., P20”. These “P01, P02, P03,..., P20” are directly used as the fire point ID in the correlation information 31 (FIG. 5).

  At this time, the fire point candidate determination unit 21 may automatically generate a fire point candidate without receiving input from the user. For example, the fire point candidate determination unit 21 draws a meridian passing through the evaluation target point on the map, and draws four meridians two by two at equal intervals on the east and west of the meridian. Furthermore, the fire point candidate determination unit 21 draws a latitude line passing through the evaluation target point, and draws four latitude lines at two equal intervals to the north and south of the latitude line. Then, a grid composed of 5 meridians and 5 latitudes is completed. There are a total of 25 intersections (lattice points) between the meridian and the latitude line, and the fire point candidate determining unit 21 selects the fire point candidates “P01, P02, P03,. -P24 "may be used.

  In step S302, the fire point candidate determination unit 21 receives fire point information and the like. Specifically, first, the fire point candidate determination unit 21 acquires fire point information for each of the fire point candidates from an external database operated by a public office or the like based on the fire point position. The wind direction, wind speed, temperature, and humidity acquired at this time are average values for a certain period for each of the candidates for the fire point.

  Second, the fire point candidate determination unit 21 creates a record of the correlation information 31 based on the data acquired in “first” in step S302. Twenty records equal to the number of candidates for the firing point are created. And the part from the fire point ID column 201 to the difference column 211 between the wind direction and the inclination direction of these 20 records is filled with numerical values. Each of the 20 records created at this time is called a “parent record”.

  In step S303, the fire point candidate determination unit 21 creates a simulation group. Specifically, first, the fire point candidate determination unit 21 selects any one of nine fire point information (wind direction, wind speed, temperature,..., Difference between wind direction and inclination direction). . The selected fire point information is referred to as “key fire point information”.

  Second, the fire point candidate determination unit 21 generates an arbitrary number of discrete values for the key fire point information. Assume that the key fire point information is “wind speed”. Furthermore, it is assumed that the average value of “wind speed” acquired in “first” in step S302 is “10 m / sec”. Then, the fire point candidate determination unit 21 generates a predetermined number (for example, 10) of discrete values of the wind speed including “10 m / sec”. Here, it is assumed that “2, 4,..., 10,. Of course, the user can arbitrarily set the number and interval of discrete values.

Third, the fire point candidate determination unit 21 copies the parent record by the number of created discrete values. The copied record is called “child record”. A set of child records is nothing but a simulation group.
Fourth, the fire point candidate determination unit 21 deletes the values in the key fire point information column of all child records, and replaces each of the discrete values generated in “second” in step S303 with each child record. Store it in the key firing point information column of the record.

  In step S304, the simulation unit 22 of the fire simulation apparatus 1 calculates the fire ray intensity. Specifically, the simulation unit 22 calculates the fire line intensity by the above-described method based on the fire point information of the simulation group. Then, the fire line intensity is calculated corresponding to each of the child records (only the value of the key firing point information is different from each other).

In step S305, the fire point candidate determination unit 21 of the fire simulation apparatus 1 calculates a correlation coefficient. Specifically, first, the fire point candidate determination unit 21 uses the value of the fire point information of the child record and the fire line intensity calculated in step S304 to determine the fire line intensity and the key fire point information by the method described above. The correlation coefficient between is calculated.
Second, the fire point candidate determination unit 21 stores the correlation coefficient calculated in step S304 in the small column corresponding to the key fire point information in the correlation coefficient column 212 of the correlation information 31.

  The fire point candidate determination unit 21 and the simulation unit 22 repeat the processing of steps S303 to S305 for all key fire point information (inner loop). Furthermore, when the fire point candidate determination unit 21 and the simulation unit 22 finish the process of the inner loop for one simulation group, the processes of steps S303 and S304 are repeated for all the simulation groups (outer loop). In FIG. 7, the description of the inner loop and the outer loop is omitted.

In step S306, the fire point candidate determination unit 21 calculates an effect coefficient. Specifically, the fire point candidate determination unit 21 calculates a product of nine correlation coefficients, and stores the calculation result as an effect coefficient in the effect coefficient column 213 of the correlation information 31. The fire point candidate determination unit 21 repeats the process for all simulation groups. At this stage, the fire point candidate determination unit 21 determines that “(fire point ID, effect coefficient) = (P01, 2.2 × 10 ⁻⁴ ), (P02, 2.0 × 10 ⁻⁴ ),. Information such as (P20, 2.1 × 10 ⁻⁴ ) ”is held. Note that the fire point candidate determination unit 21 has nine weights (0 ≦ α ₁ , α ₂ ,... Α ₉ ≦ 1 and α ₁ + α ₂ +... + Α for each of the nine correlation coefficients. ₉ = 1) and the product of the multiplication results may be used as the effect coefficient.

In step S307, the fire point candidate determination unit 21 determines a fire point candidate. Specifically, first, the fire point candidate determination unit 21 rearranges the fire point IDs in descending order of the effect coefficient.
Secondly, the fire point candidate determination unit 21 displays a predetermined number of fire point IDs from the top of the fire point IDs rearranged in “first” in step S307 on the output device 13. At this time, the fire point candidate determination unit 21 may display the fire point ID, the effect coefficient, and the rank in association with the position of the fire point candidate on the map. As described above, when the grid point is set as the candidate for the fire point in “second” in step S301, the fire point candidate determination unit 21 displays the fire point ID, the effect coefficient, and the rank for each grid point. become.
Thereafter, the processing procedure ends.

(Modification)
As a result of executing the above-described processing procedure, it is assumed that the candidate for the fire point having the maximum effect coefficient is “P01” among many fire point candidates. Of the nine correlation coefficients of P01, the largest one is assumed to be the correlation coefficient “0.8” for the wind speed. The correlation coefficient “0.8” merely indicates the overall relationship between the fire line intensity and the wind speed. For example, “if the wind speed is around a certain level, It does not indicate that the fire wire intensity changes greatly in comparison.

Now, in the implicit function described above, the fire line intensity is y ₁ and the wind speed is x ₃ . The fire point candidate determination unit 21 calculates x ₃ such that the absolute value of the values of the partial derivatives δy ₁ / δx ₃ is maximized. Before and after such x _3, y ₁ is significantly changed. In order to calculate such x ₃ , the fire point candidate determination unit 21 uses a regression analysis method for a curve that nonlinearly approximates the arrangement of a plurality of dots in the graph 271 of FIG. 6 showing the relationship between the fire line intensity and the wind speed. To identify. Then, the derivative of the specified curve is obtained. The absolute value of the derivative x ₃ to maximize the a a a a "10 m / sec." Then, the fire point candidate determination unit 21 fixes the remaining wind speed to “10 m / sec” among the nine fire point information for a given evaluation target point and a given fire point candidate. It is also possible to calculate the fire line intensity by changing the eight fire point information in various ways (first simulation method).

Further, fire point information other than wind speed (wind direction, temperature, humidity,...) Is assumed to be x ₄ , x ₅ , x ₆ ,. The fire point candidate determination unit 21 calculates x ₄ such that the absolute value of the partial derivative δy ₁ / δx ₄ is maximized. Similarly, the fire point candidate determination unit 21 calculates x ₅ such that the absolute value of the partial derivative δy ₁ / δx ₅ is maximized. Similarly, the fire point candidate determination unit 21 calculates x ₆ ,... The fire point candidate determination unit 21 fixes the values of x ₄ , x ₅ , x ₆ ,... To the values calculated in this way for a given evaluation target point and a given fire point candidate. and, by variously changing the wind velocity x _3, it is also possible to calculate the caustic strength (second simulation method).

  The fire point candidate determination unit 21 may calculate the effect coefficient for each fire point candidate using the first simulation method or the second simulation method described above. In the above description, the example of fixing the firing point information by focusing on “wind speed” and the example of fixing all the other firing point information have been described. As a matter of course, the user can arbitrarily set the fire point information to be noticed (fixed). There may be a plurality of noticeable fire point information (for example, temperature and humidity).

(Effect of this embodiment)
The fire simulation device of the present embodiment has the following effects.
(1) Without relying on empirical knowledge, fire point candidates that are truly suitable for a fire simulation can be determined in order of their degree of suitability from many fire point candidates.
(2) Among many pieces of fire point information, a fire point candidate can be determined by paying attention to the influence of a specific variable on the fire scale, such as wind speed.
(3) Since the existing technique for calculating the correlation coefficient between two variables can be used, it is possible to maintain the reliability when calculating the influence of a specific variable on the fire scale.

(4) Since only the value of one fire point information among a plurality of fire point information is discretely changed, the calculation of a significant correlation coefficient is ensured.
(5) Since the candidates for the fire point are displayed on the map, it is clear which point should be used for the simulation.
(6) If the flash point information is updated, it is reliably reflected in the processing result.
(7) Easily available data such as vegetation, altitude, wind speed, wind direction, temperature and humidity can be used as they are.

  In addition, this invention is not limited to an above-described Example, Various modifications are included. For example, the above-described embodiments have been described in detail for easy understanding of the present invention, and are not necessarily limited to those having all the configurations described. Further, a part of the configuration of one embodiment can be replaced with the configuration of another embodiment, and the configuration of another embodiment can be added to the configuration of one embodiment. Further, it is possible to add, delete, and replace other configurations for a part of the configuration of each embodiment.

Each of the above-described configurations, functions, processing units, processing means, and the like may be realized by hardware by designing a part or all of them with, for example, an integrated circuit. Each of the above-described configurations, functions, and the like may be realized by software by interpreting and executing a program that realizes each function by the processor. Information such as programs, tables, and files that realize each function can be stored in a recording device such as a memory, a hard disk, or an SSD (Solid State Drive), or a recording medium such as an IC card, an SD card, or a DVD.
In addition, the control lines and information lines are those that are considered necessary for the explanation, and not all the control lines and information lines on the product are necessarily shown. In practice, it may be considered that almost all the components are connected to each other.

1 Fire Simulation Device 11 Central Control Device 12 Input Device 13 Output Device 14 Main Storage Device 15 Auxiliary Storage Device 21 Fire Point Candidate Determining Unit 22 Simulation Unit 31 Correlation Information

Claims (9)

A simulation unit that is a mathematical model that uses as input variables fire point information including position information, topographic information, and weather information about the fire point, and information that indicates the scale of the fire at the evaluation target point;
Accept a plurality of candidates for the fire point,
For each of the plurality of received fire point candidates, receive the fire point information, input the received fire point information to the simulation unit, receive the output variable from the simulation unit,
A fire point candidate determination unit that orders the plurality of received fire point candidates according to the degree of influence of the input fire point information on the received output variable;
A fire simulation apparatus comprising: The degree of influence is
The degree of influence that one or more variables included in the flash point information has on a specific variable among the output variables;
The fire simulation apparatus according to claim 1. The degree of influence that one variable included in the flash point information has on a specific one of the output variables is:
A correlation coefficient between one variable included in the flash point information and a specific one of the output variables;
The fire simulation device according to claim 2. The fire point candidate determination unit
Generating a plurality of discrete values for one variable included in the flash point information;
Inputting the plurality of generated values to the simulation unit;
The fire simulation apparatus according to claim 3. The fire point candidate determination unit
Displaying the ordered positions of the fire point candidates along with the degree of influence on a map;
The fire simulation apparatus according to claim 4. The fire point candidate determination unit
Each time the terrain information or the weather information is updated,
Inputting the updated fire point information into the simulation unit;
The fire simulation apparatus according to claim 5. The terrain information is
Including vegetation and elevation,
The weather information is
Including wind direction, wind speed, temperature and humidity,
The fire simulation apparatus according to claim 6. The simulation part of the fire simulation device
Execute the process as a mathematical model with the fire point information including the position information about the fire point, terrain information and weather information as input variables, and the information indicating the scale of the fire at the evaluation target points as output variables,
The fire point candidate determination unit of the fire simulation device is:
Accept a plurality of candidates for the fire point,
For each of the plurality of received fire point candidates, receive the fire point information, input the received fire point information to the simulation unit, receive the output variable from the simulation unit,
Ordering the plurality of accepted fire point candidates according to the degree of influence of the input fire point information on the received output variable;
A fire simulation method for a fire simulation apparatus. For the simulation part of the fire simulation device,
Execute the process as a mathematical model with the location information, the topographic information and the weather information about the fire point as the input variable, and the information indicating the scale of the fire at the evaluation target point as the output variable.
For the fire point candidate determination unit of the fire simulation device,
Accept a plurality of candidates for the fire point,
For each of the plurality of received fire point candidates, receive the fire point information, input the received fire point information to the simulation unit, receive the output variable from the simulation unit,
According to the degree of influence of the input fire point information on the received output variable, to execute a process of ordering the received plurality of fire point candidates,
A fire simulation program for causing a fire simulation device to function.

Images