JP7473079B2 - Evacuation shelter allocation device, evacuation shelter allocation method and program - Google Patents

Description

The present invention relates to a shelter allocation device, a shelter allocation method, and a program.

Technologies being considered for guiding disaster victims to evacuation shelters include a technology that uses disaster simulations to notify disaster victims of information encouraging them to evacuate (Patent Document 1) and a technology that displays the evacuation shelters that disaster victims should head to on their mobile devices (Non-Patent Document 1).

JP 2020-52527 A

Umeki Hisato, et al., "A method for determining evacuation shelters considering the bias of crowding levels," Journal of Information Processing Society of Japan, Vol. 60, No. 2, pp. 608-616., [online], Internet <URL: http://arakawa-lab.com/wp-content/uploads/2019/03/umeki.pdf>

However, because the technology in Patent Document 1 provides information encouraging evacuation to individual disaster victims, it is up to the disaster victim to decide which evacuation shelter to go to, and there are cases in which the evacuation shelter they are directed to is full and they cannot enter. The technology in Non-Patent Document 1 makes it possible to request allocations so that evacuation shelters do not become full, but because it does not take into account the period during which disaster victims will use the shelter, when the number of disaster victims decreases and the number of shelters is reduced, the operating costs of the shelters and the costs for disaster victims to move between shelters can become excessive.

The present invention has been made in consideration of the above points, and aims to enable the allocation of evacuation shelters taking into account the period during which disaster victims will be using the shelters.

Therefore, in order to solve the above problem, the shelter allocation device has a first acquisition unit that acquires the location of each disaster victim at the time of the disaster occurrence, a second acquisition unit that acquires an estimate of the timing of when evacuation will be lifted for each of the disaster victims, and an allocation unit that acquires the allocation result to the shelter for each of the disaster victims for each divided period that divides the period by solving an integer programming problem aimed at minimizing the sum of the cost required for each of the disaster victims to move from their location to one of multiple shelters, the cost required for each of the disaster victims to move between shelters during the period until evacuation is lifted, and the operating cost of each of the multiple shelters.

It will be possible to allocate evacuation shelters taking into account the period of time that disaster victims will be using the shelters.

1 is a diagram illustrating an example of a configuration of an evacuation assistance system according to an embodiment of the present invention. 1 is a diagram illustrating an example of a hardware configuration of a shelter allocating device 10 according to an embodiment of the present invention. 2 is a diagram illustrating an example of a functional configuration of a shelter allocating device 10 according to an embodiment of the present invention. FIG. 11 is a flowchart illustrating an example of a processing procedure executed by the shelter allocating device 10. FIG. 2 is a diagram showing an example of the configuration of an evacuation shelter DB 121. FIG. 2 is a diagram showing an example of the configuration of a disaster victim information DB 122. FIG. 2 is a diagram showing an example of the configuration of a road network DB 123. FIG. 13 is a diagram showing an example of calculation results of evacuation costs for each disaster victim to each evacuation shelter. FIG. 13 is a diagram showing an example of calculation results of relocation costs between evacuation shelters. FIG. 13 is a diagram for explaining the meaning of a variable t. FIG. 13 is a diagram showing an example of an evacuation schedule. FIG. 11 is a diagram showing an example of an operation schedule.

An embodiment of the present invention will now be described with reference to the drawings. Figure 1 shows an example of the configuration of an evacuation support system in an embodiment of the present invention.

In Figure 1, the shelter allocation device 10 is connected to multiple shelter terminals 20 and multiple disaster victim terminals 30 via a network such as the Internet.

The shelter allocation device 10 is one or more computers that generate (calculate) evacuation schedules for each disaster victim and operation schedules for each evacuation shelter (hereinafter, information including these schedules is referred to as "schedule information") for a country or region where a disaster has occurred (hereinafter, referred to as "Region A") after the occurrence of the disaster (for example, immediately after the occurrence of the disaster, or the timing before each disaster victim reaches an evacuation shelter (hereinafter, simply referred to as "at the time of the disaster occurrence")). The shelter allocation device 10 distributes the evacuation schedules for each disaster victim to the victim terminal 30 of each disaster victim, and distributes the operation schedules for each evacuation shelter to the shelter terminal 20 of each evacuation shelter.

An evacuation schedule refers to information indicating the evacuation shelters in which each disaster victim should stay (the allocation of each disaster victim to an evacuation shelter) for each specified unit of time (hereinafter simply referred to as a "span"), and may differ for each disaster victim. An operation schedule refers to information indicating the span over which evacuation shelters should be open. A span is, for example, one day. In this case, the evacuation schedule and operation schedule are in units of one day. However, the span does not have to be one day. The span can be determined appropriately depending on the time granularity at which schedule information is desired to be generated, and may be, for example, one hour or one week.

When generating schedule information (i.e., allocating each disaster victim to one of the shelters), the shelter allocation device 10 determines the allocation that minimizes the sum of the cost of the disaster victim traveling to the shelter and the operating cost of the shelter.

Travel cost is the sum of evacuation cost and relocation cost. Evacuation cost refers to the cost required to travel from the location of victims to an evacuation shelter when a disaster occurs. In the first span, all victims are assumed to be accommodated in an evacuation shelter. Therefore, it is desirable for the span to be a unit of time sufficient to accommodate all victims in one of the evacuation shelters.

The relocation cost refers to the cost required for disaster victims to move between evacuation shelters in the second and subsequent spans. In other words, in this embodiment, since the reduction of evacuation shelter operating costs is also taken into consideration, it is assumed that disaster victims staying at a shelter will move (relocate) to another shelter during the evacuation period. Note that operating costs refer to the cost required to keep a shelter open (to operate) and may differ for each shelter. The operating costs of each shelter are incurred in each span from when the shelter is opened to when it is closed. Therefore, the fewer the shelters that are opened and the shorter the operating period, the smaller the operating costs will be. However, if disaster victims staying at a shelter are moved to another shelter in order to close the shelter, the cost of moving the disaster victims will be high.

For a given victim, the shelter assigned to the first span corresponds to the shelter to which the victim should first evacuate, and shelters assigned to the second and subsequent spans correspond to the shelters in which the victim should stay in those spans. If the shelter assigned to the t-th span for the given victim is different from the shelter assigned to the t+1-th span, this indicates that the victim should relocate to the t+1-th span. The number indicating the order of the spans (t for t-th span) will be referred to as the "span number" below.

The evacuation shelter terminal 20 is a terminal such as a PC (Personal Computer) that is placed at each evacuation shelter in area A. The evacuation shelter terminal 20 receives and outputs the operation schedule distributed from the evacuation shelter allocation device 10.

The victim terminal 30 is a mobile terminal such as a smartphone carried by each victim in area A. The victim terminal 30 provides the evacuation shelter allocation device 10 with information indicating the location of the victim terminal 30 (i.e., the location of the victim carrying the victim terminal 30), measured using a GPS (Global Positioning System) or the like. The victim terminal 30 also receives and outputs an evacuation schedule distributed from the evacuation shelter allocation device 10.

Figure 2 is a diagram showing an example of the hardware configuration of a shelter allocation device 10 in an embodiment of the present invention. The shelter allocation device 10 in Figure 2 has a drive device 100, an auxiliary storage device 102, a memory device 103, a processor 104, and an interface device 105, which are all interconnected by a bus B.

The program that realizes the processing in the shelter allocation device 10 is provided by a recording medium 101 such as a CD-ROM. When the recording medium 101 storing the program is set in the drive device 100, the program is installed from the recording medium 101 via the drive device 100 into the auxiliary storage device 102. However, the program does not necessarily have to be installed from the recording medium 101, but may be downloaded from another computer via a network. The auxiliary storage device 102 stores the installed program as well as necessary files, data, etc.

The memory device 103 reads out and stores the program from the auxiliary storage device 102 when an instruction to start the program is received. The processor 104 is a CPU or a GPU (Graphics Processing Unit), or a CPU and a GPU, and executes functions related to the shelter allocation device 10 in accordance with the program stored in the memory device 103. The interface device 105 is used as an interface for connecting to a network.

Figure 3 is a diagram showing an example of the functional configuration of the shelter allocation device 10 in an embodiment of the present invention. As shown in Figure 3, the shelter allocation device 10 has a shelter allocation unit 11 and a schedule distribution unit 12. Each of these units is realized by processing in which one or more programs installed in the shelter allocation device 10 are executed by the processor 104. The shelter allocation device 10 also uses databases (storage units) such as a shelter DB 121, a victim information DB 122, and a road network DB 123. Each of these databases can be realized using, for example, an auxiliary storage device 102, or a storage device that can be connected to the shelter allocation device 10 via a network.

The processing procedure executed by the shelter allocation device 10 will be described below. FIG. 4 is a flowchart for explaining an example of the processing procedure executed by the shelter allocation device 10. The processing procedure in FIG. 4 is executed when a disaster occurs. It is assumed that none of the shelters have been opened yet at the start of the processing procedure in FIG. 4 (i.e., when a disaster occurs), but in reality, preparations for opening may be underway. However, whether or not all of the shelters need to be opened is unknown at the start of the processing procedure.

In step S101, the shelter allocation unit 11 obtains the location information (point ID) and operating costs of each shelter from the shelter DB 121.

Figure 5 is a diagram showing an example of the configuration of the evacuation shelter DB 121. As shown in Figure 5, the evacuation shelter DB 121 stores (registers) for each evacuation shelter a shelter ID, a location ID, capacity, and operating costs. The evacuation shelter ID is identification information of the evacuation shelter. The location ID is the location ID of the location where the evacuation shelter is located, among the location IDs pre-registered in the road network DB 123. Each pre-registered location ID is associated with the location information (latitude and longitude) of the corresponding location, and is pre-stored, for example, in the auxiliary storage device 102. The capacity is the upper limit of the number of people that the evacuation shelter can accommodate. The operating cost is the cost required for operation per span, regardless of the number of people that can be accommodated.

Next, the evacuation shelter allocation unit 11 obtains the current location and address of each victim from the victim information DB 122 (S102).

Figure 6 is a diagram showing an example of the configuration of the victim information DB 122. As shown in Figure 6, the victim information DB 122 stores a victim ID, a location ID, an address, and an expected time of returning home for each victim. The victim ID is identification information for each victim terminal 30 (for each victim). The location ID is a location ID corresponding to the location closest to the current location (location at the time of the disaster occurrence) of the victim terminal 30 (victim). The address is the address of the victim. The expected time of returning home is information indicating an estimated value (predicted value) of the timing (span number) when the victim can start returning home (evacuation is lifted) due to the recovery of the transportation infrastructure, residences, etc. from the damage caused by the disaster. Note that the expected time of returning home is calculated in the process described below, so at the time of step S102, the value of the expected time of returning home for each victim is blank.

The information stored in the victim information DB 122 is collected, for example, from each victim terminal 30 when a disaster occurs. The ID of the victim terminal 30 may be used as the victim ID. Also, a point ID identified based on a position measured by a GPS (Global Positioning System) function of the victim terminal 30 may be stored for the victim related to the victim terminal 30. The address may be obtained from a database (not shown) that stores addresses in advance in association with the victim ID.

Next, the evacuation shelter allocation unit 11 calculates the travel distance from the current location to each evacuation shelter for each disaster victim by referring to the road network DB 123 (S103).

Figure 7 is a diagram showing an example of the configuration of road network DB123. One record in road network DB123 includes point ID1, point ID2, and distance. Point ID1 and point ID2 are point IDs corresponding to the points of one node and the other node of two adjacent nodes in a graph representing a road network. For example, intersections and branching points in a road network become nodes in the graph. Each evacuation shelter also becomes a node in the graph. The distance is the distance of the road section between point ID1 and point ID2. Note that the point ID in evacuation shelter DB121 and the point ID in disaster victim information DB122 are values registered as point ID1 or point ID2 in road network DB123.

The evacuation shelter allocation unit 11 can use, for example, a graph of the road network shown in the road network DB 123 to calculate the travel distance to each evacuation shelter for each victim by using the Dijkstra algorithm with the location ID of each victim as the start point, the location ID of each evacuation shelter as the end point, and the distance of the road section as the cost. Therefore, for victims with the same location ID (current location), the travel distance is calculated together.

Next, the evacuation shelter allocation unit 11 calculates the evacuation cost to each evacuation shelter for each disaster victim based on the travel distance of the disaster victim (S104). In this embodiment, the evacuation cost and relocation cost (i.e., travel cost) are expressed as a monetary amount. This is to ensure that the units of travel cost and operating cost are consistent in the integer programming problem described below. Therefore, the evacuation shelter allocation unit 11 converts the travel distance into a monetary amount using a specified function, and sets the converted monetary amount as the evacuation cost.

Figure 8 is a diagram showing an example of the calculation results of the evacuation costs for each disaster victim to each evacuation shelter. In Figure 8, the calculation results of the evacuation costs are represented by a matrix of M rows and M columns. Each row and each column of the matrix corresponds to a respective location ID. In other words, M is the number of location IDs (number of different locations). Of these, for rows, the row corresponding to the location ID corresponding to the current location of the disaster victim is valid. Also, for columns, the column corresponding to the location ID of any of the evacuation shelters is valid. The element values of the valid row and valid column indicate the evacuation cost required to travel from the location ID of the row to the evacuation shelter of the column.

Next, the shelter allocation unit 11 calculates the travel distance between each shelter based on the location (point ID) of each shelter by referring to the road network DB 123 (S105). For example, by using a graph of the road network shown in the road network DB 123 (Figure 7), the travel distance to each other shelter can be calculated for each shelter by using the Dijkstra algorithm with the point ID of the shelter as the start point, the point ID of each other shelter as the end point, and the distance of the road section as the cost.

Next, the shelter allocation unit 11 calculates the relocation costs between each shelter based on the travel distance calculated for each shelter (S106).

Figure 9 is a diagram showing an example of the calculation result of the relocation cost between evacuation shelters. In Figure 9, the calculation result of the relocation cost is expressed by a matrix of M rows and M columns. Each row and each column of the matrix corresponds to each location ID. However, only the row and column corresponding to the location ID of any of the evacuation shelters are valid. The value of the element of the valid row and column indicates the relocation cost required to move from the evacuation shelter in that row to the evacuation shelter in that column. In the example of Figure 9, the value of the element other than the diagonal element (i.e., the value of the relocation cost) is a constant (1 in this case). In this case, the allocation of evacuation shelters can be obtained with the aim of reducing the number of relocations. However, the relocation cost may also be calculated by converting the travel distance into a monetary amount using a function similar to the evacuation cost.

Next, the evacuation shelter allocation unit 11 acquires (calculates) an expected time of return home (span number) for each victim based on the address of the victim and the damage situation of the area related to the address (S107). For example, if it is possible to know the timing when public transportation will be restored in the area to which the victim's address belongs, that timing may be set as the expected time of return home. Alternatively, the extent of the damage may be classified into multiple stages, and the expected time of return home may be set as the timing after the passage of a period associated with each stage for each victim's location ID or address. Alternatively, the expected time of return home may be calculated by other methods. The evacuation shelter allocation unit 11 registers the expected time of return home calculated for each victim in the record corresponding to the victim in the victim information DB 122 (FIG. 6).

Next, the evacuation shelter allocation unit 11 obtains the allocation results for each disaster victim to each evacuation shelter for each span that divides the period until each disaster victim's expected return home time by solving an integer programming problem with the objective function of minimizing the sum of the travel cost (evacuation cost + relocation cost) and the operation cost (S108). The integer programming problem is shown below.

但し、各変数の意味は以下の通りである。
ｎ：被災者ＩＤ
ｍ：避難所ＩＤ又は地点ＩＤ
Ｃ_ｍ：避難所ｍの定員（図５）
ｔ：時刻（スパン番号）
τ_ｎ：被災者ｎの想定帰宅時刻（図６）
ｆ_ｍ：避難所ｍの１スパン分の運営コスト（図５）
ｄ_ｔｍｍ'：時刻ｔに被災者が地点ｍ又は避難所ｍから避難所ｍ'に移動するコスト
ｔ＝１の場合、ｄｔｍｍ'＝地点ｍから避難所ｍ'への避難コスト（図８）
ｔ＞１の場合、ｄｔｍｍ'＝避難所ｍから避難所ｍ'への移転コスト（図９）
ｘ_ｔｍｎ：ｔ＝０の場合、時刻ｔに被災者ｎが災害発生時にいる地点ｍにいるか否かを示し、ｔ＞０の場合、時刻ｔに被災者ｎが避難所ｍに滞在するか否かを示す変数
ｙ_ｔｍ：時刻ｔ避難所ｍを運営するか否かを示す変数
ｚ_{ｔｍｍ'ｎ}：時刻ｔに被災者ｎが避難所ｍから避難所ｍ'に移転するか否かを示す変数（ｚ_{ｔｍｍ'ｎ}＝ｘ_{（ｔ－１）ｍｎ}×ｘ_ｔｍ'ｎ
また、式（１）～（１２）までの各式の意味は、以下の通りである。

The meaning of each variable is as follows:
n: victim ID
m: Evacuation shelter ID or location ID
C _m : Capacity of evacuation shelter m (Figure 5)
t: time (span number)
τ _n : Estimated time of return home for victim n (Figure 6)
f _m : Operating cost for one span of shelter m (Figure 5)
d _tmm' : The cost of a disaster victim moving from point m or shelter m to shelter m' at time t. When t = 1, dtmm' = the evacuation cost from point m to shelter m' (Figure 8)
If t>1, dtmm' = relocation cost from shelter m to shelter m' (Figure 9)
x _tmn : A variable indicating whether or not victim n is at location m at the time of disaster occurrence at time t when t=0, and whether or not victim n will be staying at evacuation shelter m at time t when t>0. y _tm : A variable indicating whether or not evacuation shelter m will be operated at time t. z _tmm'n : A variable indicating whether or not victim n will move from evacuation shelter m to evacuation shelter m' at time t. (z _tmm'n = x _{(t-1) mn} × x _tm'n
The meanings of the formulas (1) to (12) are as follows:

Equation (1) is the objective function for minimizing the sum of the transportation cost of each disaster victim, represented by the first term, and the operating cost of each evacuation shelter, represented by the second term.

Equation (2) is a constraint that indicates that the number of disaster victims must not exceed the capacity of the evacuation shelter, and that the number of disaster victims cannot stay in a closed evacuation shelter. However, at the time of the disaster occurrence (t=0), this constraint does not apply because the disaster victims have not yet been accommodated in the evacuation shelter.

Equations (3) and (4) are constraints that indicate that victims will stay in one of the evacuation shelters until they return home (until the evacuation order is lifted).

The formulas (5), (6), and (7) are constraints indicating that a user moves from a relocation source to a relocation destination only if the user stays at the relocation source before the relocation and at the relocation destination after the relocation. In other words, the constraints are for (z _tmm′n =x _{(t−1) mn} ×x _tm′n ).

Formula (8) is a constraint condition for victim n to be present in a given location m ^~ (n) when a disaster occurs. Note that m ^~ is the symbol m in formula (8) with ~ added above it.

Equation (9) is a constraint that a closed shelter (a shelter with zero victims assigned at a certain time t) will not be reopened (no victims will be assigned) at a subsequent time t.

Equations (10), (11), and (12) are conditions for restricting the values of x _tmn , y _tm , and z _tmm′n to 0 or 1.

FIG. 10 is a diagram for explaining the meaning of the variable t. In FIG. 10, the value of t shown on the upper side of the rectangle representing each span indicates t for x _tmn . That is, t for x _tmn indicates the timing of the end of the span related to the span number indicated by the t. On the other hand, the value of t shown on the lower side of the rectangle representing each span indicates t for y _tm and z _tmm'n . That is, t for y _tm and z _tmm'n indicates the span related to the span number indicated by the t. Note that the length of each span does not need to be the same as the length of the other spans. Up to this point, the length of the span has been set to a "prescribed fixed period (unit time)", but by appropriately determining t for x _tmn , the length of the span can be set to an unequal time. For example, it is possible to set t=1 for one day after the occurrence of a disaster (t=0), t=2 for one week, t=3 for one month, and so on.

In the above integer programming problem, x _tmn , y _tm , and z _tmm'n are decision variables, that is, the values of x _tmn , y _tm , and z _tmm'n are calculated by solving the integer programming problem.

Here, x _tmn indicates the evacuation schedule of victim n. That is, x _tmn having a value of 1 indicates that victim n should stay at evacuation shelter m at time t. x _tmn having a value of 0 indicates that victim n should not stay at evacuation shelter m at time t. Therefore, based on x _tmn , an evacuation schedule such as that shown in FIG. 11 is obtained. FIG. 11 shows, for each evacuee, the evacuation shelter where the evacuee should stay at each time t until the expected return home time of the evacuee.

Also, _ytm indicates the operation schedule of the shelter m. That is, _ytm having a value of 1 indicates that the shelter m should be in operation at time t. _xtmn having a value of 0 indicates that the shelter m should not be in operation at time t. Therefore, based on _ytm , an operation schedule such as that shown in Fig. 12 is obtained. Fig. 12 shows, for each shelter, the span numbers to be opened and the span numbers to be closed.

Next, the schedule distribution unit 12 transmits to each victim terminal 30 the evacuation schedule of the victim associated with that victim terminal 30 (S109). Each victim can know the evacuation shelter in which to stay in each subsequent span according to the evacuation schedule received by his/her victim terminal 30.

Next, the schedule distribution unit 12 transmits to each evacuation shelter terminal 20 the operation schedule of the evacuation shelter related to that evacuation shelter terminal 20 (S110). Each evacuation shelter can know the timing to open and close according to the operation schedule received by the evacuation shelter terminal 20 of that evacuation shelter.

As described above, according to this embodiment, the allocation of evacuation shelters for each disaster victim is calculated for each span, taking into consideration not only the evacuation costs of each disaster victim, but also the relocation costs of each disaster victim and the operating costs of each evacuation shelter that arise during the period in which each disaster victim uses the evacuation shelter. Therefore, it is possible to allocate evacuation shelters taking into consideration the period in which each disaster victim will use the evacuation shelter.

In this embodiment, the evacuation shelter allocation unit 11 is an example of a first acquisition unit, a second acquisition unit, and an allocation unit. The schedule distribution unit 12 is an example of a first transmission unit and a second transmission unit. A span is an example of a divided period.

Although the embodiments of the present invention have been described in detail above, the present invention is not limited to such specific embodiments, and various modifications and variations are possible within the scope of the gist of the present invention as described in the claims.

10 Evacuation shelter allocation device 11 Evacuation shelter allocation unit 12 Schedule distribution unit 20 Evacuation shelter terminal 30 Disaster victim terminal 100 Drive device 101 Recording medium 102 Auxiliary storage device 103 Memory device 104 Processor 105 Interface device 121 Evacuation shelter DB
122 Disaster Victim Information DB
123 Road Network DB
B Bus

Claims (7)

A first acquisition unit that acquires the location of each disaster victim at the time of the occurrence of a disaster;
A second acquisition unit that acquires an estimated value of a timing for lifting evacuation for each of the disaster victims;
an allocation unit that acquires allocation results of each of the disaster victims to the evacuation shelters for each divided period by solving an integer programming problem for the purpose of minimizing the sum of a cost required for each of the disaster victims to move from the location to any one of a plurality of evacuation shelters, a cost required for each of the disaster victims to move between the evacuation shelters during the period until the evacuation order is lifted, and an operating cost of each of the plurality of evacuation shelters;
A shelter allocation device comprising: The second acquisition unit acquires an estimate of the timing of lifting the evacuation warning based on an address of each of the disaster victims.
2. The shelter allocation device according to claim 1. The integer programming problem has a constraint that a shelter to which no disaster victims are assigned in a certain divided period will not be assigned any disaster victims in the following divided periods.
3. The shelter allocation device according to claim 1 or 2. a first transmission unit that transmits the allocation result of each of the disaster victims to each of the disaster victim terminals;
4. The shelter allocation device according to claim 1, further comprising: a second transmission unit that transmits information indicating the opening timing and closing timing of each of the evacuation shelters based on the allocation result to a terminal of each of the evacuation shelters;
5. The shelter allocation device according to claim 1, further comprising: A first acquisition step of acquiring a location of each disaster victim at the time of occurrence of a disaster;
A second acquisition step of acquiring an estimate of a timing for lifting evacuation for each of the disaster victims;
an allocation procedure for obtaining an allocation result of each of the disaster victims to the evacuation shelters for each divided period by solving an integer programming problem for the purpose of minimizing the sum of a cost required for each of the disaster victims to move from the location to any one of a plurality of evacuation shelters, a cost required for each of the disaster victims to move between the evacuation shelters during the period until the evacuation order is lifted, and an operating cost of each of the plurality of evacuation shelters;
A shelter allocation method characterized in that the above steps are executed by a computer. A program that causes a computer to function as a shelter allocation device according to any one of claims 1 to 5.

Images