Traffic volume estimation device, method for estimating traffic volume, and computer program

Abstract

Using sparse information on the number of people and vehicles, and sparse information on travel routes, we estimate information on dense travel routes. [Solution] The traffic volume estimation device comprises an acquisition unit that acquires: person-vehicle count measurement information, which is the number of people and / or vehicles that pass through a point, measured by a measuring device installed at any point within the estimation target area; and location history information, which represents a series of actual location information that has changed with the movement of any person-vehicle that moves through the estimation target area; and a location estimation unit that uses the acquired person-vehicle count measurement information and location history information to estimate extended location history information, which represents a series of location information that has likely changed with the movement of each person-vehicle, for all person-vehicles within the estimation target area.

Description

Technical Field

[0001] The present invention relates to a technique for estimating the traffic volume.

Background Art

[0002] In recent years, there has been a demand for dealing with large crowds of people generated due to the holding of large-scale events and the like. Examples of dealing with large crowds of people include, for example, traffic flow design for preventing mass accidents and ensuring social distancing, traffic flow design for guiding a part of the crowd to nearby commercial facilities, and mobility service design considering the crowd flow. By predicting the crowd flow generated due to the holding of an event or the like and performing these designs in consideration of the predicted crowd flow, it is possible to improve safety and expect an effect of reducing the number of personnel required for traffic control and an economic effect.

[0003] As a device for predicting the crowd flow, for example, Patent Document 1 describes a crowd flow estimation device that acquires the number of passing people and the walking direction of pedestrians from two or more crowd flow measurement devices having different measurement characteristics and generates OD data, which is the number of pedestrians going from a virtual entrance to a virtual exit. For example, Patent Document 2 describes a crowd flow survey support system that generates a large number of hypotheses based on a crowd flow model expressing changes in the crowd flow and calculates a crowd flow distribution that does not conflict with the measurement data by evaluating the compatibility with the measurement data obtained by a crowd flow measuring instrument.

Prior Art Documents

Patent Documents

[0004]

Patent Document 1

Patent Document 2

Summary of the Invention

Problems to be Solved by the Invention

[0005] However, the device described in Patent Document 1 has the problem that it can only estimate the number of pedestrians at a given point, and cannot estimate the pedestrians' movement paths. A pedestrian's movement path is a sequence of positional information that changes as pedestrians move. Furthermore, although the system described in Patent Document 2 can estimate pedestrian movement paths, it has the problem that it is not easy to apply because it requires the results of a complete enumeration survey (such as a questionnaire survey) for estimation. It should be noted that these problems are common not only when predicting the flow of people, but also when predicting the flow of vehicles. Hereafter, people and / or vehicles will also be referred to as "people and vehicles".

[0006] The present invention has been made to solve at least some of the problems described above, and aims to estimate information on dense travel routes using sparse information on the number of people and vehicles and sparse information on travel routes. [Means for solving the problem]

[0007] The present invention has been made to solve at least some of the problems described above, and can be realized in the following forms.

[0008] (1) According to one embodiment of the present invention, a traffic volume estimation device is provided. This traffic volume estimation device comprises: an acquisition unit that acquires: a person-vehicle count measurement information representing the number of people and / or vehicles measured by a measuring device installed at any point in the estimation target area, which measures the number of people and / or vehicles passing through the point; and location history information representing a series of actual location information that has changed with the movement of any person-vehicle that moves through the estimation target area; and a location estimation unit that uses the acquired person-vehicle count measurement information and location history information to estimate extended location history information representing a series of location information that would have changed with the movement of each person-vehicle for all person-vehicles in the estimation target area.

[0009] The number of people and vehicles measured using a measuring device has a spatial constraint: it only represents the number of people and vehicles at the location where the measuring device is installed. In other words, the number of people and vehicles measured by the acquisition unit is sparse information. Furthermore, since the location history information is the history of location information from mobile terminals or in-vehicle terminals carried by people, it has a coverage constraint: the source of the information is limited to terminals that have authorized the provision of location information. In other words, the location history information acquired by the acquisition unit is sparse information about movement paths. With this configuration, the location estimation unit can use a combination of this number of people and vehicles measured and location history information to compensate for the shortcomings of both (spatial constraints and coverage constraints), and estimate extended location history information that represents the series of location information that would have changed with the movement of each person and vehicle for all people and vehicles within the estimation target area. Since the extended location history information is information about the movement paths of all people and vehicles within the estimation target area, it can be said to be dense information about movement paths. As a result, with this configuration, dense information about movement paths can be estimated using sparse information about the number of people and vehicles and sparse information about movement paths.

[0010] (2) The traffic volume estimation device of the above form further includes a person-vehicle count estimation unit that uses the acquired person-vehicle count measurement information and the location history information to estimate extended person-vehicle count measurement information that represents the number of people and vehicles at all locations in the estimation target area, including locations where the measurement device is not installed, and the location estimation unit may use the extended person-vehicle count measurement information to estimate the extended location history information instead of the person-vehicle count measurement information acquired by the acquisition unit if the extended person-vehicle count measurement information estimated by the person-vehicle count estimation unit exists. With this configuration, the passenger-vehicle count estimation unit can estimate extended passenger-vehicle count information using passenger-vehicle count measurement information (sparse passenger-vehicle count information) and location history information (sparse travel path information). Since the extended passenger-vehicle count measurement information is the number of passengers at all locations, including locations where no measurement devices are installed, it can be said to be dense passenger-vehicle count information. Furthermore, if extended passenger-vehicle count measurement information estimated by the passenger-vehicle count estimation unit exists, the location estimation unit will use the extended passenger-vehicle count measurement information to estimate the extended location history information instead of the passenger-vehicle count measurement information acquired by the acquisition unit, thereby improving the estimation accuracy of the location estimation unit.

[0011] (3) In the traffic volume estimation device of the above form, the person-vehicle count estimation unit may further include a convergence determination unit that, if the extended position history information estimated by the position estimation unit exists, estimates the extended person-vehicle count measurement information using the extended position history information instead of the position history information acquired by the acquisition unit, and repeatedly performs the estimation of the extended person-vehicle count measurement information by the person-vehicle count estimation unit and the estimation of the extended position history information by the position estimation unit until a predetermined condition is met. With this configuration, if extended location history information estimated by the location estimation unit exists, the passenger-vehicle count estimation unit will use the extended location history information instead of the location history information acquired by the acquisition unit to estimate the extended passenger-vehicle count measurement information, thereby improving the estimation accuracy of the passenger-vehicle count estimation unit. Furthermore, the convergence determination unit will repeatedly perform the estimation of extended passenger-vehicle count measurement information by the passenger-vehicle count estimation unit and the estimation of extended location history information by the location estimation unit until predetermined conditions are met, thereby further improving the estimation accuracy of both the passenger-vehicle count estimation unit and the location estimation unit.

[0012] (4) In the traffic volume estimation device of the above form, the predetermined condition may be that the prediction accuracy calculated from the comparison result between the person-vehicle count measurement information acquired by the acquisition unit and the extended person-vehicle count measurement information estimated by the person-vehicle count estimation unit is equal to or greater than a predetermined threshold. With this configuration, the prediction accuracy can be calculated by comparing the expanded passenger and vehicle count information estimated by the passenger and vehicle count estimation unit with the ground truth data, using the actual passenger and vehicle count measurement information obtained by the measuring device as the ground truth data. The convergence determination unit then terminates the iterative process when the prediction accuracy exceeds a predetermined threshold, thereby further improving the estimation accuracy by both the passenger and vehicle count estimation unit and the position estimation unit.

[0013] (5) In the traffic volume estimation device of the above form, the position estimation unit may estimate the extended position history information by generating candidate movement paths using the position history information, and selecting a set of movement paths that match the person-vehicle count measurement information from the generated candidate movement paths. With this configuration, the position estimation unit can easily estimate extended position history information using the number of people and vehicles measured and the position history information.

[0014] (6) In the traffic volume estimation device of the above form, the position estimation unit may generate candidate travel routes by considering at least one of the following: map-like conditions, environmental conditions, temporal conditions, and event-related conditions. With this configuration, the position estimation unit can generate candidate movement paths that are weighted according to the situation and the purpose of estimating the extended position history information.

[0015] Furthermore, the present invention can be realized in various forms, for example, as a traffic volume estimation device for estimating information on dense travel routes, a traffic volume estimation system, a design system for estimating information on dense travel routes and designing traffic flow, a control method for these devices and systems, a computer program executed in these devices and systems, a server device for distributing the computer program, a non-temporary storage medium storing the computer program, and so on. [Brief explanation of the drawing]

[0016] [Figure 1] This is an explanatory diagram illustrating the configuration of an estimation system as one embodiment of the present invention. [Figure 2] This diagram explains the information regarding the number of people and vehicles measured. [Figure 3] This is a diagram explaining location history information. [Figure 4] This flowchart shows an example of the processing steps in the position estimation unit. [Figure 5] This figure illustrates step S30 in Figure 4. [Figure 6] This is a diagram for explaining the extended position history information. [Figure 7] This is an explanatory diagram illustrating the configuration of the estimation system in the second embodiment. [Figure 8] This is a diagram for explaining the extended number of people and vehicles measurement information. [Figure 9] This is an explanatory diagram illustrating the configuration of the estimation system in the third embodiment.

Mode for Carrying Out the Invention

[0017] <First Embodiment> FIG. 1 is an explanatory diagram illustrating the configuration of an estimation system 1 as an embodiment of the present invention. The estimation system 1 is a system for estimating (predicting) the flow of people and vehicles within an arbitrary estimation target area. The flow of people and vehicles estimated by the estimation system 1 can be used, for example, for accident prevention, traffic flow design for ensuring social distancing, traffic flow design for guiding a part of people and vehicles to nearby commercial facilities, and mobility service design considering the flow of people and vehicles. The estimation system 1 includes a traffic volume estimation device 100, a people and vehicle number storage unit 200, and a position history storage unit 300. In FIG. 1, the transmission and reception of the people and vehicle number measurement information IN1 and the position history information IN2 are shown by thin dashed lines, and the transmission and reception of the extended position history information IN21 are shown by thick dashed lines.

[0018] In the present embodiment, "people and vehicles" means people and / or vehicles. The term "people and vehicles" hereinafter may be interpreted to mean only people, only vehicles, or both people and vehicles. The "estimation target area" of the estimation system 1 is the area where the flow of people and vehicles is to be estimated by the estimation system 1. The estimation target area is defined in advance. The estimation target area can be set arbitrarily. For example, if it is outdoors, it can be a facility where a large-scale event is held and the surrounding blocks, and if it is indoors, it can be a floor where a large-scale event is held and the surrounding area, etc. The estimation target area may be an arbitrary outdoor or indoor range unrelated to a large-scale event.

[0019] The person-vehicle count storage unit 200 is a storage medium that pre-stores person-vehicle count measurement information IN1. The person-vehicle count measurement information IN1 is the number of people and / or vehicles actually measured by a measuring device installed at any point within the target area for estimation. "Person-vehicle count" means the number of people and / or vehicles. The person-vehicle count storage unit 200 is communicated with the traffic volume estimation device 100. The person-vehicle count storage unit 200 may be a storage medium built into or connected to the traffic volume estimation device 100. The person-vehicle count storage unit 200 may be provided on a server or the like located on a network.

[0020] Figure 2 is a diagram illustrating the pedestrian / vehicle count measurement information IN1. For example, as shown in Figure 2, suppose an area of ​​arbitrary size containing multiple links L is predetermined as the estimation target area TA. Links L may be roads, sidewalks, passageways within buildings, or passageways on private property. In the example in Figure 2, links L are roads. Of all the links L included in the estimation target area TA, some links L have measuring devices installed. In Figure 2, links L with measuring devices installed are referred to as links LM. The pedestrian / vehicle count measurement information IN1 is the number of pedestrians and vehicles actually passing through each link LM as measured by the measuring devices installed on each link LM. As shown in Figure 2, if 10 measuring devices are installed within the estimation target area TA, the pedestrian / vehicle count measurement information IN1 includes the number of pedestrians and vehicles at 10 locations measured by these 10 measuring devices. In Figure 2, links with a relatively large number of pedestrians and vehicles are colored more darkly than links with a relatively small number of pedestrians and vehicles, according to the "number of pedestrians and vehicles per link" included in the pedestrian / vehicle count measurement information IN1. Therefore, link LM is a dark gray, almost black. Link L, shown in very light gray in Figure 2, does not have passenger / vehicle count data; in other words, no measuring device is installed there.

[0021] As shown in Figure 2, the passenger / vehicle count measurement information IN1 includes only the number of passengers / vehicles at the location where the measuring device is installed, and can therefore be said to be sparse passenger / vehicle count information. As a measuring device, for example, PASSER-NET (Giken Trustem Co., Ltd. and PASSER-NET are registered trademarks) can be used. The passenger / vehicle count measurement information IN1 may simply be the number of passengers / vehicles at the location where the measuring device is installed. The passenger / vehicle count measurement information IN1 may also be the number of passengers / vehicles by direction of movement at the location where the measuring device is installed. For the sake of simplicity, from here on, the passenger / vehicle count measurement information IN1 will be described as simply the number of passengers / vehicles without considering the direction of movement.

[0022] The location history storage unit 300 is a storage medium that has previously stored location history information IN2. The location history information IN2 is the actual movement path of any person or vehicle moving around the estimated target area. "Movement path" means a series (history) of location information that changes as a person or vehicle moves. The location history storage unit 300 is connected to the traffic volume estimation device 100 in a communicative manner. The location history storage unit 300 may be a storage medium built into or connected to the traffic volume estimation device 100. The location history storage unit 300 may be provided on a server or the like located on a network.

[0023] Figure 3 is a diagram illustrating location history information IN2. The estimated target area TA in Figure 3 corresponds to the estimated target area TA in Figure 2. If, for example, there are 1000 people in the estimated target area TA as "arbitrary people / vehicles," then location history information IN2 includes 1000 movement paths acquired from each of these 1000 people's terminals. Here, "arbitrary people / vehicles" refers to people / vehicles possessing terminals (mobile terminals or in-vehicle terminals) that have authorized the provision of location information. In Figure 3, for the sake of explanation, all points representing location information included in the 1000 movement paths measured over a certain 5-minute period are drawn.

[0024] Even if there are actually 10,000 people within the estimated target area, the movement paths of the 9,000 people who possess devices that do not permit the sharing of location information are not included in the location history information IN2. Therefore, as shown in Figure 3, location history information IN2 can be said to be sparse movement path information, in other words, information on sparse movement paths, which only includes the movement paths of people and vehicles possessing devices that permit the sharing of location information. One method for acquiring movement paths is to use point-type trip data provided by Agoop Co., Ltd. (Agoop is a registered trademark). Point-type trip data can acquire location (latitude and longitude), GPS accuracy, movement speed, etc. from the terminals of people and vehicles at predetermined time intervals and record them along with the time. GPS is an abbreviation for Global Positioning System.

[0025] Returning to Figure 1, let's continue the explanation. The traffic volume estimation device 100 comprises a CPU 10, a memory unit 20, a communication unit 30, and a ROM / RAM 40, and each unit is interconnected by a bus (not shown). The traffic volume estimation device 100 can be configured using, for example, an information processing device such as a personal computer.

[0026] The storage unit 20 consists of a hard disk, flash memory, memory card, etc. The storage unit 20 has a map information database 21 pre-stored in it. Hereafter, the database will also be referred to as "DB". The map information DB 21 contains information necessary for map display, such as terrain, buildings, and road shapes. The map information DB 21 also contains definition information for the estimated target area TA within the map. The definition information for the estimated target area TA may be prepared separately from the map information DB 21. The communication unit 30 controls communication with other devices via a communication interface (not shown). Other devices may include a passenger count storage unit 200 and a location history storage unit 300, as well as a server (not shown).

[0027] The CPU 10 controls each part of the traffic volume estimation device 100 by loading the computer program stored in the ROM 40 into the RAM 40 and executing it. The CPU 10 also functions as an acquisition unit 11, a position estimation unit 12, and an output unit 13. The acquisition unit 11 acquires the number of people and vehicles measurement information IN1 from the number of people and vehicles storage unit 200 and transmits the acquired number of people and vehicles measurement information IN1 to the position estimation unit 12. The acquisition unit 11 also acquires the position history information IN2 from the position history storage unit 300 and transmits the acquired position history information IN2 to the position estimation unit 12.

[0028] Figure 4 is a flowchart illustrating an example of the processing performed by the position estimation unit 12. The position estimation unit 12 uses the information acquired by the acquisition unit 11 (person-vehicle count measurement information IN1 and position history information IN2) to estimate extended position history information IN21, which represents the sequence of position information that would have changed with the movement of each person-vehicle, for all people-vehicles within the target area TA. The processing shown in Figure 4 may be executed at predetermined triggers. For example, the processing shown in Figure 4 may be executed when the position estimation unit 12 acquires information from the acquisition unit 11, or it may be executed when a predetermined application installed on the traffic volume estimation device 100 is started. First, in step S10, the position estimation unit 12 acquires the person-vehicle count measurement information IN1 and the position history information IN2 from the acquisition unit 11.

[0029] In step S20, the position estimation unit 12 generates candidate movement paths using the position history information IN2. The position estimation unit 12 can generate candidate movement paths, for example, by following the steps a1 and a2 below. (a1) The position estimation unit 12 compares the travel paths included in the position history information IN2 with the map information DB21, for example by map matching, and recognizes them as travel paths using links L within the target area TA. Procedure a1 is performed for all travel paths included in the position history information IN2. (a2) The position estimation unit 12 generates a large number of travel routes similar to each travel route obtained in step a1. The specific number of candidate routes to be generated can be determined arbitrarily. For example, in order to obtain a travel route similar to a certain travel route, the position estimation unit 12 can perform at least one combination of changing the starting point, changing the destination, changing the waypoints, and changing the links L that pass through the travel route. It is preferable that the position estimation unit 12 continues to generate candidate travel routes until all links L included in the estimation target area TA of the map information DB 21 are covered by the candidate travel routes generated in step a2.

[0030] The position estimation unit 12 may generate weighted candidate movement paths in step a2. Specifically, the position estimation unit 12 may weight the number of movement paths generated in step a2 using any of the following conditions b1 to b4, or a combination of conditions b1 to b4. (b1) Geographical conditions: For example, conditions such as road width, sidewalk width, and the number of facilities adjacent to the road. (b2) Environmental conditions: for example, conditions such as season, temperature, and weather. (b3) Time conditions: For example, conditions related to the time of day (early morning, morning, lunchtime, afternoon, evening, night, etc.). (b4) Event-related conditions: For example, conditions regarding the location where the event will be held.

[0031] For example, when using condition b1, the position estimation unit 12 refers to the road width (or sidewalk width) of each link L from the map information DB 21 and generates more candidate travel routes through links L with wider roads than travel routes through links L with narrower roads. For example, when using condition b2, the position estimation unit 12 refers to the presence or absence of a roof (a roof covering the sidewalk) at each link from the map information DB 21 and generates more candidate travel routes through links L with roofs than travel routes through links L without roofs. For example, when using condition b4, the position estimation unit 12 refers to the address (or latitude and longitude) of the location where the event is held and generates more candidate travel routes heading towards that location than candidate travel routes not heading towards that location. Conditions b1 to b4 may be applied according to the estimation purpose of the extended location history information IN 21. The estimation purpose of the extended location history information IN 21 is, for example, to estimate the flow of people towards a fireworks display held on a summer night, or to estimate the flow of people towards a festival held during the daytime of Golden Week.

[0032] Figure 5 is a diagram illustrating step S30 in Figure 4. In step S30 of Figure 4, the position estimation unit 12 selects a set of movement paths that match the passenger / vehicle count measurement information IN1 from the candidate movement paths generated in step S20. For example, the case in which the position estimation unit 12 selects a set of movement paths in the range consisting of links L1 to L14 shown in Figure 5(A) will be explained as an example. In the example of Figure 5(A), the passenger / vehicle count measurement information IN1 exists only in link L1 connecting nodes N1 and N2, and the passenger / vehicle count measurement information IN1 represents 3. In this case, the position estimation unit 12 selects a combination of three paths that pass through link L1 from the candidate movement paths generated in step S30. For example, the position estimation unit 12 can select a set of movement paths consisting of two paths RT1 shown in Figure 5(B) and one path RT1 shown in Figure 5(C), for a total of three paths. Path RT1 is a movement path that passes through links L2, L4, L1, and L10. Route RT2 is a travel path that passes through links L6, L1, and L11.

[0033] The position estimation unit 12 considers all the number of people and vehicles included in the person and vehicle count measurement information IN1 (for example, the number of people and vehicles at 10 locations in the example above) and selects a set of travel paths with the smallest error for all the number of people and vehicles, in the manner described in Figure 5. The magnitude of the error may be determined by whether or not it is below a predetermined threshold, or the content of the process that produced the smallest error after executing a predetermined number of processes may be selected. The position estimation unit 12 can perform the above selection of the set of travel paths using, for example, the solution method of an integer linear programming problem.

[0034] The position estimation unit 12 defines the set of movement paths selected in step S30 as "extended position history information IN21". After transmitting the extended position history information IN21 to the output unit 13, the position estimation unit 12 terminates processing. The position estimation unit 12 may repeatedly perform candidate generation in step S20 and set selection in step S30. For example, if the error in step S30 (error relative to the number of people / vehicles) is greater than a predetermined allowable number, the position estimation unit 12 may repeatedly perform steps S20 and S30 until the error becomes less than or equal to the allowable number.

[0035] Figure 6 is a diagram illustrating the extended location history information IN21. The extended location history information IN21 represents the movement paths (series of location information) that would have changed with the movement of each person or vehicle within the estimated target area. In other words, following the specific example described above, the extended location history information IN21 includes 9,000 movement paths estimated by the location estimation unit 12, in addition to the 1,000 actual movement paths included in the location history information IN2. For the sake of explanation, Figure 6 shows all the points representing location information included in the 10,000 movement paths measured over a certain 5-minute period. In Figure 6, it can be seen that links with relatively high pedestrian traffic are drawn thicker due to the overlap of many points. Thus, the extended location history information IN21 is "dense movement path information" that includes the movement paths of all people and vehicles within the estimated target area TA.

[0036] The output unit 13 in Figure 1 outputs the extended location history information IN21 acquired from the location estimation unit 12 using any output method. The output unit 13 outputs the extended location history information IN21 by, for example, displaying it on an output unit (such as a display) connected to the traffic volume estimation device 100. The output unit 13 may also output the extended location history information IN21 by, for example, saving it to the storage unit 20 or an externally connected storage medium.

[0037] As described above, according to the estimation system 1 of the first embodiment, the traffic volume estimation device 100 can estimate dense travel route information using sparse information on the number of people and vehicles and sparse information on travel routes. The number of people and vehicles measurement information IN1 obtained using the measuring device has a spatial constraint that it is only the number of people and vehicles at the point where the measuring device is installed. In other words, the number of people and vehicles measurement information IN1 acquired by the acquisition unit 11 is sparse information on the number of people and vehicles. In addition, the location history information IN2 is a history of location information of a mobile terminal or in-vehicle terminal carried by a person, and therefore has a coverage constraint that the source of the information is limited to terminals that have been authorized to provide location information. In other words, the location history information acquired by the acquisition unit 11 is sparse information on travel routes. According to the traffic volume estimation device 100 of the first embodiment, the position estimation unit 12 uses a combination of such person-vehicle count measurement information IN1 and position history information IN2 to compensate for the shortcomings of both (spatial constraints, coverage constraints), and can estimate extended position history information IN21, which represents the sequence of position information that would have changed with the movement of each person-vehicle, for all people-vehicles within the estimation target area TA. Since the extended position history information IN21 is information about the movement paths of all people-vehicles within the estimation target area TA, it can be said to be information about dense movement paths. In other words, according to the traffic volume estimation device 100 of the first embodiment, information about dense movement paths can be estimated using sparse person-vehicle count information and sparse movement path information.

[0038] Furthermore, according to the estimation system 1 of the first embodiment, the position estimation unit 12 generates candidate movement paths using the position history information IN2, and estimates the extended position history information IN21 by selecting a set of movement paths that match the number of people and vehicles measured information IN1 from the generated candidate movement paths (Figures 4 and 5). Therefore, the position estimation unit 12 can easily estimate the extended position history information IN21 using the number of people and vehicles measured information IN1 and the position history information IN2.

[0039] Furthermore, in the estimation system 1 of the first embodiment, if the position estimation unit 12 generates candidate movement paths by considering at least one of the conditions b1 to b4, the position estimation unit 12 can generate weighted candidate movement paths according to the situation and the estimation purpose of the extended position history information IN21.

[0040] <Second Embodiment> Figure 7 is an explanatory diagram illustrating the configuration of the estimation system 1A in the second embodiment. In the second embodiment, a configuration capable of estimating extended passenger and vehicle count measurement information IN11 will be described. The estimation system 1A is configured as described in the first embodiment, but with a traffic volume estimation device 100A instead of the traffic volume estimation device 100. The traffic volume estimation device 100A is configured as described in the first embodiment, but with a passenger and vehicle count estimation unit 14, and with a position estimation unit 12A instead of the position estimation unit 12. In Figure 7, the transmission and reception of passenger and vehicle count measurement information IN1 and position history information IN2 are shown with thin dashed lines, and the transmission and reception of extended passenger and vehicle count measurement information IN11 and extended position history information IN21 are shown with thick dashed lines.

[0041] The person-vehicle count estimation unit 14 uses the information acquired by the acquisition unit 11 (person-vehicle count measurement information IN1 and location history information IN2) to estimate extended person-vehicle count measurement information IN11, which represents the number of people at all locations in the target area TA, including locations where no measuring devices are installed. The person-vehicle count estimation unit 14 can calculate the number of people at locations not measured by measuring devices, for example, by using the matrix decomposition method described in Reference 1. The person-vehicle count estimation unit 14 merges the calculated number of people (number of people at locations not measured by measuring devices) with the person-vehicle count measurement information IN11 to obtain the extended person-vehicle count measurement information IN11.

[0042] [Reference 1]: Taguchi et al., "Estimation of total traffic volume using simultaneous matrix decomposition," IEICE Technical Report, December 2023, vol. 123, no. 311, pp. 18-24.

[0043] Figure 8 is a diagram illustrating the extended passenger count measurement information IN11. The extended passenger count measurement information IN11 represents the number of passengers at all locations within the estimation target area TA, including locations where measuring devices are installed and locations where measuring devices are not installed. In Figure 8, as with Figure 2, links with a relatively large number of passengers are shown in a darker color than links with a relatively small number of passengers, according to the "number of passengers per link" included in the extended passenger count measurement information IN11. As can be seen from Figure 8, in the extended passenger count measurement information IN11, in addition to link LM where the measuring device is installed, all other links L are also displayed in a darker gray compared to link L in Figure 2. This is because there is a number of passengers estimated by the passenger count estimation unit 14 for all other links L as well. In Figure 8, the number of passengers at link LM and the number of passengers at the other links L are both based on the number of passengers estimated by the passenger count estimation unit 14. The "number of people / vehicles in link LM" input to the position estimation unit 12A may use the number of people / vehicles measurement information IN1, or the number of people / vehicles estimated by the number of people / vehicles estimation unit 14. The differences in the density of links L and LM in Figure 8 are due to differences in the number of people / vehicles. The extended number of people / vehicles measurement information IN11, as shown in Figure 8, is "dense information on the number of people / vehicles" that includes the number of people / vehicles at all points within the target estimation area TA.

[0044] The position estimation unit 12A estimates the extended position history information IN21 using the extended passenger count measurement information IN11 estimated by the passenger count estimation unit 14 and the position history information IN2 acquired by the acquisition unit 11. In other words, the position estimation unit 12A estimates the extended position history information IN21 using the extended passenger count measurement information IN11 estimated by the passenger count estimation unit 14 instead of the passenger count measurement information IN1 acquired by the acquisition unit 11. The estimation method is as described in the first embodiment. The passenger count estimation unit 14 may also transmit the estimated extended passenger count measurement information IN11 to the output unit 13. In this case, the output unit 13 may output the extended passenger count measurement information IN11 in addition to, or instead of, the extended position history information IN21. Any output method as described in the first embodiment can be used.

[0045] Thus, the configuration of the traffic volume estimation device 100A can be modified in various ways, and it may also include a person-vehicle number estimation unit 14. The estimation system 1A and traffic volume estimation device 100A of the second embodiment can also achieve the same effects as the first embodiment described above.

[0046] Furthermore, according to the traffic volume estimation device 100A of the second embodiment, the person-vehicle count estimation unit 14 can estimate extended person-vehicle count measurement information IN11 using person-vehicle count measurement information IN1 (sparse person-vehicle count information) and location history information IN2 (sparse travel path information). Since the extended person-vehicle count measurement information IN11 is the number of people-vehicles at all locations, including locations where no measurement device is installed, it can be said to be dense person-vehicle count information. Moreover, if the extended person-vehicle count measurement information IN11 estimated by the person-vehicle count estimation unit 14 exists, the location estimation unit 12A estimates the extended location history information IN21 using the extended person-vehicle count measurement information IN11 instead of the person-vehicle count measurement information IN1 acquired by the acquisition unit 11, thereby improving the estimation accuracy of the location estimation unit 12A.

[0047] <Third Embodiment> Figure 9 is an explanatory diagram illustrating the configuration of the estimation system 1B in the third embodiment. In the third embodiment, a convergence determination unit 15 is provided, and a configuration in which the estimation process by the position estimation unit 12B and the person-vehicle count estimation unit 14B is repeated is described. The estimation system 1B is the same as the configuration described in the second embodiment, but with a traffic volume estimation device 100B instead of the traffic volume estimation device 100A. The traffic volume estimation device 100B is the same as the configuration described in the second embodiment, but with a position estimation unit 12B instead of the position estimation unit 12A, a person-vehicle count estimation unit 14B instead of the person-vehicle count estimation unit 14, and further with a convergence determination unit 15. In Figure 9, the transmission and reception of person-vehicle count measurement information IN1 and position history information IN2 are shown by thin dashed lines, and the transmission and reception of extended person-vehicle count measurement information IN11 and extended position history information IN21 are shown by thick dashed lines.

[0048] In the initial processing (t1), the passenger count estimation unit 14B estimates the extended passenger count measurement information IN11 using the passenger count measurement information IN1 acquired by the acquisition unit 11 and the location history information IN2 acquired by the acquisition unit 11. The estimation method is as described in the second embodiment. In the initial processing (t1), the location estimation unit 12B estimates the extended location history information IN21 using the extended passenger count measurement information IN11 estimated by the passenger count estimation unit 14B and the location history information IN2 acquired by the acquisition unit 11. The estimation method is as described in the first embodiment.

[0049] In the second to n-1th processing iterations (t2 to tn-1), the passenger-vehicle count estimation unit 14B estimates the extended passenger-vehicle count measurement information IN11 using the passenger-vehicle count measurement information IN1 acquired by the acquisition unit 11 and the extended location history information IN21 estimated by the location estimation unit 12B. In other words, in the second processing iteration (t2), the passenger-vehicle count estimation unit 14B estimates the extended passenger-vehicle count measurement information IN11 using the extended location history information IN21 estimated by the location estimation unit 12B instead of the location history information IN2 acquired by the acquisition unit 11. By using the dense movement path information estimated by the location estimation unit 12B as input, the estimation accuracy of the passenger-vehicle count estimation unit 14B can be improved. Note that "n" is a natural number with an initial value of n=2, and is incremented along with the repeated processing of the passenger-vehicle count estimation unit 14B and the location estimation unit 12B until the convergence determination unit 15 determines that "a predetermined condition has been met".

[0050] The convergence determination unit 15 repeatedly performs the estimation of extended passenger count measurement information IN11 by the passenger count estimation unit 14B and the estimation of extended position history information IN21 by the position estimation unit 12B until predetermined conditions are met. The convergence determination unit 15 can adopt either of the following c1 or c2 as predetermined conditions. (c1) When the prediction accuracy of the extended passenger count measurement information IN11 exceeds a predetermined first threshold. The convergence determination unit 15 can calculate the prediction accuracy from the comparison result between the actual passenger count measurement information IN1 (number of passengers) acquired by the acquisition unit 11 and the extended passenger count measurement information IN11 (number of passengers) estimated by the passenger count estimation unit 14B for link LM (Figure 8) where the measurement device is installed. The first threshold can be set arbitrarily. (c2) When the number of iterations of the person-vehicle number estimation unit 14B and the position estimation unit 12B exceeds a predetermined second threshold. The second threshold can be set arbitrarily.

[0051] The passenger-vehicle count estimation unit 14B may use the extended passenger-vehicle count measurement information IN11(tn-1) estimated in the previous processing to estimate the extended passenger-vehicle count measurement information IN11(tn) for the current processing, instead of the passenger-vehicle count measurement information IN1 acquired by the acquisition unit 11. The passenger-vehicle count estimation unit 14B may also transmit the estimated extended passenger-vehicle count measurement information IN11 to the output unit 13. In this case, the output unit 13 may output the extended passenger-vehicle count measurement information IN11 in addition to, or instead of, the extended location history information IN21. Any output method as described in the first embodiment can be used.

[0052] Thus, the configuration of the traffic volume estimation device 100B can be modified in various ways, and the convergence determination unit 15 may be used to enable iterative processing of the number of people and vehicles estimation unit 14B and the position estimation unit 12B. The estimation system 1B and traffic volume estimation device 100B of the third embodiment can also achieve the same effects as those of the first and second embodiments described above.

[0053] Furthermore, according to the traffic volume estimation device 100B of the third embodiment, if there is extended position history information IN21 estimated by the position estimation unit 12B, the person-vehicle count estimation unit 14B estimates the extended person-vehicle count measurement information IN11 using the extended position history information IN21 instead of the position history information IN2 acquired by the acquisition unit 11, thereby improving the estimation accuracy of the person-vehicle count estimation unit 14B. Moreover, the convergence determination unit 15 repeatedly performs the estimation of the extended person-vehicle count measurement information IN11 by the person-vehicle count estimation unit 14B and the estimation of the extended position history information IN21 by the position estimation unit 12B until predetermined conditions are met, thereby further improving the estimation accuracy of both the person-vehicle count estimation unit 14B and the position estimation unit 12B.

[0054] Furthermore, the convergence determination unit 15 can calculate the prediction accuracy by comparing the extended passenger count measurement information IN11 estimated by the passenger count estimation unit 14B with the ground truth data, using the passenger count measurement information IN1 actually obtained by the measuring device as the ground truth data. The convergence determination unit 15 then terminates the iterative process when the prediction accuracy reaches or exceeds a predetermined first threshold, thereby further improving the estimation accuracy by the passenger count estimation unit 14B and the estimation accuracy by the position estimation unit 12B.

[0055] <Modified form of this embodiment> The present invention is not limited to the embodiments described above, and can be implemented in various forms without departing from its spirit. For example, a part of the configuration implemented by hardware may be replaced with software, or conversely, a part of the configuration implemented by software may be replaced with hardware. In addition, the following modifications are also possible, for example.

[0056] [Example 1] The above embodiment shows an example of the configuration of the estimation systems 1, 1A, 1B and the traffic volume estimation devices 100, 100A, 100B. However, the configuration of the estimation systems 1, 1A, 1B and the traffic volume estimation devices 100, 100A, 100B can be modified in various ways. For example, at least one of the person / vehicle count storage unit 200 and the location history storage unit 300 may be built into the traffic volume estimation device 100 or provided in the storage unit 20. For example, at least a part of each functional unit of the traffic volume estimation device 100, 100A, 100B (acquisition unit 11, location estimation unit 12, output unit 13, person / vehicle count estimation unit 14, convergence determination unit 15) may be implemented by an independent device provided externally (for example, a server on the cloud). In other words, multiple devices may cooperate to implement the functions of the traffic volume estimation device 100.

[0057] For example, multiple passenger / vehicle count storage units 200 may be connected to the traffic volume estimation device 100. In this case, the acquisition unit 11 can acquire multiple passenger / vehicle count measurement information IN1 from each of the multiple passenger / vehicle count storage units 200, generate merged data of the multiple passenger / vehicle count measurement information IN1, and transmit it to other functional units. For example, multiple location history storage units 300 may be connected to the traffic volume estimation device 100. In this case, the acquisition unit 11 can acquire multiple location history information IN2 from each of the multiple location history storage units 300, generate merged data of the multiple location history information IN2, and transmit it to other functional units.

[0058] For example, the traffic volume estimation devices 100, 100A, and 100B do not need to have at least some of the functional units described above. For example, the traffic volume estimation device 100 does not need to have an output unit 13. For example, the traffic volume estimation device 100 does not need to have an acquisition unit 11. If the acquisition unit 11 is omitted, the position estimation unit 12 and the person-vehicle count estimation unit 14 can each acquire person-vehicle count measurement information IN1 from the person-vehicle count storage unit 200 and position history information IN2 from the position history storage unit 300, respectively.

[0059] [Differentiation 2] In the above embodiment, an example of the processing content in the position estimation unit 12 (Figure 4) is shown. However, the processing content in the position estimation unit 12 can be modified in various ways. For example, the execution order of each step may be changed, some steps may be omitted, and other steps not described may be executed. For example, instead of the method described in steps S20 and S30, the position estimation unit 12 may estimate the extended position history information IN21 using a learning model that has been built in advance by machine learning. The learning model is a model that outputs the extended position history information IN21 when it receives the number of people and vehicles measurement information IN1 and the position history information IN2 as input.

[0060] [Difference 3] The estimation systems 1, 1A, 1B and traffic volume estimation devices 100, 100A, 100B of the first to third embodiments described above, and the estimation systems 1, 1A, 1B and traffic volume estimation devices 100, 100A, 100B of the modified examples 1, 2 described above may be combined as appropriate.

[0061] The embodiments of this specification have been described above based on the embodiments and modifications described above. The embodiments described above are for the purpose of facilitating understanding of this specification and do not limit it. This specification may be modified and improved without departing from its spirit and the scope of the claims, and equivalents thereof are included in this specification. Furthermore, any technical features that are not described as essential in this specification may be deleted as appropriate.

[0062] The present invention can also be realized in the following forms. [Application Example 1] Traffic volume estimation device, An acquisition unit that acquires: person-vehicle count measurement information representing the number of people and / or vehicles measured by a measuring device installed at any point within the estimated target area, which measures the number of people and / or vehicles passing through that point; and location history information representing a series of actual location information that changes with the movement of any person-vehicle moving around in the estimated target area; A position estimation unit that uses the acquired passenger vehicle count measurement information and the position history information to estimate extended position history information representing the sequence of position information that would have changed with the movement of each passenger vehicle for all passenger vehicles within the estimated target area, A traffic volume estimation device equipped with the following features. [Application Example 2] The traffic volume estimation device described in Application Example 1, further, The system includes a passenger count estimation unit that estimates extended passenger count measurement information representing the number of passengers at all locations within the target area, including locations where the measuring device is not installed, using the acquired passenger count measurement information and location history information. Traffic volume estimation device wherein, if the position estimation unit has expanded passenger and vehicle count measurement information estimated by the passenger and vehicle count estimation unit, it estimates the expanded position history information using the expanded passenger and vehicle count measurement information instead of the passenger and vehicle count measurement information acquired by the acquisition unit. [Application Example 3] A traffic volume estimation device as described in Application Example 1 or Application Example 2, If the passenger count estimation unit has the extended location history information estimated by the location estimation unit, it will use the extended location history information to estimate the extended passenger count measurement information instead of the location history information acquired by the acquisition unit. Traffic volume estimation device further comprising a convergence determination unit that repeatedly performs the estimation of the extended number of people and vehicles by the person and vehicle number estimation unit and the estimation of the extended location history information by the location estimation unit until predetermined conditions are met. [Application Example 4] A traffic volume estimation device according to any one of Application Examples 1 to 3, Traffic volume estimation device, wherein the predetermined condition is that the prediction accuracy calculated from the comparison result between the person-vehicle count measurement information acquired by the acquisition unit and the extended person-vehicle count measurement information estimated by the person-vehicle count estimation unit is equal to or greater than a predetermined threshold. [Application Example 5] A traffic volume estimation device according to any one of Application Examples 1 to 4, When the aforementioned position estimation unit refers to the sequence of position information as a movement path, Using the location history information, candidate movement paths are generated. Traffic volume estimation device that estimates the extended location history information by selecting a set of travel routes that match the number of people and vehicles measured from the generated candidate travel routes. [Application Example 6] A traffic volume estimation device according to any one of Application Examples 1 to 5, The traffic volume estimation device generates candidate travel routes by considering at least one of the following: map-like conditions, environmental conditions, temporal conditions, and event-related conditions. [Application Example 7] A method for estimating traffic volume, wherein an information processing device is An acquisition step of acquiring: person-vehicle count measurement information representing the number of people and / or vehicles measured by a measuring device installed at any point within the estimated target area, which measures the number of people and / or vehicles passing through that point; and location history information representing a series of actual location information that changed as a person-vehicle moved through the estimated target area; A position estimation step in which, using the acquired passenger vehicle count measurement information and the position history information, an extended position history information representing a series of position information that would have changed with the movement of each passenger vehicle is estimated for all passenger vehicles within the estimated target area. How to do it. [Application Example 8] A computer program, for use in an information processing device. An acquisition function that acquires: person-vehicle count measurement information representing the number of people and / or vehicles measured by a measuring device installed at any point within the estimated target area, which measures the number of people and / or vehicles passing through that point; and location history information representing a series of actual location information that changes with the movement of any person-vehicle moving around in the estimated target area; A position estimation function that uses the acquired passenger vehicle count measurement information and the position history information to estimate extended position history information representing the sequence of position information that would have changed with the movement of each passenger vehicle for all passenger vehicles within the estimated target area, A computer program that makes this possible. [Explanation of Symbols]

[0063] 1,1A,1B…Estimation System 10…CPU 11…Acquisition part 12,12A,12B…Position estimation part 13…Output section 14,14B…Person and vehicle number estimation section 15...Convergence determination unit 20...Storage section 21…Map Information Database 30... Communications Department 40…ROM / RAM 100, 100A, 100B... Traffic volume estimation device 200... Personnel / Vehicle Count Memory Unit 300...Location history memory unit

Claims

1. Traffic volume estimation device, An acquisition unit that acquires: person-vehicle count measurement information representing the number of people and / or vehicles measured by a measuring device installed at any point within the estimated target area, which measures the number of people and / or vehicles passing through that point; and location history information representing a series of actual location information that changes as a person-vehicle moves through the estimated target area; A position estimation unit that uses the acquired passenger vehicle count measurement information and the position history information to estimate extended position history information representing the sequence of position information that would have changed with the movement of each passenger vehicle for all passenger vehicles within the estimated target area, A traffic volume estimation device equipped with the following features.

2. A traffic volume estimation device according to claim 1, further, The system includes a passenger count estimation unit that estimates extended passenger count measurement information representing the number of passengers at all locations within the target area, including locations where the measuring device is not installed, using the acquired passenger count measurement information and location history information. Traffic volume estimation device wherein, if the position estimation unit has expanded passenger and vehicle count measurement information estimated by the passenger and vehicle count estimation unit, it estimates the expanded position history information using the expanded passenger and vehicle count measurement information instead of the passenger and vehicle count measurement information acquired by the acquisition unit.

3. A traffic volume estimation device according to claim 2, If the passenger count estimation unit has the extended location history information estimated by the location estimation unit, it will use the extended location history information to estimate the extended passenger count measurement information instead of the location history information acquired by the acquisition unit. Traffic volume estimation device further comprising a convergence determination unit that repeatedly performs the estimation of the extended number of people and vehicles by the person and vehicle number estimation unit and the estimation of the extended location history information by the location estimation unit until predetermined conditions are met.

4. A traffic volume estimation device according to claim 3, Traffic volume estimation device, wherein the predetermined condition is that the prediction accuracy calculated from the comparison result between the person-vehicle count measurement information acquired by the acquisition unit and the extended person-vehicle count measurement information estimated by the person-vehicle count estimation unit is equal to or greater than a predetermined threshold.

5. A traffic volume estimation device according to any one of claims 1 to 4, When the aforementioned position estimation unit refers to the sequence of position information as a movement path, Using the location history information, candidate movement paths are generated. Traffic volume estimation device that estimates the extended location history information by selecting a set of travel routes that match the number of people and vehicles measured from the generated candidate travel routes.

6. A traffic volume estimation device according to claim 5, The traffic volume estimation device generates candidate travel routes by considering at least one of the following: map-like conditions, environmental conditions, temporal conditions, and event-related conditions.

7. A method for estimating traffic volume, wherein an information processing device is An acquisition step of acquiring: person-vehicle count measurement information representing the number of people and / or vehicles measured by a measuring device installed at any point within the estimated target area, which measures the number of people and / or vehicles passing through that point; and location history information representing a series of actual location information that changed as a person-vehicle moved through the estimated target area; A position estimation step in which, using the acquired passenger vehicle count measurement information and the position history information, an extended position history information representing a series of position information that would have changed with the movement of each passenger vehicle is estimated for all passenger vehicles within the estimated target area. How to do it.

8. A computer program, for use in an information processing device. An acquisition function that acquires: person-vehicle count measurement information representing the number of people and / or vehicles measured by a measuring device installed at any point within the estimated target area, which measures the number of people and / or vehicles passing through that point; and location history information representing a series of actual location information that changes with the movement of any person-vehicle moving around in the estimated target area; A position estimation function that uses the acquired passenger vehicle count information and the position history information to estimate extended position history information representing the sequence of position information that would have changed with the movement of each passenger vehicle for all passenger vehicles within the estimated target area, A computer program that makes this possible.

Images