JP2016200973A - Estimation method for od traffic of pedestrian - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide an estimation method for OD traffic of pedestrians in which OD traffic and a congestion state of pedestrians can be grasped in a space having a plurality of gateways in real time based upon cross-sectional traffic.SOLUTION: An estimation method for OD traffic of pedestrians includes: a measurement value acquisition step of acquiring the number of coming-in persons and the number of going-out persons at each gateway which are measured by a traffic measurement device installed at each gateway; a traffic estimation step of estimating OD traffic at each gateway by multiplying the number of coming-in persons by an OD branch rate of the gateway; a flow state prediction step of probabilistically simulating an average flow state of pedestrians based upon OD traffic estimated for each gateway; and a branch rate correction step of finding a correction value for an OD branch rate for each gateway in predetermined cycles and correcting the OD branch rate so that the difference between the number of going-out persons at each gateway and an outflow prediction value at each gateway becomes small.SELECTED DRAWING: Figure 1

Description

  The present invention relates to a method for estimating a pedestrian's OD traffic volume in which a pedestrian's OD traffic volume and a congestion situation in a space having a plurality of entrances and the like can be grasped in real time based on a cross-sectional traffic volume.

  OD traffic volume of pedestrians in buildings and other spaces (Origin Destination traffic volume: traffic volume between the pedestrian's starting point (starting point) and ending point (destination point)) and the congestion of the space in real time In order to grasp the temporal change, it is necessary to install a sensor such as a laser scanner for detecting the position of the pedestrian and to integrate and process the obtained position information of the pedestrian.

  Conventionally, for the estimation of traffic volume (or traffic volume), the OD table that gives the traffic volume between regions in a matrix is divided, and the split distribution that sequentially loads the OD traffic volume on the shortest route while updating the link cost. Has been widely used. However, with split allocation, there is a problem that the estimation results differ depending on the split conditions, and studies are underway on estimation methods using balanced allocation.

  For example, it is an estimation method using such equilibrium distribution, and a total data generation method that makes it possible to total various amounts depending on traffic after distribution in a stochastic user equilibrium distribution for a predetermined target Is disclosed (for example, see Patent Document 1).

  Further, for example, a method for estimating the OD traffic volume of a vehicle in a single road section of an expressway using a macro traffic simulation model is disclosed (for example, see Non-Patent Document 1).

Japanese Patent No. 3931190

Takashi Nakajo, Hironori Suzuki: Online estimation of OD traffic volume incorporating macro traffic simulation model, Proceedings of traffic engineering research presentation 26, 233-236 (2006)

  However, the total data generation method of Patent Document 1 estimates a route when a pedestrian moves from one node to another node, and the pedestrian in a space such as a building having a plurality of entrances and exits. There is a problem that the OD traffic volume cannot be estimated based on the cross-section traffic volume.

  In addition, the online estimation of the OD traffic volume in Non-Patent Document 1 estimates the OD traffic volume based on the inflow traffic volume. However, since the target vehicle is an automobile traveling on a highway, for example, Pedestrian OD traffic volume that cannot flow out of the entrance or reverse on the highway, is expected to flow out of a certain entrance and exit through the same entrance and exit. There was a problem that it was impossible to estimate.

  The subject of this invention is providing the estimation method of the OD traffic volume of a pedestrian which can grasp | ascertain the OD traffic volume and congestion condition of a pedestrian in the space which has a some entrance / exit in real time based on cross-sectional traffic volume. .

In order to achieve the above object, the present invention provides:
A method for estimating the OD traffic volume of a pedestrian in a space having a plurality of doorways,
A measurement value acquisition step for acquiring the number of inflows and outflows for each of the entrances and exits measured by the traffic measuring device installed for each of the entrances and exits;
A traffic volume estimating step of multiplying the number of inflows by the OD branching rate for each doorway to estimate the amount of OD traffic for each doorway;
A flow situation prediction step for stochastically simulating an average flow situation of pedestrians based on the estimated OD traffic volume for each doorway to obtain an outflow prediction value for each doorway;
A branch that corrects the OD branching rate by obtaining a correction value of the OD branching rate for each doorway so that a difference between the number of outflowers at each doorway and the predicted outflow value for each doorway becomes small at a predetermined cycle. Rate correction step;
Is included.

  OD branching rate is multiplied by the measured number of inflows to estimate the amount of OD traffic, and the outflow predicted value obtained by stochastic simulation of the average situation of pedestrians and the number of outflows at each entrance and exit By calculating the correction value of the OD branching rate for each entrance and exit at a predetermined cycle so as to reduce the difference, and correcting the OD branching rate, the OD trafficing amount estimated based on the number of inflows as the cross-sectional traffic amount is actually The amount of OD traffic can be followed, and the OD traffic amount of a pedestrian in a space having a plurality of doorways can be accurately grasped in real time.

Preferably, in the branching rate correction step, when the value obtained by subtracting the average value of the difference from the difference is positive, the correction value is negative, and the value obtained by subtracting the average value of the difference from the difference is negative. In this case, the correction value is made positive.
When the value obtained by subtracting the average value of the difference from the difference is positive, the correction value is negative. When the value obtained by subtracting the average value of the difference from the difference is negative, the correction value is set positive. It can be calculated approximately, and the estimated OD traffic volume can be quickly followed by the actual OD traffic volume, and the OD traffic volume of pedestrians in a space with multiple doorways can be accurately grasped in real time. can do.

Preferably, in the branching rate correction step, the correction value of the entrance / exit having a large difference is increased.
By preferentially increasing the correction value of the entrance / exit with a large difference, the correction value can be obtained approximately and efficiently, and the estimated OD traffic volume can be made to quickly follow the actual OD traffic volume, The amount of OD traffic of a pedestrian in a space having a plurality of doorways can be grasped more accurately in real time.

Desirably, in the branching rate correction step, the correction value of the entrance / exit with a large number of inflowing persons is increased.
By preferentially increasing the correction value at the entrance / exit with a large number of inflows, the correction value can be obtained efficiently and approximately, and the estimated OD traffic volume can quickly follow the actual OD traffic volume. The OD traffic volume of a pedestrian in a space having a plurality of doorways can be grasped more accurately in real time.

Preferably, in the branching rate correction step, the value of the correction value for each of the doorways is adjusted so that the total of the correction values becomes zero.
By adjusting the value of the correction value for each doorway so that the total correction value becomes zero, the consistency of the estimated OD traffic volume can be maintained, and the estimated OD traffic volume can be changed to the actual OD traffic volume. Therefore, it is possible to grasp the OD traffic amount of a pedestrian in a space having a plurality of doorways in real time.

Preferably, a map is created for acquiring the number of pedestrians in each mesh obtained by dividing the space obtained by the simulation of the flow state prediction step, and creating a map showing the distribution of the number of pedestrians in the space Steps are included.
By creating a map that shows the distribution of the number of pedestrians in each mesh in the space, the degree of congestion of the number of pedestrians in the space can be visually understood, so pedestrian congestion in a space with multiple entrances It is possible to grasp the situation in real time and grasp temporal changes.

  According to the present invention, the OD branching rate is multiplied by the measured number of inflows to estimate the amount of OD traffic, the predicted outflow value obtained by probabilistic simulation of the average situation of pedestrians, the entrance and exit Estimated based on the number of inflows, which is the cross-sectional traffic volume, by obtaining a correction value for the OD branching rate at each entrance and exit and correcting the OD branching rate so that the difference from the number of people flowing out is small. Since the OD traffic volume can be made to follow the actual OD traffic volume, the OD traffic volume of a pedestrian in a space having a plurality of entrances can be grasped in real time.

  In addition, by creating a map showing the distribution of pedestrians for each mesh obtained by simulation, the degree of crowding of the number of pedestrians in the space can be seen visually, so pedestrians in spaces with multiple entrances It is possible to grasp the congestion situation in real time and grasp temporal changes.

It is a schematic block diagram which shows an example of a structure of the estimation apparatus of OD traffic amount which concerns on this Embodiment. It is a top view which shows an example of the space which has the some entrance / exit which estimates OD traffic volume. It is a flowchart which shows an example of operation | movement of an estimation apparatus. It is explanatory drawing which shows an example of the space which has the some entrance and exit divided | segmented into the mesh in a flow condition prediction part. It is explanatory drawing which shows an example of deterministic simulation and probabilistic simulation. It is explanatory drawing which shows an example of the map which shows distribution of the number of pedestrians in the space which has a some entrance / exit.

(Embodiment)
[1. Description of configuration]
Hereinafter, a pedestrian OD traffic amount estimation method and estimation device according to an embodiment of the present invention will be described in detail with reference to the drawings. However, the scope of the invention is not limited to the illustrated examples.

  The configuration of the pedestrian OD traffic amount estimating apparatus according to the embodiment of the present invention will be described with reference to FIG. FIG. 1 is a schematic configuration diagram showing the function of the pedestrian OD traffic estimation device 100 as a block diagram.

  As shown in FIG. 1, the pedestrian OD traffic amount estimation device 100 (hereinafter simply referred to as device 100) includes an input unit 1, a storage unit 2, a display unit 3, a communication unit 4, an operation unit 5, and a measurement. It has a value acquisition unit 6, a traffic amount estimation unit 7, a flow situation prediction unit 8, a branching rate correction unit 9, and a map creation unit 10.

  The functions of the measurement value acquisition unit 6, the traffic amount estimation unit 7, the flow condition prediction unit 8, the branching rate correction unit 9, and the map creation unit 10 are realized by the calculation control unit 50. Specifically, the arithmetic control unit 50 includes a central processing unit (CPU), a read only memory (ROM), a random access memory (RAM), and the like, and is stored in a ROM or the like developed in the work area of the RAM. In cooperation with the various program data and the CPU, the input unit 1, the storage unit 2, the display unit 3, the communication unit 4, and the operation unit 5 are collectively controlled, and the measured value acquisition unit 6, the traffic amount estimation unit 7 The processes in the flow state prediction unit 8, the branching rate correction unit 9, and the map creation unit 10 are executed to realize the functions of the respective units.

The input unit 1 is controlled by the arithmetic control unit 50, for example, for each pedestrian entrance measured by a traffic measuring device installed at each entrance of a space having a plurality of entrances, from a device outside the device 100. Get the number of inflows and outflows.
The input unit 1 is an input port such as a USB (Universal Serial Bus) connector for inputting information. The traffic volume measuring device is an apparatus that extracts the traffic volume (number of people) of pedestrians from image information acquired by a ticket gate, a camera, an optical sensor, or the like installed at the entrance of a space.

The storage unit 2 stores information in a readable and writable manner from the calculation control unit 50 (measurement value acquisition unit 6, traffic amount estimation unit 7, flow state prediction unit 8, branch rate correction unit 9, and map creation unit 10). For example, the storage unit 2 is an HDD (Hard Disk Drive), an SSD (Solid State Drive), a semiconductor memory, or the like, and stores information such as data collected by the input unit 1. For example, the storage unit 2 may be an HDD (Hard Disk Drive), a semiconductor memory, or the like.
Specifically, the storage unit 2 stores the OD branching rate of the pedestrian for each doorway used when estimating the OD traffic amount for the pedestrian for each doorway, and is obtained from the traffic amount measuring device. The number of inflows and outflows of pedestrians at each doorway, and the predicted outflow values of pedestrians at each doorway acquired from the flow state prediction unit 8 are stored as appropriate.

The display unit 3 is information based on display control signals output from the calculation control unit 50 (measurement value acquisition unit 6, traffic amount estimation unit 7, flow state prediction unit 8, branch rate correction unit 9, and map creation unit 10). And display images on the display screen.
For example, the display unit 3 is a CRT (Cathode Ray Tube), an LCD (Liquid Crystal Display), an FPD (Flat Panel Display) using an organic EL (Organic Electro-Luminescence) element, or the like.
The display unit 3 may be a touch panel formed integrally with the operation unit 5 such as a smartphone or a tablet terminal.

  The communication unit 4 is connected to a network such as the Internet or wireless communication under the control of the arithmetic control unit 50, and receives the number of inflows and outflows of pedestrians at each entrance / exit measured by the traffic measurement device outside the device 100. And store it in the storage unit 2.

The operation unit 5 operates a display screen displayed on the display unit 3 under the control of the calculation control unit 50. For example, a keyboard and a mouse for operating the display screen.
The operation unit 5 may be a touch panel formed integrally with the display unit 3 such as a smartphone or a tablet terminal.

  The measured value acquisition unit 6 controls the input unit 1 or the communication unit 4 to walk from the outside of the device 100 for each entrance and exit measured by the traffic measuring devices respectively installed at the entrance and exit of the space having a plurality of entrances and exits. The number of inflows and outflows of the person is acquired and stored in the storage unit 2. For example, the measurement value acquisition unit 6 acquires the number of inflows and outflows of pedestrians at each entrance and exit measured by the traffic measurement device at intervals of 1 second and stores them in the storage unit 2.

  The traffic amount estimation unit 7 reads the OD branching rate of pedestrians for each entrance and exit stored in the storage unit 2 and multiplies the number of pedestrian inflows for each entrance and obtained from the traffic amount measuring device. Estimate the amount of pedestrian traffic. In addition, the traffic amount estimation unit 7 stores the estimated OD traffic amount of the pedestrian for each entrance / exit in the storage unit 2 and appropriately displays it on the display unit 3.

  As will be described later, the flow situation prediction unit 8 probabilistically simulates the average flow situation of pedestrians based on the estimated OD traffic amount of pedestrians at each entrance and exit, and predicts the outflow of pedestrians at each entrance and exit. Find the value.

The branching rate correction unit 9 is a predetermined period, and the difference between the number of pedestrian outflows per entrance and exit measured by the traffic measuring device and the predicted outflow value of pedestrians per entrance and exit simulated by the flow state prediction unit 8 The correction value of the OD branching rate of the pedestrian for each entrance is obtained so that the OD is reduced, and the OD branching rate of the pedestrian for each entrance is corrected using the correction value.
For example, the branching rate correction unit 9 obtains a correction value of the OD branching rate of the pedestrian for each entrance and exit so that the difference between the number of pedestrians outflowing for each entrance and the predicted outflow value becomes small at an interval of 1 minute. The OD branching rate of the pedestrian at each entrance is corrected using the correction value.

  The map creation unit 10 acquires the number of pedestrians in each mesh obtained by dividing the space obtained by the simulation of the flow state prediction unit 8, and creates a map indicating the distribution of the number of pedestrians in the space. For example, based on the number of pedestrians in each mesh obtained by the simulation of the flow situation prediction unit 8, a map showing the distribution of the number of pedestrians by obtaining the distribution of pedestrians in the space is created, and the display unit 3 is displayed.

[2. Explanation of operation of estimation device]
Here, a specific operation of the estimation device 100 for the pedestrian OD traffic amount according to the embodiment of the present invention will be described in detail with reference to FIGS.
FIG. 2 is a plan view showing an example of a space having a plurality of doorways for estimating the OD traffic volume of a pedestrian, and the space SP21 has four doorways ET21, ET22, ET23, and ET24.

  A ticket gate GT21 is installed at the entrance / exit ET21, which measures the amount of traffic (outflow and inflow) of pedestrians, which is a traffic volume measuring device. Devices EQ21, EQ22, and EQ23 that extract the amount of pedestrian traffic (number of people outflow and number of people inflow) from image information acquired by a camera or an optical sensor as a device are installed.

[2-1. Explanation of operation]
As illustrated in the flowchart of FIG. 3, the measurement value acquisition unit 6 controls the input unit 1 or the communication unit 4 to measure the traffic amount measurement device (GT21, EQ21) installed for each entrance (ET21 to ET24) of the space SP21. To EQ23), the number of inflows and outflows of pedestrians measured are acquired and stored in the storage unit 2 (step S31: measurement value acquisition step).

  Next, the traffic amount estimation unit 7 reads the OD branching rate of the pedestrian for each entrance and exit stored in the storage unit 2 and multiplies the number of pedestrian inflows for each entrance and obtained by the measurement value acquisition unit 6. Then, the OD traffic volume of the pedestrian at each doorway is estimated (step S32: traffic volume estimation step). Further, the traffic amount estimation unit 7 stores the estimated OD traffic amount of the pedestrian for each entrance / exit in the storage unit 2 and appropriately displays the pedestrian traffic amount of the pedestrian on the display unit 3 as necessary.

  Then, the flow state prediction unit 8 probabilistically simulates the average flow state of the pedestrian based on the OD traffic amount of the pedestrian for each doorway estimated by the traffic amount estimation unit 7 (step S33: flow state prediction step) ).

[2-2. Explanation of simulation]
Here, the stochastic simulation in the flow state prediction unit 8 will be described in detail. The flow situation prediction unit 8 divides the space SP21 having a plurality of entrances and exits shown in FIG. 2 into a plurality of meshes (for example, ME41, ME42, ME43, ME44, etc.) as shown in FIG. A simulation is made of the flow situation or the like of which gateway (ET21 to ET24) flows out while moving to which mesh the pedestrian flowing in from the gateway ET21.

  In general, when simulating the movement of a pedestrian that has flowed into such a plurality of meshes, for example, as shown in FIG. 5A, when the pedestrian moves from the mesh ME51 to another mesh, the walking Since the user cannot be divided, the user moves to one mesh ME53 as shown in FIG. Such a simulation is called a deterministic simulation, and it takes time for the calculation for determining which mesh the pedestrian is moved to, and is inferior in real time.

On the other hand, for example, as shown in FIG. 5 (c), when a pedestrian moves from the mesh ME51 to another mesh, the pedestrian himself is divided probabilistically, as shown in FIG. 5 (d). The meshes ME52 to ME54 are divided and moved.
Specifically, “0.2” pedestrians move to mesh ME52 and ME54, respectively, and “0.6” pedestrian moves to mesh ME53. This kind of simulation is called probabilistic simulation, and the average movement situation of pedestrians is probabilistically predicted based on the OD traffic volume of pedestrians at each entrance and exit. Is excellent.

  For this reason, in the pedestrian OD traffic estimation device 100 according to the embodiment of the present invention, the flow situation prediction unit 8 uses the probability to grasp the pedestrian OD traffic in a space having a plurality of doorways in real time. Simulation (probabilistic prediction of average pedestrian movement based on pedestrian OD traffic at each doorway).

  Here, various settings of the space, time, pedestrian attributes and state for performing the stochastic simulation, and the flow map obtained by the stochastic simulation and the state change of the flow map will be described in detail. In addition, the stochastic simulation in the flow state prediction unit 8 is performed by calculation using general spreadsheet software.

と表記する。複数の出入口を有する空間内の位置は単位Δｘの離散した値をとり、その間の距離は扱わない。また、Δｘは、混雑していない時の歩行速度に対する時間の刻み幅との条件を満足すれば自由に設定可能である。 [2-2-1. Space settings]
A space having a plurality of entrances to be simulated is equally divided into sections (one mesh) of vertical and horizontal length Δx, and a mesh with the upper left as the origin is

Is written. The position in the space having a plurality of doorways takes discrete values of the unit Δx, and does not handle the distance between them. Further, Δx can be freely set as long as the condition of the time increment with respect to the walking speed when not crowded is satisfied.

と表記する。時刻は離散的な値をとり、その間の時刻は扱わない。また、Δｔについても、混雑していない時の歩行速度に対するメッシュサイズとの条件を満足すれば自由に設定可能である。 [2-2-2. Time setting]
Based on a certain time, the time is expressed as the number of steps in increments of time Δt.

Is written. The time takes discrete values, and the time between them is not handled. Also, Δt can be freely set as long as the condition of the mesh size with respect to the walking speed when not crowded is satisfied.

と表記する。なお、複数の出入口を有する空間に流入した後、歩行者は流出地点種類（Ｄ）のみの属性を持つことになる。 [2-2-3. Setting pedestrian attributes]
By combining the inflow point type (O) into the space having a plurality of entrances and the outflow point type (D) from the space having a plurality of entrances, the attribute of the pedestrian is determined.

Is written. In addition, after flowing into the space having a plurality of doorways, the pedestrian has only the attribute of the outflow point type (D).

と表記する。また、属性毎の人数を集計したメッシュ（ｉ，ｊ）内の歩行者の人数を、

と表記する。歩行者の人数は連続量として表記する。 [2-2-4. Status setting]
Consider the number and flow of pedestrians for each mesh. The number of pedestrians of attribute d in mesh (i, j) at time t is

Is written. In addition, the number of pedestrians in the mesh (i, j) that totals the number of people for each attribute,

Is written. The number of pedestrians is expressed as a continuous quantity.

Further, eight neighboring meshes for a certain mesh (i, j) are set as neighboring meshes. The pedestrian flow for mesh (i, j) is
(1) Flow from mesh (i, j) to neighboring mesh
(2) Flow from neighboring mesh to mesh (i, j)
(3) Inflow from outside the space to mesh (i, j)
(4) The flow is limited to outflow from the mesh (i, j). That is, the flow to a place farther than the neighboring mesh at one step time is not handled.

と表記する。また、逆向きの流動として、メッシュ（ｉ，ｊ）から近傍メッシュｋへ流出する属性（ｏ，ｄ）の歩行者の人数を、

と表記する。 The number of pedestrians with attribute d flowing from the neighboring mesh k to the mesh (i, j) between times t and t + 1 is

Is written. In addition, as the flow in the opposite direction, the number of pedestrians with attributes (o, d) flowing out from the mesh (i, j) to the neighboring mesh k is

Is written.

と表記する。また、複数の出入口を有する空間外への流出を、

と表記する。 In addition, the number of pedestrians with attributes (o, d) flowing into the mesh (i, j) from outside the space having a plurality of entrances is conveniently referred to as inflow from the neighborhood 0,

Is written. In addition, the outflow outside the space having a plurality of entrances,

Is written.

と表記する。 As with the number of people,

Is written.

と表記され、これらをまとめて広義の流動マップと呼ぶ。 [2-2-5. Flow map]
By performing various settings as described above, the number of pedestrians (q) and flow (f) at time t in a space having a plurality of entrances and exits,

These are collectively called a flow map in a broad sense.

を狭義の流動マップと呼ぶ。 In addition, the number of pedestrians at time t in each mesh,

Is called a flow map in a narrow sense.

が成り立つ。 [2-2-6. State change of flow map]
A change in the flow map due to a flow in a space having a plurality of entrances and a flow in and out of the space is called a state change. Organize state changes by number and flow. As a condition
(1) Pedestrians inside the mesh for reasons other than in-space flow and inflow / outflow to / from outside the space
Does not increase or decrease
(2) Considering that there is no movement from a mesh other than the neighboring mesh at the time of Δt,

Holds.

Ｄｄ：目的地点ｄのメッシュ集合
として得られることになる。 In addition, since the value of [Equation 4] has already been obtained, the number of pedestrians flowing from outside the space having a plurality of entrances represented by [Equation 8] is the number of inflows actually measured at each entrance / exit. Is multiplied by the OD branching rate.
On the other hand, as for the number of people outflowing from a space with multiple doorways (predicted outflow value), all the pedestrians of the mesh corresponding to the destination point will flow out of the space.

Dd: obtained as a mesh set of the destination point d.

  Returning to the flowchart of FIG. 3, the branching rate correction unit 9 determines whether or not a predetermined correction cycle (for example, one minute interval) has been reached (step S34), and determines that it is not the predetermined correction cycle. In the case (step S34: No), the process returns to step S31.

  On the other hand, when the branching rate correction unit 9 determines that it is a predetermined correction cycle (step S34: Yes), the branching rate correction unit 9 includes the number of outflows per entrance / exit measured by the traffic amount measuring device, In the flow situation prediction unit 8, a correction value for the OD branching rate for each entrance / exit is calculated so that the difference from the predicted outflow value for each entrance / exit obtained by the stochastic simulation based on the estimated OD traffic volume is reduced. The value is used to correct the OD branching rate for each doorway and store it in the storage unit 2 (step S35: branching rate correction step).

  Finally, the branching rate correction unit 9 (calculation control unit 50) determines whether or not the estimation of the OD traffic amount of the pedestrian is finished (for example, whether the upper limit value of the user-specified branching rate correction step number has been reached). Judgment is made (step S36), and the branching rate correction unit 9 (calculation control unit 50) does not end the estimation of the OD traffic amount of the pedestrian (for example, the user has reached the upper limit value of the predetermined branching rate correction step number). (No at step S36), the process returns to step S31, and on the other hand, the estimation of the OD traffic amount of the pedestrian is terminated (for example, the upper limit value of the user-specified branching rate correction step number is reached). If it is determined (step S36: Yes), the pedestrian OD traffic amount estimation process is terminated.

Ｏ_ｉ（ｔ）：流入地点ｉへ、時刻ｔ〜ｔ＋Δｔに流入する歩行者の人数（計測値）
＿
Ｒ_ｉ，ｊ（ｔ）：ｉからｊへのＯＤ分岐率（推計値） [2-3. Explanation of how to find the correction value]
Here, how to obtain the correction value of the OD branching rate for each entrance / exit in step S35 (branch rate correcting step) will be described in more detail. The estimated OD traffic volume is expressed by [Equation 16].

O _i (t): Number of pedestrians (measured value) flowing into the inflow point i from time t to t + Δt
_
R _{i, j} (t): OD branching rate from i to j (estimated value)

Ｄ_ｊ（ｔ）：流出地点ｊで、時刻ｔ〜ｔ＋Δｔに流出する歩行者の人数（計測値）
＿
Ｄ_ｊ（ｔ）：流出地点ｊで、時刻ｔ〜ｔ＋Δｔに流出する歩行者の人数（予測値） Then, the difference between the number of outflows per doorway measured by the traffic measuring device and the outflow prediction value for each doorway acquired from the flow state prediction unit 8 is [Equation 17].

D _j (t): Number of pedestrians (measured value) flowing out at time t to t + Δt at outflow point j
_
D _j (t): Number of pedestrians that flow out at time t to t + Δt at the outflow point j (predicted value)

ΔＲ_ｉ，ｊ（ｔ）：ＯＤ分岐率の補正値
＿
Ｒ_ｉ，ｊ（ｔ）：ＯＤ分岐率（推計値）
但し、

を満たすために、

を満たす必要がある。 Further, the correction of the OD branching rate is performed by [Equation 18].

ΔR _{i, j} (t): Correction value of OD branching ratio _
R _{i, j} (t): OD branching rate (estimated value)
However,

In order to meet the,

It is necessary to satisfy.

Further, in order to obtain the correction value of the OD branching rate, the following is assumed as a precondition.
Precondition 1: There is no extreme bias in the moving speed, and the average speed is almost uniform depending on the density.
If you move with v
Precondition 2: The OD branching rate changes slowly and is almost the same value for a short time.

だけ遡った時刻を用いて、［数２２］と表される。

ε：誤差 Here, the number of outflows at time t at the outflow point j is determined by the past number of inflows and the OD branching rate according to the distance. This time is determined by using the distances d _{i, j} and the speeds v between the inflow points i and the outflow points j.

It is expressed as [Equation 22] using a time that is only traced back.

ε: error

The simulation (predicted value) is the same, and is expressed as [Equation 23].

  However, if the correction value of the OD branching rate for each entrance / exit is strictly determined so that the difference becomes smaller based on the preconditions, it takes time to calculate and the real-time property deteriorates. A correction value for the branching rate is obtained approximately and efficiently.

[2-3-1. Setting conditions]
The following conditions are set in order to efficiently and approximately obtain the correction value of the OD branching ratio for each doorway.
Condition 1: The sign of the correction value is determined according to the sign of the difference.
Condition 2: Increase the correction value for doorways with large differences
Condition 3: Increase the correction value for entrances / exits with large inflows

[2-3-2. Condition 1]
For example, when the difference between the number of outflows at a certain doorway and the predicted outflow value is positive, the correction value is made positive because the branching rate for the predicted outflow value of the OD branching rate at that doorway is insufficient. On the other hand, when the difference between the number of outflows at a certain doorway and the predicted outflow value is negative, the correction value is made negative because the branching rate of the OD branching rate at the doorway with respect to the predicted outflow value is too large.

However, when all the differences are positive or negative, if all the correction values are positive or negative, the condition of [Equation 20] may not be satisfied. Therefore, the average value of the differences is expressed as [Equation 24]. To do.

  If the value obtained by subtracting the average value of the difference from the difference between the number of outflows at a certain doorway and the predicted outflow value is positive, the branching rate for the predicted outflow value of the OD branching rate at that doorway is insufficient. Make the value positive. On the other hand, if the value obtained by subtracting the average value of the difference from the difference between the number of outflows at a certain doorway and the predicted outflow value is negative, the branching rate for the predicted outflow value of the OD branching rate at that doorway is too large. Make the value negative.

[2-3-3. Condition 2]
In order to increase the correction value as the entrance / exit with a larger difference between the number of outflows at a certain entrance / exit and the predicted outflow value, the magnitude relationship of the difference is expressed by [Equation 25].

  Then, the magnitude of the difference is relatively evaluated by [Equation 25] to determine the magnitude relationship of the correction values.

[2-3-4. Condition 3]
In order to preferentially increase the correction value of the inflow point i where the number of inflowing people is large, the relationship between the number of inflowing people is expressed by [Equation 26].

  Then, the size of the number of inflows is relatively evaluated by [Equation 26], and the magnitude relationship of the correction values is determined.

As an intermediate stage for obtaining the correction value in consideration of such conditions 1 to 3, the basic value before satisfying the condition of [Equation 20] is represented by [Equation 27].

In order to make the basic value of [Equation 27] correspond to the magnitude of the difference, it is expressed by [Equation 28] using the correction value k _j .

と一定の値を与える。 Then, a correction value k _j is obtained from [Equation 28]. However, if the difference and the positive and negative of the basic value are different, it is not necessary to align with the size of the difference,

And give a certain value.

Here, “k _j × ΔR ′ _{i, j} ” does not satisfy the condition of [Equation 20], and is finally corrected. For the outflow point j, the sum of the positive values of “k _j × ΔR ′ _{i, j} ” is S +, the sum of the negative values is S−, and the absolute value of S + and S− is obtained. _{When big} and small are S _small , the correction value is obtained by [Equation 31].

As a result of the estimation of the pedestrian OD traffic amount estimation apparatus 100 according to the embodiment of the present invention, the OD branching rate at each doorway and the OD traffic amount at a specified time (t _{from to} t _to ) ] And [Equation 33].

  As described above, the OD branching rate is multiplied by the measured number of inflows to estimate the OD traffic volume, and the average flow situation of pedestrians is stochastically simulated based on the estimated OD traffic volume. The correction value is efficiently obtained by calculating the correction value of the OD branching rate at each entrance and exit and correcting the OD branching rate so that the difference between the predicted outflow value and the number of outflows at each entrance and exit is reduced. Since the OD traffic volume estimated based on the number of inflowers, which is the cross-sectional traffic volume, can be made to follow the actual OD traffic volume, the OD traffic volume of pedestrians in a space having a plurality of doorways Can be accurately grasped in real time.

[3. Explanation of pedestrian congestion map]
The map creating unit 10 obtains a narrowly defined flow map ([Formula 13]) which is the number of pedestrians in each mesh at the time t, which is obtained at the time of simulation in a space where the flow state prediction unit 8 has a plurality of entrances. Acquired, a map showing the distribution of the number of pedestrians at time t in the mesh divided in FIG. 4 is created and displayed on the display unit 3. For example, the map creation unit 10 creates a map in which each mesh is displayed by an identifiable display method such as color-coded display or grayscale display according to the number of pedestrians in each mesh, and is displayed on the display unit 3. Let

  For example, FIG. 6 is an explanatory diagram illustrating an example of a map created by the map creation unit 10. In FIG. 6, DS 61 indicates an area where the number of pedestrians in the mesh is the largest and the congestion is high, and DS 64 is It shows an area where the number of pedestrians in the mesh is the smallest and no congestion occurs. DS62 and DS63 in FIG. 6 indicate an area that is an intermediate congestion state between DS61 and DS64, and the size relationship of the number of pedestrians in the mesh is DS61> DS62> DS63> DS64.

  As described above, by creating and displaying a map showing the distribution of the number of pedestrians in each mesh in the space, the degree of congestion of the number of pedestrians in the space can be visually understood, so a plurality of entrances It is possible to grasp a pedestrian's congestion situation in a space having a real time or grasp a temporal change.

  In the description of the embodiment, the flow situation prediction unit 8 performs a probabilistic simulation (probably predicts the average movement situation of the pedestrian based on the OD traffic volume of the pedestrian at each entrance / exit). Of course, other existing simulation methods may be adopted.

  In the description of the embodiment, conditions 1 to 3 are set in order to efficiently and approximately obtain the correction value of the OD branching rate for each doorway, but of course, the correction of the OD branching rate for each doorway. You may set the other conditions which can obtain | require a value efficiently and approximately.

  In the description of the embodiment, the space SP21 is exemplified as a space having four doorways. Of course, the number of doorways is not limited, and any space having a plurality of doorways may be used.

DESCRIPTION OF SYMBOLS 1 Input part 2 Memory | storage part 3 Display part 4 Communication part 5 Operation part 6 Measurement value acquisition part 7 Traffic volume estimation part 8 Flow condition prediction part 9 Branch rate correction | amendment part 10 Map creation part 50 Calculation control part 100 Apparatus

Claims (6)

A method for estimating the OD traffic volume of a pedestrian in a space having a plurality of doorways,
A measurement value acquisition step for acquiring the number of inflows and outflows for each of the entrances and exits measured by the traffic measuring device installed for each of the entrances and exits;
A traffic volume estimating step of multiplying the number of inflows by the OD branching rate for each doorway to estimate the amount of OD traffic for each doorway;
A flow situation prediction step for stochastically simulating an average flow situation of pedestrians based on the estimated OD traffic volume for each doorway to obtain an outflow prediction value for each doorway;
A branch that corrects the OD branching rate by obtaining a correction value of the OD branching rate for each doorway so that a difference between the number of outflowers at each doorway and the predicted outflow value for each doorway becomes small at a predetermined cycle. Rate correction step;
The estimation method characterized by including. In the branching rate correction step,
When the value obtained by subtracting the average value of the difference from the difference is positive, the correction value is negative. When the value obtained by subtracting the average value of the difference from the difference is negative, the correction value is positive. The estimation method according to claim 1, wherein: In the branching rate correction step,
The estimation method according to claim 1, wherein the correction value of the entrance / exit having a large difference is increased. In the branching rate correction step,
The estimation method according to any one of claims 1 to 3, wherein the correction value of the entrance / exit with a large number of inflowing persons is increased. In the branching rate correction step,
5. The estimation method according to claim 1, wherein the correction value is adjusted for each of the entrances and exits so that the sum of the correction values becomes zero. 6.   Including a map creation step of acquiring the number of pedestrians in each mesh obtained by dividing the space obtained by the simulation of the flow state prediction step, and creating a map indicating the distribution of the number of pedestrians in the space. The estimation method according to claim 1, wherein the estimation method is a characteristic.

Images