JP5600511B2 - Evacuation time prediction device and evacuation time prediction method - Google Patents

Description

  The present invention relates to an evacuation time predicting apparatus and an evacuation time predicting method for predicting an evacuation time from a multi-tier building.

  Conventionally, evacuation time from a building has been predicted for a multi-level building having stairs. For example, Non-Patent Document 1 discloses a method for predicting an evacuation time by a multi-agent simulation in which a distribution of refugees on each floor is input and each refugee is simulated. In addition, as shown in Non-Patent Documents 2 and 3, some building standard laws make predictions using the evacuation safety verification method. Here, the method shown in Non-Patent Document 2 calculates the evacuation for each floor, and the method shown in Non-Patent Document 3 calculates the evacuation of the entire building together.

Naohiro Takeichi, Yoshikazu Minegishi, Katsuyuki Yoshida, Yuki Sano, Kazuto Hayashida, Prediction of the whole evacuation behavior by multi-agent model-Basic performance of SimTread, a pedestrian simulation system Part 3-, Annual Meeting of the Fire Society of Japan, 2008 May Ministry of Construction Notification No. 1441 on the Floor Evacuation Safety Verification Act Ministry of Construction Notification No. 1442 on the Floor Evacuation Safety Verification Act

  The simulation method shown in Non-Patent Document 1 predicts the evacuation time to some degree of accuracy, but has the following problems. First, it is not practical due to the complexity of input and the long calculation time. Second, even if the flow of people on each floor can be predicted with a certain degree of accuracy, there is little knowledge about merging at the stairs, especially in high-rise buildings. There is a problem that the prediction accuracy after the merging of can not catch up.

  The methods shown in Non-Patent Documents 2 and 3 are relatively simple and practical in calculation, but have the following problems. First, it is not practical due to the complexity of input. Since the stairs and floors are not considered separately, it is impossible to calculate the bias of the concentration of evacuees on the stairs. Secondly, there is a problem that the effective flow coefficient becomes excessively small in the calculation of the exit passage time for evacuation of the entire building. In many cases, the first can be on the dangerous side and the second on the safe side, and the two can cancel each other, resulting in a somewhat safe evaluation. Third, because the floor evacuation and the whole building evacuation are calculated separately, the distribution of the number of evacuees in the hierarchy direction is not reflected.

  As described above, the conventional method has left a problem in the balance between the prediction accuracy of the evacuation time and the calculation speed. The present invention has been made in view of the above, and in the prediction of evacuation time from a multi-level building, an evacuation time prediction device capable of obtaining a highly accurate prediction result in a shorter time than the conventional method, and It aims at providing the evacuation time prediction method.

  In order to achieve the above object, an evacuation time predicting apparatus according to the present invention is an evacuation time predicting apparatus for predicting an evacuation time from a multi-level building having stairs. Model generating means for generating a model of a building with one or more continuous cells for each and each floor connected to one of the cells corresponding to the floor; and the building generated by the model generating means In the model, the initial state input means for inputting the initial state of each cell and the number of evacuees on each floor, and the model of the building generated by the model generating means, in order from the cell closest to the evacuation exit, for each unit time The status of the number of evacuees in each cell, either the initial status entered by the initial status input means or the status at the previous time (time), the status of the inflow from the cell one level above, and the evacuees from the floor Transition state calculating means for calculating based on the state of advancement and the horizontal approach rate indicating the rate of evacuees entering the stairs cell from the floor with respect to the inflow, and evacuation of each cell calculated by the transition state calculating means It is determined whether the number of evacuees indicates that the evacuation has been completed, and if it is determined that the evacuation has not been completed, the transition state calculation means determines whether the evacuee of each cell at the next time When the evacuation end determination means for calculating the number of persons and the evacuation end determination means determine that the evacuation is completed, the evacuation time is output according to the number of unit times calculated by the transition state calculation means. And an output means.

  In the evacuation time predicting apparatus according to the present invention, the evacuation time is predicted in consideration of the mutual influences of advancement from each floor to the staircase as an evacuation route and movement on the stairs. Therefore, according to the evacuation time prediction apparatus according to the present invention, it is possible to obtain a prediction result with higher accuracy than the simple methods shown in Non-Patent Documents 2 and 3. Further, in the evacuation time prediction apparatus according to the present invention, the evacuation time is predicted by the number of people included in the cell without simulating each refugee, so that the simulation by the multi-agent as shown in Non-Patent Document 1 is performed. Compared with the method according to, even in a large building, the prediction result can be obtained in a realistic time. That is, according to the evacuation time predicting apparatus according to the present invention, it is possible to obtain a highly accurate prediction result in a shorter time compared to the conventional method in the prediction of the evacuation time from a multi-tiered building.

  The initial state input means also inputs information indicating the timing at which evacuation is started at least on any floor, and the transition state calculation means is based on information indicating the timing input by the initial state input means. It is desirable to calculate the status of the number of evacuees in each cell. According to this structure, the time difference of the evacuation start in each floor can be provided. Thereby, the evacuation time that is more realistic can be predicted.

  The model generation means preferably determines the number of cells based on the number of people that can flow between cells per unit time. According to this configuration, more appropriate evacuation time can be predicted.

  The output means preferably outputs information indicating the state of the number of evacuees in each cell for each unit time calculated by the transition state calculation means. According to this configuration, it is possible to grasp not only the final result of the final evacuation time but also the evacuation process.

  By the way, the present invention can be described as the invention of the evacuation time prediction apparatus as described above, as well as the invention of the evacuation time prediction method as follows. This is substantially the same invention only in different categories, and has the same operations and effects.

That is, the evacuation time prediction method according to the present invention is an evacuation time prediction method that is an operation method of an evacuation time prediction device that predicts evacuation time from a multi-level building having stairs. A model generation step for generating a model of a building in which one or more continuous cells for each floor are connected and each floor is connected to one of the cells corresponding to the floor. In the building model, the initial state input step for inputting the initial state of the number of evacuees in each cell and each floor, and the unit in the order from the cell closest to the evacuation exit based on the building model generated in the model generation step. The status of the number of evacuees in each cell for each hour, the initial state input in the initial state input step and the state of the previous time and inflow from the cell one level above A transition state calculating step for calculating based on the state and the state of the refugees entering from the floor and the horizontal approach rate indicating the rate at which the refugees enter the stairs cell from the floor with respect to the inflow, and the transition state calculating step If the status of the number of evacuees calculated in each cell indicates whether the evacuation has been completed, and if it is determined that the evacuation has not been completed, the transition state calculation step The evacuation end determination step for calculating the number of evacuees in each cell, and when it is determined that the evacuation has been completed in the evacuation end determination step, according to the number of unit times calculated in the transition state calculation step And an output step of outputting the evacuation time.

  In the present invention, since the evacuation time is predicted in consideration of the mutual influence of the advance to the stairs which is the evacuation route and the movement on the stairs from each floor, a highly accurate prediction result can be obtained. In the present invention, since the evacuation time is predicted by the number of people included in the cell without simulating each evacuee, the prediction result can be obtained in a realistic time even for a large-scale building. . That is, according to the present invention, in prediction of evacuation time from a multi-level building, a highly accurate prediction result can be obtained in a shorter time compared to the conventional method.

It is a figure which shows the functional structure of the evacuation time prediction apparatus which concerns on embodiment of this invention. It is a figure which shows notionally the model used for prediction of evacuation time. It is a figure which shows notionally the movement of the refugee between cells. It is a figure which shows the matrix used in order to calculate the transition state of each cell in each time. It is a figure which shows the calculation of the transition state of each cell between the time using a matrix. It is an example of the display which changes and shows the accommodation degree of the evacuees of each floor and stairs for every unit time. It is an example of the display of the line graph which shows the accommodation degree according to the elapsed time of each floor. It is a flowchart which shows the whole process (evacuation time prediction method) performed with the evacuation time prediction apparatus which concerns on embodiment of this invention. It is a flowchart which shows the process which calculates the transition state of the accommodation state of the evacuee of a cell, and the process which judges the completion of evacuation. It is a flowchart which shows the process which calculates the transition state of the accommodation state of the evacuee of a cell, and the process which judges the completion of evacuation. It is a flowchart which shows the process which calculates the transition state of the accommodation state of the evacuee of a cell, and the process which judges the completion of evacuation. It is the figure shown according to the case about the transition of the accommodation state of the evacuee of the cell when there is an approach from the floor to the stairs. It is a graph which shows the actual flow coefficient of each floor which is a result of having actually calculated evacuation time by the method which concerns on embodiment of this invention. It is a line graph which shows the accommodation degree according to the elapsed time of each floor which is the result of having actually calculated evacuation time by the method which concerns on embodiment of this invention. It is the figure which showed the accommodation degree of the refugee of each floor and stairs which is the result of having actually calculated evacuation time by the method which concerns on embodiment of this invention. It is a graph which shows the evacuation time according to the horizontal approach rate which is the result of having actually calculated evacuation time by the method which concerns on embodiment of this invention.

  Hereinafter, preferred embodiments of an evacuation time prediction apparatus and an evacuation time prediction method according to the present invention will be described in detail with reference to the drawings. In the description of the drawings, the same elements are denoted by the same reference numerals, and redundant description is omitted.

  FIG. 1 shows an evacuation time prediction apparatus 10 according to the present embodiment. The evacuation time prediction device 10 is a device that predicts the evacuation time of an evacuee located in a building from a multi-level building having stairs. Such a building is, for example, a high-rise building such as a complex facility. The predicted evacuation time is used for verification of building safety.

  The evacuation time is predicted by information processing (simulation calculation) by the evacuation time prediction device 10. Specifically, the evacuation time prediction apparatus 10 corresponds to an information processing apparatus such as a workstation or a PC (Personal Computer), and is configured by hardware such as a CPU (Central Processing Unit) or a memory. The evacuation time predicting device 10 exhibits each function described later when these components are operated by a program or the like. In the present embodiment, the evacuation time prediction device 10 is realized by a single device, but may be realized by an information processing system configured by connecting a plurality of information processing devices to each other via a network.

  As shown in FIG. 1, the evacuation time prediction apparatus 10 includes a model generation unit 11, an initial state input unit 12, a transition state calculation unit 13, an evacuation end determination unit 14, and an output unit 15 as functional configurations. And is configured.

  The model generation unit 11 is a model generation unit that generates a model of a building to be predicted for evacuation time. Specifically, the model is generated on a memory included in the evacuation time prediction device 10 as will be described later, and simulation calculation is performed based on the generated model to predict the evacuation time.

  FIG. 2 conceptually shows a building model to be predicted for evacuation time in this embodiment. As shown in FIG. 2, the staircase 20 provided in the building is divided into a plurality of cells (small spaces) 21 continuous in the up and down direction (moving direction of the staircase) that accommodate the evacuees. The evacuees move the cell in the evacuation direction (usually the direction from the top to the bottom because there is an exit on the basement (first floor) of the building). The plurality of cells 21 include a cell 21 a that is provided with an entrance to each floor 30 and is connected to each floor 30 so that evacuees can enter and leave the floor 30, and other cells 21 b. The cell 21 is provided with one or more cells for each floor. In addition, as shown in FIG. 2, each floor 30 is also provided with one or more cells 31 for accommodating evacuees in the order closer to the stairs. Each cell 21 and 31 is set with the maximum number of evacuees that can be accommodated.

  FIG. 3 conceptually shows the movement of evacuees between the cells 21 and 31. FIG. 3 shows that there are evacuees in the colored portions of each cell (indicating the number of evacuees) and there are no evacuees in the uncolored portions. As shown in FIG. 3, in the cells 21a and 21b of the staircase, if the lower cells 21a and 21b are empty, the evacuee moves from the upper cells 21a and b (hereinafter, this movement is referred to as “march”). To do. In addition, the evacuees move (march) from the upper cell 21b to the cell 21a connected to the floor, and the evacuees move from the cell 31 on the floor (hereinafter, this movement is referred to as “(horizontal) approach”). (See the third and fourth figures from the left in FIG. 3). When the evacuees reach the evacuation floor (usually the basement), they leave the building from the exit of the evacuation floor (this movement is hereinafter referred to as “exit”). In the present embodiment, the movement of the evacuees between the cells 21 and 31 as described above is repeated, and the time (on the simulation) until all the evacuees evacuate from the building is calculated.

  The model generation unit 11 generates a model by determining each parameter indicating the model. The parameter is input to the evacuation time prediction apparatus 10 or is determined by the model generation unit 11 based on information input to the evacuation time prediction apparatus 10. The input of information to the evacuation time prediction apparatus 10 is performed, for example, by an input operation on the evacuation time prediction apparatus 10 by the user of the evacuation time prediction apparatus 10.

  Information input to the evacuation time prediction apparatus 10 includes, for example, the number of floors of the building, the number of people (as an initial state), the stairs width, the stairs area and the walking distance (for one layer), and the door from each floor to the stairs The width, the exit width of the stairs on the stairs, the horizontal entrance rate, and the evacuation start time of each floor. For example, the model generation unit 11 determines information regarding the cell 21 based on the above information. First, the number of people who can move in the stairs per unit time (for example, 1 second) is determined as the number of people that can be accommodated in the cell 21. Specifically, it is obtained from the flow coefficient at the stairs × the width of the stairs. Here, the flow coefficient at the stairs is a value indicating the number of people moving to 1 m per second, and is input by the user of the evacuation time predicting device 10, for example. For example, a value of the flow coefficient N = 80/60 person / second m (1.33 person / second m) at the stairs is used.

Subsequently, the model generation unit 11 stores the staircase by dividing the staircase area A _st (the staircase area from the current floor to the immediately lower floor (directly upper floor in the case of the basement)) by a predetermined value (for example, 0.25). The total number of cells is calculated by calculating the number of persons and dividing the number of persons accommodated by the number of cells 21. That is, it is desirable that the model generation unit 11 determines the number of cells 21 based on the number of people that can flow between the cells 21 per unit time. If cells are set in this way, more appropriate evacuation time can be predicted from the viewpoint of calculation accuracy and the like. However, the number of cells 21 does not necessarily have to be as described above. For example, any number may be used as long as one or more cells are set on one floor.

  In addition, parameters indicating how the “entry”, “march”, and “exit” are performed are also determined by the model generation unit 11. This will be described in more detail. The number of evacuees in each cell 21 and 31 is managed in a matrix as shown in FIGS. 4 and 5, etc., which will also be described later. The model generated by the model generation unit 11 is referred to by the following functional means and used for calculating the evacuation time.

  The initial state input unit 12 is an initial state input unit that inputs an initial state of the number of evacuees in the building model generated by the model generation unit 11. Of the information input to the evacuation time prediction device 10, the initial state input unit 12 inputs, for example, the number of people on each floor as the initial state (the number of stairs is 0). Specifically, this information is given as information indicating the number of cells 21 and 31. Information indicating the initial state input by the initial state input unit 12 is referred to by the following functional means and used for calculating the evacuation time. This initial state is preferably determined in consideration of the fire prevention section, various divisions, the use of each floor, and the like.

  Based on the building model generated by the model generation unit 11, the transition state calculation unit 13 calculates the state of the number of cells 21 and 31 at the next time from the state of the number of cells 21 and 31 at the previous time. It is a transition state calculation means. The interval between the previous time and the next time is a predetermined unit time, for example, 1 second. For example, this unit time is input in advance to the evacuation time prediction device 10 and stored in the transition state calculation unit 13. The state of the number of cells 21 and 31 at the previous time t−1 is calculated by the initial state (the time before time t = 0) input by the initial state input unit 12 and the transition state calculation unit 13 itself. One of the states.

The building model generated by the model generation unit 11 includes a matrix as shown in FIG. A step cell 21 at a time t is represented by the following vector V (t) _i .

Here, each component C (t) of the matrix indicates the number of evacuees accommodated in each cell 21, and the subscripts 1, 2,..., N identify the cell 21 and are from cells close to the basement. It is given to be in order. n is the number of cells and is a given number determined as described above. The subscript i indicates the number of times the calculation is performed sequentially at a certain time t (for the number n of cells). That is, at a certain time t, the transition state calculation unit 13 repeats n operations from the previous time t−1 to calculate the state of the cell 21 for the next unit time t (ie, V (t) _n. Shows the state of the number of cells 21 at time t). In the process, V (t) _{i corresponding} to the number n of cells 21 is obtained, and the n column vectors are set as [V (t) ₁ , V (t) ₂ ,..., V (t) _n ]. What is represented is the matrix shown in FIG.

The state of the number of cells is calculated in order from the cell 21 with the closest escape port. As shown in FIG. 5, from V (t−1) _n indicating the number of cells 21 at the previous time t−1, first, C (t) _{1 of the} cell 21 closest to the basement at the next time t. _{, 1} is calculated. Next, C (t) ₂ , ₂ of the cell above the cell 21 is calculated based on the calculation result. This calculation is performed sequentially (n times), and finally 21 _{n, n of the} cell farthest from the escape port is calculated.

In addition, for the cells 31 on each floor, the number of evacuees to be accommodated is stored as F (t) _{i, j in the same} manner as C (t) _{i, j} in the cell 21 of the staircase, and the transition state is changed. calculate. For convenience of matrix operation, the number of cells of the floor 31 (F (t) _{i, j} ) (number of subscripts j) is also equal to the number of cells of the staircase cell 21 (C (t) _{i, j} ). It is better to do the same. In that case, F (t) _{i, j} for the cell 31 on the floor not connected to the cell 21 of the staircase is always 0.

  The calculation of the state of the number of people in each cell 21 is performed based on the state of march (inflow) from the cell 21 above. As the information indicating the state of march (inflow) from the cell 21 that is one level above, the number marching at maximum in one unit time calculated by the flow coefficient N in the staircase × the step width B can be used.

Further, regarding the calculation of the state of the number of cells 21a connected to the stairs, the status of the refugees entering from the cell 31 on the floor and the rate at which the refugees enter the cell 21a on the stairs from the floor with respect to the advance. This is also performed based on the horizontal approach rate indicating. The state of the evacuees entering from the cell 31 to the cell 21a on the floor is the number of maximum entry in one unit time calculated by the flow coefficient N _i from the floor to the stairs × the door width B _i from each floor to the stairs. Can be used. Here, for example, a value of the flow coefficient N _i = 90/60 person / second m (1.5 person / second m) from the floor to the stairs is used. The horizontal approach rate α is a value between 0 and 1 (0 ≦ α ≦ 1), and is set and input in advance by the user of the evacuation time prediction device 10 based on the structure of the building and the like. Value.

Further, the cell 21 on the evacuation floor is performed based on the state of exit from the building. As the state of exit from the building, the maximum exit number in one unit time calculated by the flow coefficient N _{1 at} the evacuation floor × the evacuation floor exit width B ₁ of the stairs can be used.

Moreover, it is good also as setting the timing which starts evacuation for every floor. In that case, evacuation for each floor (advance from the floor to the stairs) is started according to the timing set for each floor. More specifically, for each cell 21 of the staircase, the time t _{i. A delay} (where i is a subscript specifying a cell) is set, and a transition state is calculated based on the _delay . For example, evacuation may be started all at once from time 0. Alternatively, evacuation starts at time 0 on certain floors (for example, a fire has occurred), and on other floors,

Evacuation may be started after a delay of (seconds). Here, A _floor is the floor area.

  The evacuation end determination unit 14 refers to the state of the number of evacuees in the cells 21 and 31 calculated by the transition state calculation unit 13 and determines whether or not the state indicates that the evacuation from the building has been completed. This is an evacuation end determination means. If it is determined that the evacuation has not been completed, the evacuation end determination unit 14 causes each cell 21 at the next time (time advanced by one unit time from the previous time) to the transition state calculation unit 13. The number of 31 evacuees is calculated. When it is determined that the evacuation is completed, the evacuation end determination unit 14 outputs information indicating the number of unit times (simulated) that has passed so far to the output unit 15. Further, the evacuation end determination unit 14 may output information indicating the progress of calculation by the transition state calculation unit 13 to the output unit 15.

  The specific contents of the calculation process (simulation) by the transition state calculation unit 13 and the evacuation end determination unit 14 will be described in detail later.

  The output unit 15 is an output unit that outputs an evacuation time corresponding to the number of unit times input from the evacuation end determination unit 14. For example, if the unit time is 1 second, the evacuation time is output as the number of unit times (seconds). This output is performed, for example, by displaying on a display device included in the evacuation time prediction device 10 so that the user of the evacuation time prediction device 10 can refer to the prediction result. Further, not only the final evacuation time but also a more detailed display may be performed based on the result calculated by the transition state calculation unit 13.

  For example, as shown in FIG. 6, the illustrated display may be performed by changing the accommodation levels of evacuees on each floor and stairs for each unit time (that is, display as a moving image). The part shown in dark color indicates that the evacuees are being accommodated at that time. Or as shown in FIG. 7, it is good also as displaying the accommodation degree according to the elapsed time of each floor as a line graph (each line respond | corresponds to each floor). In addition, the number of intruders per unit time from each floor to the stairs may be displayed as a line graph.

  That is, the output unit 15 may output information indicating the state of the number of evacuees in the cells 21 and 31 at each time calculated by the transition state calculation unit 13. According to this configuration, it is possible to grasp not only the final result of the final evacuation time but also the evacuation process. Thereby, the safety of various patterns can be confirmed in the design practice, and can be used for evacuation safety design.

  Further, the output by the output unit 15 is not necessarily a display output, and an arbitrary output such as transmitting information to another apparatus can be performed.

  Then, the process (evacuation time prediction method) performed with the evacuation time prediction apparatus 10 which concerns on this embodiment is demonstrated using the flowchart of FIGS. First, the entire process executed by the evacuation time prediction apparatus 10 will be described using the flowchart of FIG. 8, and then the cell refugee accommodation status by the transition state calculation unit 13 using the flowcharts of FIGS. 9 to 11. A process for calculating the transition state of the evacuation and a process for determining completion of evacuation by the evacuation end determination unit 14 will be described.

  In the evacuation time prediction device 10, first, parameters for predicting the evacuation time are input (S01). As described above, the input parameters are the number of floors of the building, the number of people in each floor (as the initial state), the stairs width, the stairs area and walking distance (for one layer), the door width from each floor to the stairs, and the evacuation of the stairs Information such as floor exit width, horizontal approach rate, and evacuation start time of each floor. The parameters input to the evacuation time prediction device 10 are output to each functional means according to the contents.

  Subsequently, the model generation unit 11 generates a model of the building to be predicted for the evacuation time (S02, model generation step). Specifically, this model is generated by generating a matrix as described above according to the input parameters on the memory of the evacuation time prediction apparatus 10.

  Subsequently, an initial state of the number of evacuees in the above model is input by the initial state input unit 12 (S03, initial state input step). Specifically, this input is performed by storing information indicating the initial state according to the input parameters in matrix data indicating the initial state of each of the cells 21 and 31 of the model.

  Subsequently, the state of the number of evacuees in the cells 21 and 31 of the building at a time obtained by advancing a preset unit time by one unit from the initial state input by the initial state input unit 12 by the transition state calculating unit 13 (Transition state) is calculated (S04, transition state calculation step).

  Subsequently, the state of the number of evacuees in each cell 21, 31 calculated by the transition state calculation unit 13 is referred to by the evacuation end determination unit 14, and the state indicates that the evacuation from the building has been completed. Is determined (S05, evacuation end determination step). When it is determined that the evacuation is not completed, the cells 21 and 31 at the time advanced by one unit time from the time when the state is calculated from the evacuation end determination unit 14 to the transition state calculation unit 13. Control is performed to calculate the status of the number of evacuees (S06, evacuation end determination step).

  Each cell 21 of the building at a time obtained by advancing a preset unit time by one unit from the state of the number of evacuees of the cells 21 and 31 calculated before by the transition state calculation unit 13 under the control. The state (transition state) of the number of 31 evacuees is calculated (S04, transition state calculation step). The above determination (S05) by the evacuation end determination unit 14 is repeatedly performed on the state.

  If it is determined in S05 that the evacuation has been completed, the number of unit times calculated by the transition state calculation unit 13 from the evacuation end determination unit 14 (the number of times the calculation of the repetition S04 has been performed). ) Is output to the output unit 15. Subsequently, the evacuation time corresponding to the input information is output by the output unit 15 (S07, output step). For example, by referring to this output, the user of the evacuation time prediction apparatus 10 can grasp the evacuation time from the building. The above is the entire processing executed by the evacuation time prediction apparatus 10.

  Subsequently, using the flowcharts of FIGS. 9 to 11, processing for calculating the transition state of the evacuee accommodation state of the cell by the transition state calculation unit 13 and processing for determining completion of evacuation by the evacuation end determination unit 14 (FIG. 9). The process of S04 to S06 in FIG. In this process, at the start of the process, the simulation time t = 0 and the unit time is 1 (second).

In addition, the notation in the following description is defined as follows.

First, it is determined whether or not t ≧ min ( _ti.delay ) is satisfied for the simulation time t (S101). Here, min is a function whose function value is the minimum value of all the variable values. If the time t does not satisfy the above relationship, that is, if evacuation has not yet started on any floor, t = t + 1 is set (S102). That is, in this case, since the number of people accommodated in the cells 21 and 31 is not changed, the calculation for the next time advanced by one unit time is started.

If it is determined that the relationship is satisfied in the determination in S101, the state of the number of evacuees in the cells 21 and 31 at time t is obtained thereafter. First, i = 1 is set in order to perform processing on the evacuation floor (S103). Subsequently, C ₁ (t) due to exit in the cell 21 of the staircase on the evacuation floor is calculated (S104). V ₁ (t) and U ₁ (t) are calculated using final column vectors V _n (t−1) and U _n (t−1) indicating the transition state at time t−1 as follows.

That is, in the previous time t-1, number C _(t-1) of stairs cells 21 in the evacuation floor _{n, 1} is that exceeds the maximum exit number 1.3B ₁ from the evacuation floor outlet in a unit time In this case, C (t) _{1,1 is set} to C (t-1) _{n, 1} -1.3B _1, and 0 otherwise. When t = 0, the state of t−1 is assumed to be the initial state by the initial state input unit 12 (the same applies hereinafter).

Subsequently, C ₁ (t) and C ₂ (t) due to the march from the cell 21 above the evacuation floor to the cell 21 at the evacuation floor in the cell 21 and the cell 21 above the evacuation floor are calculated ( S105). As described below, V ₂ (t) and U ₂ (t) are calculated using column vectors V ₁ (t) and U ₁ (t) indicating the transition state after the calculation in S104.

That is, after the calculation of S104, the number of vacant persons “CC (t) _1,1 ” in the cell 21 of the staircase on the evacuation floor is equal to _or less than the number of persons C (t) ₁ , _{2 in} the cell 21 above The number C (t) _2,2 of the cell 21 above the cell 21 on the stairs on the evacuation floor is set to "C (t) _1,1 + C (t) _1,2- C", and the number of the cell 21 on the evacuation floor _Let C (t) _2,1 be C. Otherwise, the number C (t) _2,2 of the cell 21 above the cell 21 on the evacuation floor is set to 0, and the number C (t) _2,1 of the cell 21 on the stairs on the evacuation floor is set to “C (t ) _1,1 + C (t) _1,2 . Subsequently, i = 2 is set in order to perform processing on the next floor (S106).

Subsequently, it is determined whether or not the i-th cell 21 of the stairs 20 is connected to the floor (S107). Specifically, it is determined whether or not F _i > 0 is satisfied. If it is determined that the i-th cell of the staircase is not connected to the floor (if it is determined that F _i > 0 is not satisfied), then the i-th cell 21 and the cell above it in the staircase In step 21, C _i (t) and C _{i + 1} (t) are calculated by marching from the upper cell 21 to the lower cell 21 (S108). As described below, new V _{i + 1} (t) and U _{i + 1} (t) are calculated using the column vectors V _i (t) and U _i (t) indicating the i-th transition state.

That is, in the i-th transition state, the vacant number “C−C (t) _{i, i} ” of the i-th cell 21 is the number of people C (t) _{i, i + 1} of the cell 21 (i + 1-th cell 21) above it. In the case of the following, the number C (t) _{i, i + 1} of the ( _{i + 1} ) th cell 21 is “C (t) _{i, i} + C (t) _{i, i + 1−} C”, and the number C ( t) _{Let i + 1, i} be C. In other cases, the number C (t) _{i + 1, i + 1} of _{the i + 1} cell 21 is set to 0, and the number C (t) _{i, i} of the _i cell 21 is set to “C (t) _{i, i} + C (t ) _{I, i + 1} ”. Subsequently, i = i + 1 is set to perform processing on the next floor (S109).

When it is determined in step S107 that the i-th cell 21 of the staircase is connected to the floor (when it is determined that F _i > 0 is satisfied), the i-th cell 21 is subsequently changed to the n-th cell. It is determined whether it is neither the cell 21 nor the (n-1) th cell 21 (S110).

If it is determined that the i-th cell 21 is neither the n-th cell 21 nor the (n-1) -th cell 21, then t starts to flow from the floor to the stairs in the i-th cell. Time t _i. It is determined whether or not it is greater than or _{equal to delay} (S111). Here, t is the time t _{i. When} the flow starts from the floor to the stairs in the i-th cell _. If it is determined that it is not greater than the _delay , evacuation has not yet started on the floor, and the same processing as S108 and S109 described above is performed.

In t111, t starts to flow from the floor to the stairs in the i-th cell _{. If} it is determined that it is greater than or equal to _delay, it is then determined whether there is an evacuee on the floor connected to the i-th cell of the stairs (S112). Specifically, it is determined whether or not F _i (t)> 0 is satisfied. If it is determined that there are no evacuees on the floor connected to the i-th cell of the stairs (if it is determined that F _i (t)> 0 is not satisfied), then the floor evacuation has already ended for that floor (S113). Further, the same processing as S108 and S109 described above is performed.

If it is determined that there are refugees on the floor connected to the i-th cell of the stairs (if it is determined that F _i (t)> 0), the refugees will enter the stairs from the floor horizontally. Considering this, the status of evacuees in each cell is calculated. Processing in that case will be described with reference to FIG.

First, the vacant number “C−C (t) _{i, i} ” of the i-th cell 21 is the maximum horizontal intruder 1.5B _i and the cell 21 from the top (i + 1-th cell 21) to the i-th cell 21. It is determined whether or not the number of people marching to C (t) _{i, i + 1} is greater than or equal to the sum (whether C (t) _{i, i + 1} ≧ C (t) _{i, i + 1} + 1.5B _i is satisfied) (S121). ). In the present embodiment, since the maximum number of cells matches the number of evacuees who can march between cells in one unit time, the determination by the above formula is possible.

When it is determined that C (t) _{i, i + 1} ≧ C (t) _{i, i + 1} + 1.5B _i is satisfied, the state of each cell is calculated in the case of state (a) described later (S122). If it is determined that C (t) _{i, i + 1} ≧ C (t) _{i, i + 1} + 1.5B _i is not satisfied, then, the vacant number “C−C (t) _{i, i} ” of the _i -th cell 21 Is the sum C (t) of the maximum horizontal intruder 1.5B multiplied by the horizontal approach rate (α) and the number of people marching from the top cell 21 (i + 1th cell 21) to the i th cell 21 _{It is} determined whether or not the sum is greater than or equal to _{i and i + 1} (whether or not C (t) _{i, i + 1} ≧ C (t) _{i, i + 1} + α1.5B _i is satisfied) (S123).

When it is determined that C (t) _{i, i + 1} ≧ C (t) _{i, i + 1} + α1.5B _i is satisfied, the state of each cell is calculated in the case of state (b) described later (S124). If it is determined that C (t) _{i, i + 1} ≧ C (t) _{i, i + 1} + α1.5B _i is not satisfied, then, the vacant number “C−C (t) _{i, i} ” of the _i -th cell 21 Is determined to be equal to or greater than the value obtained by multiplying the maximum horizontal intruder 1.5B by the horizontal approach rate (α) (whether or not C (t) _{i, i + 1} ≧ α1.5B _i is satisfied) (S125). .

When it is determined that C (t) _{i, i + 1} ≧ α1.5B _i is satisfied, the state of each cell is calculated in the case of state (c) described later (S126). If it is determined that C (t) _{i, i + 1} ≧ α1.5B _i is not satisfied, the state of each cell is calculated in the case of state (d) described later (S127).

As described above, new V _{i + 1} (t) and U _{i + 1} (t) are calculated using the column vectors V _i (t) and U _i (t) indicating the i-th transition state as described below.

This calculation will be described more specifically with reference to FIG. The entry from the floor to the stairs is prioritized by multiplying the maximum entry number of 1.5B _i per unit time by the horizontal entry rate α, and then the march from the cell above it on the stairs After that, if there is a vacant space from the floor to the stairs, the remaining portion will be entered.

As shown in FIG. 12A, in the state (a), only C (t) _{i, i + 1} is transferred from the upper cell 21 (i + 1th cell 21) to the lower cell (ith cell 21). Since march, C (t) _{i + 1, i + 1 is set} to 0. In addition, to the stairs of the cell from the floor, so to enter only 1.5B _i at the maximum, the _{F (t) i + 1,} i, and _{max [F (t) i,} i -1.5B i, 0]. Here, max is a function having the maximum value among all the variable values as a function value. Since the i-th cell 21 has an inflow from each of the above cells, C (t) _{i + 1, i} is set to C (t) _{i, i} + C (t) _{i, i + 1} + 1.5B _i . The i-th cell 21 is not filled with evacuees.

As shown in FIG. 12B, in the state (b), only C (t) _{i, i + 1} is transferred from the upper cell 21 (i + 1th cell 21) to the lower cell (ith cell 21). Since march, C (t) _{i + 1, i + 1 is set} to 0. In addition, from the floor to the stair cell, “C− (C (t) _{i, i} + C (t)”, which is the number of people (≧ α1.5B _i ) obtained by subtracting the number of cells directly above the “vacant” of the stair cell. _{i, i + 1} ) ”, so that F (t) _{i + 1, i} is changed to max [F (t) _{i, i} − (C− (C (t) _{i, i} + C (t) _{i, i + 1} )), 0]. Since the i-th cell 21 is filled with evacuees by the inflow from each of the above cells, it is assumed as C.

As shown in FIG. 12C, in the case of the state (c), the upper cell 21 (i + 1st cell 21) is changed to C- (C (t) _i from the lower cell (ith cell 21). _{, I} + α1.5B _i ), C (t) _{i + 1, i + 1 is set} to C (t) _{i, i + 1} − (C− (C (t) _{i, i} + α1.5B _i )). Further, since a maximum of α1.5B _{i is} entered from the floor to the staircase cell, F (t) _{i + 1, i} is set to max [F (t) _{i, i} −α1.5B _i , 0]. Since the i-th cell 21 is filled with evacuees by the inflow from each of the above cells, it is assumed as C.

As shown in FIG. 12 (d), in the state (d), all evacuees remain without marching from the upper cell 21 (i + 1th cell 21) to the lower cell (ith cell 21). Therefore, C (t) _{i + 1, i + 1} is set as C (t) _{i, i + 1} . Further, since the empty cell C (t) _{i, i} of the i-th cell 21 enters from the floor to the staircase cell at the maximum, F (t) _{i + 1, i} is set to max [F (t) _{i, i i−} (C−C (t) _{i, i} ), 0]. Since the i-th cell 21 is filled with evacuees by the inflow from each of the above cells, it is assumed as C.

  As shown in FIG. 10, when the transition state is calculated in any one of S112, S124, S126, and S127, i = i + 1 is set to perform processing on the next floor (S128). Subsequently, as shown in FIGS. 9 and 10, the processes after S107 described above are repeated.

The processing when it is determined in S110 of FIG. 9 that the i-th cell 21 is neither the n-th cell 21 nor the (n−1) -th cell 21 will be described with reference to FIG. In this case, the time t _{n. At which} t starts to flow from the floor to the stairs in the nth cell _. It is determined whether or not it is greater than or _{equal to delay} (S131). the time t when t begins to flow from the floor to the stairs in the nth cell _{. If} it is determined that it is greater than or equal to _delay, it is then determined whether there is an evacuee on the floor connected to the nth cell of the stairs (S132). Specifically, it is determined whether or not F _n (t)> 0 is satisfied.

In S131, t starts to flow from the floor to the stairs in the nth cell _{tn. When} it is determined that it is not greater than _delay , and when it is determined in S132 that there are no evacuees on the floor connected to the nth cell of the stairs (when it is determined that F _i (t)> 0 is not satisfied) Then, calculation of the ( _n−1 ) th and nth cells C (t) _{n, n−1} , C (t) _{n, n in} the staircase is performed (S133). Specifically, using the column vectors V _n−1 (t) and U _n−1 (t) indicating the ( _n−1 ) th transition state, new V _n (t), U _n (t) is calculated. Subsequently, the processing returns to the processing of S102 shown in FIG. 9, t = t + 1, and the processing is repeated.

On the other hand, in S132, when it is determined that there is an evacuee on the floor connected to the nth cell of the stairs (when it is determined that F _i (t)> 0), the (n−1) th transition Using column vectors V _n−1 (t) and U _n−1 (t) indicating the states, new V _n (t) and U _n (t) are calculated as follows.

First, similarly to the process of S108, calculation of the ( _n-1 ) _th cell C (t) _{n, n-1} in the staircase is performed (S134). In addition, the n-th cell C (t) _n, the empty “2C-C (t) _{n−1, n−1−} C (t) _{n−1, n} ” of the n-cell is the n-th cell C (t). _It is determined whether or not the entry from the floor to _{n, n} is 1.5B _i or more (S135). In accordance with the determination, C (t) _{n, n} and F (t) _{n, n} are calculated according to the above table (S136, S137).

Subsequently, it is determined whether or not the number of accommodated cells in each cell after the transition state calculation is 0 (S139). That is, it is determined whether C _{n, i} (t) = 0 (i = 1 to n) and F _i (t) = 0 (i = 1 to n). If it is determined that C _{n, i} (t) = 0 (i = 1 to n) and F _i (t) = 0 (i = 1 to n) are satisfied, it is determined that the evacuation from the building has ended. Thus, the process of calculating the transition state of the evacuee accommodation state of the cell by the transition state calculation unit 13 and the process of determining completion of evacuation by the evacuation end determination unit 14 are completed.

On the other hand, if it is determined that C _{n, i} (t) = 0 (i = 1 to n) and F _i (t) = 0 (i = 1 to n) are not satisfied, the processing of S102 shown in FIG. Returning to t = t + 1, the process is repeated. The above is the process for calculating the transition state of the evacuee accommodation state of the cell by the transition state calculation unit 13 and the process for determining completion of evacuation by the evacuation end determination unit 14.

  As described above, according to the present embodiment, the evacuation time is predicted in consideration of the mutual influences of advancement from each floor to the stairs that are evacuation routes and movement on the stairs. Therefore, according to the present embodiment, it is possible to obtain a prediction result with higher accuracy than the simple methods as shown in Non-Patent Documents 2 and 3. Further, in the evacuation time prediction device according to the present invention, the evacuation time is predicted by the number of persons included in the cells 21 and 31 without individually simulating each refugee. Compared with a simulation method using a multi-agent, a prediction result can be obtained in a realistic time even in a large building. That is, according to this embodiment, in the prediction of the evacuation time from a multi-level building, a highly accurate prediction result can be obtained in a shorter time compared to the conventional method.

  Evacuation is an extremely difficult field to reproduce through experiments. The evacuation start time, etc. on the fire floor and non-fire floor, depending on the situation of the site, such as the characteristics of the refugee, the degree of recognition of the facility, the location of the fire and the scale of the fire, etc. The swing is large and it is difficult to obtain laws that contribute to evacuation calculations. According to the present embodiment, these are treated as parameters, and merging on the stairs and evacuation start can be arbitrarily set. While changing the parameters, it is possible to check against the safest choice and the intuition that “this kind of movement is possible / impossible”. Thereby, in design practice, for example, the safety of various patterns in which the initial state, the horizontal approach rate, and the like are changed can be confirmed and used for evacuation safety design.

An example in which the evacuation time is actually calculated by the method according to the embodiment described above will be shown. Calculation was performed under the following conditions. In addition, the following calculation can be used to consider how horizontal is actually performed when merging on the stairs. A value obtained by dividing the number of intruders per unit time (one second) by the entrance to the stairs from the floor is called the actual flow coefficient _Nr .
N _r × entrance width = actual number of intruders actual flow coefficient N _r is a variable (function) determined by the floor and the elapsed time.
・ Number of floors …… 9th floor ・ Standard floor number of persons …… 125 people ・ Stair width …… 2m
・ Floor → Stair flow coefficient ... 1.5 people / msec ・ Floor → Stairway entrance width B _i ...... 0.8 m
· Stairs outlet at the evacuation floor flow coefficient ...... 1.3 people / m stairs outlet width in seconds and evacuation floor _B 1 ...... 0.8 m (stair width less than), 2m (greater stairs width)
・ Horizontal approach rate α …… 0%, 20%, 50%, 100%
・ Evacuation start time: Simultaneous evacuation (simultaneous evacuation from time 0)

The calculation results are shown in FIG. In each graph in FIG. 13, the vertical axis represents the floor, and the horizontal axis shows the actual flow coefficient N _r. When α = 0% and B ₁ = 0.8 m, immediately after the start of evacuation, _Nr has a normal flow coefficient of 1.5, but then changes to 1.3. Since the exit at the evacuation floor is 1.3 × the door width, N _r = 1.3 even at the entrance floor. On the other floors, N _r = 0. When the stairs are clogged, marching takes precedence over approach. The approach phenomenon gradually moves down from the top floor to the lower floor.

When α = 100% and B ₁ = 0.8 m, immediately after the start of evacuation is the same as when α = 0%. After that, since the approach has priority over the march in the stairs, the process is completely reversed. The entry moves up from the second floor to the upper floor. When α = 20% and B ₁ = 0.8 m, immediately after the start of evacuation, N _r has a normal flow coefficient of 1.5, but thereafter, N _{r for the} lower four layers changes at 0.3. . On the first floor, _Nr is a small value of about 0.1. When α = 50% and B ₁ = 0.8 m, immediately after the start of evacuation, N _r has a normal flow coefficient of 1.5, but thereafter, N _r for one layer in the lower layer is 0.75 (1. 5 × 50%), the _{N r} downstairs thereon to remain at slightly lower value (0.55) degrees.

When the exit door width at the evacuation floor the same 2m stairs width, although _{N r} is the normal flow coefficient 1.5, then the _{N r} of four layers of low-rise section 0.75 (1.5 × 50%). On the first floor, _Nr is a small value of about 0.3. At the end of the evacuation, the number of people on the stairs decreases, and N _r = 1.5 again at the high floor. In other words, the entry is 5 layers from the lower layer.

Thus, even if α is a constant, the actual flow coefficient _Nr changes with the elapsed time and also with the floor. If the exit width to the stairs is equal to the exit width at the evacuation floor, you can enter with the actual flow coefficient N _r = 1.5 immediately after the start of evacuation. Note that this is an upper limit.

"Akihiko Kitagogo, Kosuke Kubo, Masuki Murasaki" Experimental study on the merging of two crowds on the stairs ", Architectural Institute of Japan, Planning pp. 37-43, December 1985 ”,“ Based on experiments, the flow rate confluence ratio in the steady stage is about 60% (hereinafter referred to as “Kitago”) ”. The relationship between the flow rate confluence ratio and the horizontal approach rate is as follows.

Assuming that φ is 60%, B = 1.2 m, B _i = 0.8 m, α≈65% is obtained from the above equation. When φ is a little smaller and 50%, α≈52% is obtained.

Another case study is shown. The following example shows the impact on the evacuation time of setting up a large number of people in the skyscraper, such as the floor where the conference rooms are concentrated and the floor where the employee cafeteria is located. Calculation was performed under the following conditions.
・ Number of floors: 20 floors ・ Standard floor number of people, area A _floor: 375 people, 3000 m ²
・ Singular floor (the number of people in the building) …… 2100 ・ Singular floor (fire floor) …… 3rd floor, 10th floor, 18th floor ・ Stairs number …… 2 locations ・ Stair width, area …… 1.2 m / location, 15 m ² / floor / location / floor → Stairway entrance width B _i ...... 0.8 m
・ Stairway exit width B _{1 on} the evacuation floor 1.2 m (total exit width to the outside 10 m)
・ Flow coefficient …… Same as above ・ Horizontal approach rate α …… Parameter (0-100%)
・ Evacuation start time ... Simultaneous evacuation ... Simultaneous evacuation starts from time 0
Timed evacuation ... Non-fire floor is fire floor (evacuation starts at time 0)

Using the delayed horizontal approach rate α in (seconds) and the evacuation start time as parameters, floor evacuation times and evacuation times for the third, tenth and 18th floors were obtained.

  FIG. 14 shows a graph in the case where there is a singular floor on the 10th floor, α is 60%, and evacuation is performed simultaneously. In addition, FIG. 15 shows the number of evacuees on each floor and stairs when 300 seconds have elapsed, 600 seconds have elapsed, 900 seconds have elapsed, and 1500 seconds have elapsed. The part shown in dark color indicates that the evacuees are being accommodated at that time. FIG. 16 shows a graph of the floor evacuation time on the third floor, the 10th floor, and the 18th floor by changing α from 0 to 100%.

  In the simultaneous evacuation, when the horizontal approach rate α is 30% or less, the third floor takes the most time for the floor evacuation, and then the 10th floor. When α was 50% or more, the 18th floor took the longest time for floor evacuation, followed by the 10th floor. In this way, the aspect is reversed depending on the horizontal approach rate.

  The same applies to time evacuation. Since all other floors start evacuation 330 seconds after the start of evacuation on the singular floor, the actual flow coefficient Nr to the stairs on the singular floor decreases (the merging speed decreases). This time, the number of singular floors was made extremely large compared to the general floors, but if the difference in the number of people from the general floors narrows, the entry of the singular floor may end before the general floor that is a non-fire floor is there.

Regarding merging at the stairs, “Yuki Sano, Naoyuki Takeichi, Ken Kimura, Yoshifumi Omiya, Katsuyuki Yoshida, Hitoshi Watanabe“ Merging characteristics of the crowds in the staircase assuming evacuation of high-rise buildings ” pp. 51-56, 2005 ”,“ Merge along flow ”and“ Merge in reverse direction ”are compared, and forward merge along flow is easy. In addition, if the door to the staircase opens to nearly 90 degrees, the flow coefficient of the person entering the staircase increases. The flow coefficient is a real flow coefficient N _r defined above. In this way, the staircase structure such as staircases and doors, and the direction of the stairway when evacuating, is considered to have a significant effect on the horizontal approach rate. From the above formula and the like, it is estimated that the horizontal approach rate α takes a range of about 50 to 70% depending on the staircase structure. According to this, the floor evacuation time becomes longer when the anomalous floor is at a higher level. The anomalous floor should be placed in the lower part as much as possible. In addition, the floor evacuation time of the special floor in the high-rise part is not easily affected by the horizontal approach rate (except when the horizontal approach rate is extremely small).

  It can be read from the graph shown in FIG. 16 that the floor evacuation time is prolonged when the flow coefficient of horizontal approach is small in a complex public facility where there are floors containing many vulnerable people in the lower floors. It is necessary to make it easier to open the doors of the stairs and to install low-level dedicated stairs that do not merge with the upper floors.

A comparison of the calculation results of the staircase simulator and the evacuation safety verification method (Non-Patent Documents 2 and 3) according to this embodiment is shown in the following table.

The final evacuation time for the entire building as a result of the notification was large (10m) at the end of the evacuation floor, which was slightly shorter than the staircase simulator. On the other hand, the floor evacuation was 269 (minutes) longer than the whole building evacuation. This is because the effective flow coefficient of the singular floor becomes extremely small in calculation.

  As described above, an appropriate calculation result is obtained by an actual calculation example. Moreover, the knowledge regarding the influence on the evacuation time by the horizontal approach rate and whether the floor (singular floor) with many evacuees exists in a low-rise part or a high-rise part can be obtained.

  In the above-described example, the building does not have an underground floor (basement). However, the present invention can be applied even when the basement has a basement and joins the ground stairs on the evacuation floor. At the cell connected to the evacuation exit, the merging from the ground floor and the basement floor is divided by the number of people entering from the cell directly above and below, and the exit to the outdoor area is calculated at each time.

  DESCRIPTION OF SYMBOLS 10 ... Evacuation time prediction apparatus, 11 ... Model production | generation part, 12 ... Initial state input part, 13 ... Transition state calculation part, 14 ... Evacuation end determination part, 15 ... Output part.

Claims (5)

An evacuation time prediction device for predicting an evacuation time from a multi-level building having stairs,
A model for generating a model of the building in which the stairs are one or more continuous cells for each floor accommodating evacuees, and each floor is connected to one of the cells corresponding to the floor. Generating means;
An initial state input means for inputting an initial state of the number of evacuees in each cell and each floor in the building model generated by the model generating means;
Based on the model of the building generated by the model generation means, the state of the number of evacuees in each cell per unit time is input by the initial state input means in order from the cell with the closest exit. Either the initial state or the previous time state, the state of inflow from the cell one level above, the state of advancement of the refugee from the floor, and the refugee from the floor to the stairs cell with respect to the inflow Transition state calculating means for calculating based on a horizontal approach rate indicating a rate of entering;
When the state of the number of evacuees in each cell calculated by the transition state calculation means indicates whether or not the evacuation has been completed, and it is determined that the evacuation has not been completed, the transition state calculation means Evacuation end determination means for calculating the number of evacuees in each cell at the next time,
An output means for outputting the evacuation time according to the number of unit times calculated by the transition state calculation means when it is determined that the evacuation has been completed by the evacuation end determination means;
Evacuation time prediction device comprising: The initial state input means also inputs information indicating the timing at which evacuation is started at least in any of the floors,
The transition state calculation means calculates the state of the number of evacuees in each cell per unit time based also on the information indicating the timing input by the initial state input means.
The evacuation time prediction apparatus according to claim 1.   The said model production | generation means determines the number of said cells based on the number of people who can flow in between cells per said unit time, The evacuation time prediction apparatus of Claim 1 or 2 characterized by the above-mentioned.   The said output means outputs the information which shows the state of the number of the refugees of each said cell of each said unit time calculated by the said transition state calculation means, The Claim 1 characterized by the above-mentioned. The evacuation time prediction device described. An evacuation time prediction method that is an operation method of an evacuation time prediction device that predicts an evacuation time from a multi-level building having stairs,
A model for generating a model of the building in which the stairs are one or more continuous cells for each floor accommodating evacuees, and each floor is connected to one of the cells corresponding to the floor. Generation step;
An initial state input step of inputting an initial state of the number of evacuees in each cell and each floor in the building model generated in the model generation step;
Based on the model of the building generated in the model generation step, the state of the number of evacuees in each cell per unit time is input in the initial state input step in order from the cell with the closest exit. Either the initial state or the previous time state, the state of inflow from the cell one level above, the state of advancement of the refugee from the floor, and the refugee from the floor to the stairs cell with respect to the inflow A transition state calculating step for calculating based on a horizontal approach rate indicating a rate of entering;
When the state of the number of evacuees in each cell calculated in the transition state calculation step indicates whether or not the evacuation has been completed and it is determined that the evacuation has not been completed, the transition state calculation step Evacuation end determination step for calculating the number of evacuees in each cell at the next time at
An output step of outputting the evacuation time according to the number of unit times calculated in the transition state calculation step when it is determined that the evacuation is completed in the evacuation end determination step;
Evacuation time prediction method including