OD determination method and OD determination program

By employing a two-step traffic simulation process that adjusts OD traffic volume in coarse then fine time frames, the method efficiently estimates OD traffic volume, addressing the time-consuming nature of conventional methods.

JP2024170260APending Publication Date: 2024-12-06FUJITSU LTD
View PDF 1 Cites 0 Cited by

Patent Information

Application Number
JP2023087326
Authority / Receiving Office
JP · JP
Patent Type
Applications
Current Assignee / Owner
Filing Date
2023-05-26
Publication Date
2024-12-06

AI Technical Summary

Technical Problem

Conventional methods for estimating OD traffic volume require significant processing time due to the large matrix sizes involved, especially when fine time frames are considered.

Method used

The method involves performing a first traffic simulation for coarse time frames, adjusting the OD traffic volume to minimize errors, and then using the results to estimate OD traffic volume in finer time frames through a second simulation, thereby reducing the number of iterations and processing time.

Benefits of technology

This approach significantly reduces the processing time required for estimating OD traffic volume in fine time frames while maintaining accuracy by leveraging the results from coarser time frame simulations.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure 2024170260000001_ABST
    Figure 2024170260000001_ABST
Patent Text Reader

Abstract

To shorten a processing time for estimation of an OD traffic volume.SOLUTION: An information processing device 101 executes first simulation 110 to estimate a traffic volume of each link for each first time frame with the use of a first OD traffic volume q1. The information processing device 101 determines the OD traffic volume for each first time frame by repeating the first simulation 110 while adjusting the first OD traffic volume q1. The information processing device 101 calculates a second OD traffic volume q2 for each second time frame obtained by dividing the first time frame based on the determined OD traffic volume for each first time frame. The information processing device 101 executes a second simulation 120 for estimation of the traffic volume of each link with the use of the calculated second OD traffic volume q2. The information processing device 101 determines the OD traffic volume for each second time frame by repeating the second simulation 120 while adjusting the second OD traffic volume q2.SELECTED DRAWING: Figure 1
Need to check novelty before this filing date? Find Prior Art

Description

[Technical field]

[0001] The present invention relates to an OD determination method and an OD determination program. [Background technology]

[0002] In recent years, various transportation policies have been considered by local governments and companies to reduce traffic congestion in urban areas and cut CO2 emissions from transportation. For example, various traffic conditions are reproduced on a digital twin to estimate transportation demand and evaluate the effectiveness of policies. Transportation demand is expressed, for example, by OD traffic volume, such as the number of vehicles departing from an origin and arriving at a destination.

[0003] Prior art techniques include estimating OD traffic volumes based on probe information and statistical information, simulating traffic conditions based on road network information, signal information, and the estimated OD traffic volumes, correcting the OD traffic volumes so as to reduce the difference between the calculated link traffic volumes based on the simulated traffic conditions and the target link traffic volumes, and simulating traffic conditions using the corrected OD traffic volumes. [Prior art documents] [Patent documents]

[0004] [Patent Document 1] JP 2009-259158 A Summary of the Invention [Problem to be solved by the invention]

[0005] However, in the conventional technology, it may take a huge amount of processing time to estimate OD traffic volume (traffic movement demand).

[0006] In one aspect, the present invention aims to shorten the processing time required for estimating OD traffic volume. [Means for solving the problem]

[0007] In one embodiment, a first traffic simulation is performed for each first time frame obtained by dividing a target period, using a first OD traffic volume departing from a first point and arriving at a second point during the first time frame, to estimate a traffic volume for each link included in a route from the first point to the second point, and the first traffic simulation is repeatedly performed while adjusting the first OD traffic volume so that an error between the estimated traffic volume of the link and an observed traffic volume for the link is reduced, thereby determining an OD traffic volume for each first time frame for a pair of the first point and the second point, and calculating an OD traffic volume for each first time frame based on the determined OD traffic volume for each first time frame. the second OD traffic volume departing from the first point and arriving at the second point during the second time frame is calculated for each second time frame obtained by dividing the first time frame, the second traffic simulation is performed to estimate the traffic volume for each link for each second time frame using the calculated second OD traffic volume, and the second traffic simulation is repeatedly performed while adjusting the second OD traffic volume so that an error between the estimated traffic volume of the link and the traffic volume observed for the link is reduced, thereby determining the OD traffic volume for each second time frame for the pair of the first point and the second point. Effect of the Invention

[0008] Advantageous Effects of Invention According to one aspect of the present invention, an effect is achieved in that the processing time required for estimating OD traffic volume can be shortened. [Brief description of the drawings]

[0009] [Figure 1] FIG. 1 is an explanatory diagram showing an example of an OD determination method according to an embodiment. [Diagram 2] FIG. 2 is an explanatory diagram showing an example of a system configuration of an information processing system 200. As shown in FIG. [Diagram 3]FIG. 3 is a block diagram showing an example of the hardware configuration of the OD estimation device 201. [Figure 4] FIG. 4 is an explanatory diagram showing a specific example of a road network. [Diagram 5] FIG. 5 is an explanatory diagram showing a specific example of a road network table. [Figure 6] FIG. 6 is an explanatory diagram showing a specific example of a starting point table. [Figure 7] FIG. 7 is an explanatory diagram showing a specific example of a destination point table. [Figure 8] FIG. 8 is an explanatory diagram showing a specific example of the OD pair table. [Figure 9] FIG. 9 is an explanatory diagram showing a specific example of a routing table. [Figure 10] FIG. 10 is a block diagram showing an example of the functional configuration of the OD estimation device 201. [Figure 11] FIG. 11 is an explanatory diagram showing a calculation example of the second OD flow. [Figure 12] FIG. 12 is an explanatory diagram showing a first operation example of the OD estimation device 201. [Figure 13] FIG. 13 is an explanatory diagram showing a specific example of the initial OD flow. [Figure 14] FIG. 14 is an explanatory diagram (part 1) showing a specific example of a route selection table. [Figure 15] FIG. 15 is an explanatory diagram (part 1) showing a specific example of link flow (ratio). [Figure 16] FIG. 16 is an explanatory diagram (part 1) showing a specific example of link flow (traffic volume). [Figure 17] FIG. 17 is an explanatory diagram (part 1) showing an example of an observation matrix. [Figure 18] FIG. 18 is an explanatory diagram (part 1) showing a specific example of observable link flows (traffic volume). [Figure 19] FIG. 19 is an explanatory diagram (part 1) showing a specific example of an error in link flow (traffic volume). [Figure 20] FIG. 20 is an explanatory diagram (part 1) showing an example of updating the OD flow. [Figure 21]FIG. 21 is an explanatory diagram showing the estimation result of OD flow in a coarse time frame. [Figure 22] FIG. 22 is an explanatory diagram showing a second operation example of the OD estimation device 201. [Diagram 23] FIG. 23 is an explanatory diagram showing an example of OD flow distribution. [Figure 24] FIG. 24 is an explanatory diagram (part 2) showing a specific example of the route selection table. [Diagram 25] FIG. 25 is an explanatory diagram (part 2) showing a specific example of link flow (ratio). [Figure 26] FIG. 26 is an explanatory diagram (part 2) showing a specific example of link flow (traffic volume). [Figure 27] FIG. 27 is an explanatory diagram (part 2) showing an example of the observation matrix. [Figure 28] FIG. 28 is an explanatory diagram (part 2) showing a specific example of observable link flow (traffic volume). [Figure 29] FIG. 29 is an explanatory diagram (part 2) showing a specific example of an error in link flow (traffic volume). [Diagram 30] FIG. 30 is an explanatory diagram (part 2) showing an example of updating the OD flow. [Diagram 31] FIG. 31 is an explanatory diagram showing the estimation results of OD flows in a fine time frame. [Diagram 32] FIG. 32 is a flowchart (part 1) showing an example of the OD determination process procedure of the OD estimation device 201. [Diagram 33] FIG. 33 is a flowchart (part 2) showing an example of the OD determination process procedure of the OD estimation device 201. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0010] DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS Hereinafter, an embodiment of an OD determination method and an OD determination program according to the present invention will be described in detail with reference to the drawings.

[0011] (Embodiment) Fig. 1 is an explanatory diagram showing an example of an OD determination method according to an embodiment. In Fig. 1, an information processing device 101 is a computer that determines OD (Origin Destination) traffic volume for a pair of a departure point and a destination point for each specified time frame.

[0012] Here, OD traffic volume is an index that indicates traffic demand, and indicates the traffic volume departing from a departure point and arriving at a destination point in a certain time frame. OD traffic volume is estimated, for example, by using traffic simulation. A time frame is a time section (time slot) that divides a certain period into equal intervals. Traffic volume is expressed, for example, by the number of vehicles or the number of people.

[0013] Existing traffic demand estimation technologies include, for example, a technology (conventional technology 1) that uses traffic simulation to estimate OD traffic volume for each time frame. Traffic simulation reproduces traffic phenomena that occur on roads on a computer, for example, and simulates how traffic conditions such as traffic congestion and traffic disruptions change.

[0014] Prior Art 1 first performs a traffic simulation using OD traffic volumes given as initial values ​​to estimate the number of vehicles that have passed through each link on the route from the departure point to the destination point. Next, Prior Art 1 updates the OD traffic volumes based on the error between the estimated number of vehicles that have passed through the link and the number of vehicles that have passed through the link that was actually observed.

[0015] Then, Conventional Technology 1 performs traffic simulation again using the updated OD traffic volume. Conventional Technology 1 repeats this series of processes (traffic simulation → update of OD traffic volume → traffic simulation) until the error in the number of vehicles passing through the link falls within a predetermined range.

[0016] In this way, the conventional technique 1 estimates OD traffic volume for each time frame. The time frame is, for example, a time interval of 15 minutes or 30 minutes. For the conventional technique 1, see Non-Patent Document 1 below.

[0017] Non-patent document 1: Wei Ma, Xidong Pi, Sean Qian: Estimating multi-class dynamic origin-destination demand through a forward-backward algorithm on computational graphs, Transportation Research Part C 119 (2020) 102747.

[0018] However, in conventional technology 1, it may take a huge amount of processing time to estimate OD traffic volume for each time frame (e.g., 30 minutes, 15 minutes, etc.). In conventional technology 1, the size of the matrix handled in estimating OD traffic volume increases quadratically with the number of nodes, links, OD pairs, and travel routes in the road network. Furthermore, in conventional technology 1, when time is divided and OD traffic volume in finer time frames is required, the size of the matrix further increases by the square of the number of time frames, and it takes a huge amount of processing time to estimate OD traffic volume.

[0019] In this embodiment, an OD determination method for estimating OD traffic volume for each time frame is described, which reduces the number of iterations of traffic simulation to shorten the processing time required to obtain an optimal solution. Here, an example of processing by the information processing device 101 (corresponding to the following processes (1) to (5)) is described.

[0020] (1) The information processing device 101 uses the first OD traffic volume q1 to execute a first simulation 110. Here, the first simulation 110 is a traffic simulation that estimates the traffic volume of each link included in a route from a first point to a second point for each first time frame obtained by dividing a target period.

[0021] The first point corresponds to a starting point. The second point corresponds to a destination point. The route from the starting point to the destination point is represented, for example, by a road network. The road network is information about a road network, and is a network in which intersections are nodes and roads between the intersections are links.

[0022] The first OD traffic volume q1 represents the OD traffic volume departing from a first point and arriving at a second point during a first time frame for each first time frame obtained by dividing a target period. The first OD traffic volume q1 may be, for example, a predetermined constant, or may be represented by a randomly generated random number. The target period can be set arbitrarily, for example, to a period between 6:00 and 11:00. The time span of the first time frame can be set arbitrarily, for example, to one hour.

[0023] (2) The information processing device 101 repeatedly executes the first simulation 110 while adjusting the first OD traffic volume q1 so as to reduce the error between the estimated traffic volume of the link and the traffic volume observed for the link. In this way, the information processing device 101 determines the OD traffic volume for each first time frame for the pair of the first location and the second location.

[0024] Specifically, for example, the information processing device 101 calculates an error between the estimated traffic volume of a link and the traffic volume observed for the link. If the calculated error is greater than a threshold, the information processing device 101 updates the first OD traffic volume q1 based on the calculated error (reference numeral 111 in FIG. 1). The threshold can be set arbitrarily.

[0025] Then, the information processing device 101 uses the updated first OD traffic volume q1 to execute the first simulation 110. On the other hand, when the calculated error is equal to or smaller than the threshold, the information processing device 101 determines the first OD traffic volume q1 used in the immediately preceding first simulation 110 as the OD traffic volume for each first time frame.

[0026] (3) The information processing device 101 calculates, for each second time frame obtained by dividing the first time frame, a second OD traffic volume q2 that departs from the first point and arrives at the second point during the second time frame based on the determined OD traffic volume for each first time frame. The time width of the second time frame can be set arbitrarily according to the time width of the first time frame, and is set to, for example, 30 minutes.

[0027] For example, the target period is "6:00 to 11:00." The first time frame is each time section obtained by dividing "6:00 to 11:00" into one-hour segments. In this case, the second time frame is each time section obtained by dividing each of the first time frames into 30-minute segments (e.g., 6:00 to 6:30).

[0028] Specifically, for example, the information processing device 101 calculates the second OD traffic volume q2 by allocating the determined first OD traffic volume q1 to each second time frame according to a predetermined ratio. The predetermined ratio is set based on the ratio of the time width of the first time frame to the time width of the second time frame, for example.

[0029] For example, the time span of the first time frame is "60 minutes" and the time span of the second time frame is "30 minutes." In this case, the information processing device 101 calculates the second OD traffic volume q2 for each second time frame by allocating (dividing) the first OD traffic volume q1 for each first time frame at a ratio of "1:1" to (by two).

[0030] For example, the first OD traffic volume q1 in the first time frame of "6:00-7:00" is defined as "first OD traffic volume q1_1." Also, the second OD traffic volume q2 in the second time frame of "6:00-6:30" is defined as "second OD traffic volume q2_1." Also, the second OD traffic volume q2 in the second time frame of "6:30-7:00" is defined as "second OD traffic volume q2_2."

[0031] In this case, the first OD traffic volume q1_1 is allocated to the second time frame of "6:00-6:30" and the second time frame of "6:30-7:00" at a ratio of "1:1", and the second OD traffic volumes q2_1 and q2_2 are calculated. The second OD traffic volume q2_1 is "q2_1=q1_1×0.5". Also, the second OD traffic volume q2_2 is "q2_2=q1_0×0.5".

[0032] (4) The information processing device 101 executes a second simulation 120 using the calculated second OD traffic volume q2. Here, the second simulation 120 is a traffic simulation that estimates the traffic volume of each link included in the route from the first point to the second point for each second time frame obtained by dividing the first time frame.

[0033] (5) The information processing device 101 repeatedly executes the second simulation 120 while adjusting the second OD traffic volume q2 so as to reduce the error between the estimated traffic volume of the link and the traffic volume observed for the link. In this way, the information processing device 101 determines the OD traffic volume for each second time frame for the pair of the first location and the second location.

[0034] Specifically, for example, the information processing device 101 calculates an error between the estimated traffic volume of a link and the traffic volume observed for the link. If the calculated error is greater than a threshold, the information processing device 101 updates the second OD traffic volume q2 based on the calculated error.

[0035] Then, the information processing device 101 executes the second simulation 120 using the updated second OD traffic volume q2. On the other hand, when the calculated error is equal to or smaller than the threshold, the information processing device 101 determines the second OD traffic volume q2 used in the immediately preceding second simulation 120 as the OD traffic volume for each second time frame. The determined OD traffic volume for each second time frame is output as, for example, an OD estimation result.

[0036] In this way, the information processing device 101 can estimate the OD traffic volume in a finer time frame (second time frame) by using the OD traffic volume estimated in the coarse time frame (first time frame). This allows the information processing device 101 to reduce the number of iterations required to obtain an optimal solution and shorten the processing time when estimating the OD traffic volume in the finer time frames.

[0037] (System configuration example of information processing system 200) Next, a system configuration example of the information processing system 200 according to the embodiment will be described. In the following description, a case where the information processing device 101 shown in FIG.

[0038] Fig. 2 is an explanatory diagram showing an example of a system configuration of an information processing system 200. In Fig. 2, the information processing system 200 includes an OD estimation device 201 and a client device 202. In the information processing system 200, the OD estimation device 201 and the client device 202 are connected via a wired or wireless network 210. The network 210 is, for example, the Internet, a LAN (Local Area Network), or a WAN (Wide Area Network).

[0039] The OD estimation device 201 is, for example, a server. The client device 202 is a computer used by a user of the information processing system 200. The user is, for example, a person who considers transportation measures or evaluates the transportation measures. The client device 202 is, for example, a PC (Personal Computer), a tablet PC, etc.

[0040] In this embodiment, the OD estimation device 201 and the client device 202 are provided separately, but the present invention is not limited to this. For example, the OD estimation device 201 may be realized by the client device 202. Furthermore, the information processing system 200 may include a plurality of client devices 202.

[0041] (Hardware configuration example of OD estimation device 201) Next, an example of the hardware configuration of the OD estimation device 201 will be described.

[0042] Fig. 3 is a block diagram showing an example of the hardware configuration of the OD estimation device 201. In Fig. 3, the OD estimation device 201 has a CPU (Central Processing Unit) 301, a memory 302, a disk drive 303, a disk 304, a communication I / F (Interface) 305, a portable recording medium I / F 306, and a portable recording medium 307. In addition, each component is connected to each other by a bus 300.

[0043] Here, the CPU 301 is responsible for the overall control of the OD estimation device 201. The CPU 301 may have multiple cores. The memory 302 has, for example, a Read Only Memory (ROM), a Random Access Memory (RAM), and a flash ROM. Specifically, for example, the flash ROM stores an OS program, the ROM stores an application program, and the RAM is used as a work area for the CPU 301. The programs stored in the memory 302 are loaded into the CPU 301 to cause the CPU 301 to execute the coded processing.

[0044] The disk drive 303 controls reading / writing data from / to the disk 304 under the control of the CPU 301. The disk 304 stores data written under the control of the disk drive 303. The disk 304 is, for example, a magnetic disk, an optical disk, or the like.

[0045] The communication I / F 305 is connected to the network 210 through a communication line, and is connected to an external computer (e.g., the client device 202 shown in FIG. 2) via the network 210. The communication I / F 305 manages the interface between the network 210 and the inside of the device, and controls input and output of data from the external computer. The communication I / F 305 is, for example, a modem or a LAN adapter.

[0046] The portable recording medium I / F 306 controls reading / writing of data from / to the portable recording medium 307 under the control of the CPU 301. The portable recording medium 307 stores data written under the control of the portable recording medium I / F 306. The portable recording medium 307 is, for example, a CD (Compact Disc)-ROM, a DVD (Digital Versatile Disk), or a USB (Universal Serial Bus) memory.

[0047] In addition to the above-mentioned components, the OD estimation device 201 may have, for example, an input device, a display, etc. Also, the client device 202 shown in Fig. 2 can be realized by the same hardware configuration as the OD estimation device 201. However, in addition to the above-mentioned components, the client device 202 has, for example, an input device, a display, etc.

[0048] (Example of a road network) Next, the road network will be described. The road network corresponds to the roads for which traffic demand is to be estimated.

[0049] Fig. 4 is an explanatory diagram showing a specific example of a road network. In Fig. 4, a road network NW1 includes nodes N1 to N4 and links L1 to L3. Each of the nodes N1 to N4 represents an intersection. Each of the links L1 to L3 represents a road between the intersections.

[0050] In FIG. 4, node1 is the name (node ​​name) of node N1. node2 is the name of node N2. node3 is the name of node N3. node4 is the name of node N4. link1 is the name (link name) of link L1. link2 is the name of link L2. link3 is the name of link L3.

[0051] The road network NW1 is represented, for example, by a road network table 500 as shown in FIG.

[0052] Fig. 5 is an explanatory diagram showing a specific example of a road network table. In Fig. 5, a road network table 500 shows the network structure of the road network NW1 shown in Fig. 4. In the road network table 500, for each combination of two nodes included in the road network NW1, the name of the link connecting the nodes is shown. However, between nodes that are not connected, "None" is displayed.

[0053] For example, for the combination of node1 and node3, link1 is indicated, and it is possible to specify that nodes N1 and N3 are connected by link L1. Also, for the combination of node1 and node2, None is indicated, and it is possible to specify that nodes N1 and N2 are not connected.

[0054] (Example of starting point table) Next, a specific example of a departure point table used by the OD estimation device 201 will be described.

[0055] Fig. 6 is an explanatory diagram showing a specific example of a departure point table. In Fig. 6, a departure point table 600 shows node names "node1, node2, node3" which are O (Origin) in a road network NW1. O (Origin) represents a departure point. # in Fig. 6 is an identifier that identifies O (Origin).

[0056] (Example of a destination list) Next, a specific example of a destination point table used by the OD estimation device 201 will be described.

[0057] Fig. 7 is an explanatory diagram showing a specific example of a destination table. In Fig. 7, a destination table 700 shows a node name "node4" which is D (Destination) in a road network NW1. D (Destination) represents a destination point. # in Fig. 7 is an identifier that identifies D (Destination).

[0058] (Example of OD pair table) Next, a specific example of the OD pair table used by the OD estimation device 201 will be described.

[0059] Fig. 8 is an explanatory diagram showing a specific example of an OD pair table. In Fig. 8, an OD pair table 800 shows OD pairs in a road network NW1. An OD pair is a combination of O (Origin) and D (Destination). # in Fig. 8 is an identifier that identifies an OD pair.

[0060] The OD pair is identified, for example, from the departure point table 600 shown in Fig. 6 and the destination point table 700 shown in Fig. 7. For example, the OD pair 1 is a combination of O: node N1 (node1) and D: node N4 (node4).

[0061] (Example of a routing table) Next, a specific example of a routing table used by the OD estimation device 201 will be described.

[0062] Fig. 9 is an explanatory diagram showing a specific example of a routing table. In Fig. 9, a routing table 900 shows paths corresponding to OD pairs in a road network NW1. A path is a path from O (Origin) to D (Destination). # in Fig. 9 is an identifier that identifies a path.

[0063] For example, path1 is the route "node1-link1-link3-node4" from O: node N1 (node1) to D: node N4 (node4). Note that in the path, the display of nodes (e.g., node3) that are intermediate points is omitted.

[0064] (Example of functional configuration of OD estimation device 201) Next, an example of the functional configuration of the OD estimation device 201 will be described.

[0065] FIG. 10 is a block diagram showing an example of a functional configuration of the OD estimation device 201. In FIG. 10, the OD estimation device 201 includes an acquisition unit 1001, a setting unit 1002, a first execution unit 1003, a first determination unit 1004, a calculation unit 1005, a second execution unit 1006, a second determination unit 1007, and an output unit 1008. The acquisition unit 1001 to the output unit 1008 are functions that constitute the control unit 1000, and specifically, the functions are realized by having the CPU 301 execute a program stored in a storage device such as the memory 302, the disk 304, or the portable recording medium 307 shown in FIG. 3, or by the communication I / F 305. The processing results of each functional unit are stored in a storage device such as the memory 302 or the disk 304.

[0066] The acquisition unit 1001 acquires a road network table. Here, the road network table is information that indicates roads (for example, the road network NW1 shown in FIG. 4) for which traffic demand is to be estimated. Specifically, for example, the acquisition unit 1001 acquires a road network table 500 as shown in FIG. 5 by receiving it from the client device 202 shown in FIG. 2. The acquisition unit 1001 may also acquire the road network table 500 by a user's operation input using an input device (not shown).

[0067] The acquisition unit 1001 also acquires a departure point table and a destination point table. Here, the departure point table is information that indicates a departure point, for example, a node that is O (Origin) in a road network. The destination point table is information that indicates a destination point, for example, a node that is D (Destination) in a road network.

[0068] Specifically, for example, the acquiring unit 1001 acquires the departure point table 600 as shown in Fig. 6 and the destination point table 700 as shown in Fig. 7 by receiving from the client device 202. The acquiring unit 1001 may also acquire the departure point table 600 and the destination point table 700 by a user's operation input using an input device (not shown).

[0069] The acquiring unit 1001 also acquires an OD pair table. Here, the OD pair table is information that indicates a combination (OD pair) of O (Origin) and D (Destination). Specifically, for example, the acquiring unit 1001 acquires an OD pair table 800 as shown in FIG. 8 by creating it from the departure point table 600 and the destination point table 700.

[0070] Also, the acquiring unit 1001 may acquire the OD pair table 800, for example, by receiving it from the client device 202. Also, the acquiring unit 1001 may acquire the OD pair table 800 by a user's operation input using an input device (not shown).

[0071] The acquisition unit 1001 also acquires a routing table. Here, the routing table is information indicating a path from an origin O (origin) to a destination D (destination) corresponding to an OD pair in a road network. Specifically, for example, the acquisition unit 1001 acquires a routing table 900 as shown in FIG. 9 by creating the routing table 900 from the road network table 500 and the OD pair table 800.

[0072] Also, the acquisition unit 1001 may acquire the routing table 900 by, for example, receiving it from the client device 202. Also, the acquisition unit 1001 may acquire the routing table 900 by a user's operation input using an input device (not shown).

[0073] The acquisition unit 1001 also acquires a route selection table. Here, the route selection table is information indicating the probability (route selection probability) that a route (path) corresponding to an OD pair in a road network will be selected. Specifically, for example, the acquisition unit 1001 acquires the route selection table by receiving it from the client device 202. The acquisition unit 1001 may also acquire the route selection table by a user's operation input using an input device (not shown).

[0074] A specific example of the route selection table will be described later with reference to FIG.

[0075] The acquiring unit 1001 also acquires link flow observation data. Here, the link flow observation data is information indicating traffic volume actually observed for links in a road network. The traffic volume is represented, for example, by the number of vehicles passing through a link. The traffic volume is observed, for example, by using a sensor (for example, a vehicle counter) that is installed on a road and detects vehicles traveling on the road.

[0076] However, since it is difficult to install sensors on all links in a road network, sensors are installed on at least some of the links in the road network to observe the number of vehicles passing through those links.

[0077] Specifically, for example, the acquiring unit 1001 acquires the link flow observation data by receiving it from the client device 202. The acquiring unit 1001 may also acquire the link flow observation data by a user's operation input using an input device (not shown). The acquiring unit 1001 may also acquire the link flow observation data from a data providing service (not shown).

[0078] The setting unit 1002 sets an initial OD flow when executing a first traffic simulation. Here, the traffic simulation (for example, the first traffic simulation) is to reproduce traffic phenomena occurring on a road on a computer and to simulate how traffic conditions change. The traffic simulation is executed, for example, by using a traffic simulator.

[0079] The initial OD flow is the OD flow given as an initial value to the traffic simulation. The OD flow is information that represents the traffic volume departing from a departure point (first point) and arriving at a destination point (second point) in each time frame, and corresponds to the OD traffic volume.

[0080] In the traffic simulation, the OD flow is used to estimate the traffic volume of each link included in the route from the starting point to the destination point. The traffic volume is represented by, for example, the number of vehicles. The initial OD flow may be, for example, a predetermined constant, or may be represented by a randomly generated number.

[0081] Specifically, for example, the setting unit 1002 creates an OD flow for each first time frame into which the target period is divided, in which the number of vehicles that depart from a departure point and arrive at a destination point in the first time frame is expressed by a constant or a random number. Then, the setting unit 1002 sets the created OD flow as an initial OD flow.

[0082] Here, the target period is a period that is arbitrarily set in advance, and may be, for example, any time period in one day, or may be a period of two or more days. The first time frame is a period that is arbitrarily set in advance, and is set to a coarser time frame (longer time span) than the time frame of the OD flow that is ultimately to be estimated. The combination (OD pair) of the departure point (first point) and the destination point (second point) is identified, for example, from the OD pair table 800.

[0083] A specific example of the initial OD flow will be described later with reference to FIG.

[0084] The first execution unit 1003 sets the set initial OD flow as a first OD flow. The first OD flow corresponds to a first OD traffic volume. The first execution unit 1003 then executes a first traffic simulation using the first OD flow. The first traffic simulation is a traffic simulation that estimates the traffic volume of each link included in the route from a departure point (first point) to a destination point (second point) for each first time frame into which a target period is divided.

[0085] Specifically, for example, the first execution unit 1003 executes the first traffic simulation by providing a road network table (road network), an initial OD flow, a route table, and a route selection table to a traffic simulator. As the traffic simulator, for example, any existing simulator may be used.

[0086] This results in a link flow. The link flow is information that represents the estimated traffic volume of a link. For example, the link flow represents the traffic volume of each link included in the route from the starting point (first point) to the destination point (second point) for each first time frame that divides the target period.

[0087] A specific example of the link flow will be described later with reference to FIG.

[0088] The first determination unit 1004 repeatedly executes the first traffic simulation while adjusting the first OD flow so that the error between the estimated traffic volume of the link and the traffic volume observed for the link is reduced. In this way, the first determination unit 1004 determines the OD flow for each first time frame for the OD pair. The OD pair is a combination of a departure point (first point) and a destination point (second point). The traffic volume observed for the link is identified from the link flow observation data.

[0089] Specifically, for example, the first determination unit 1004 calculates an error between the estimated traffic volume (link flow) of a link and the traffic volume observed for the link. However, there are cases where sensors (e.g., vehicle counters) are installed only on some of the links in a road network, and traffic volumes are observed only on those links.

[0090] In this case, the first determination unit 1004 extracts links for which traffic volume (e.g., the number of vehicles) can be observed from the road network. For example, an observation matrix is ​​used to extract links for which traffic volume in a target period can be observed from among the links in the road network.

[0091] Next, the first determining unit 1004 extracts a link flow of the extracted link from among the link flows of each link estimated for each first time frame. Then, the first determining unit 1004 calculates an error between the link flow of the extracted link (estimated link flow) and the link flow observed for the link (actual link flow) for each first time frame.

[0092] Next, the first determination unit 1004 judges whether the calculated error is within the allowable range. To explain in more detail, for example, the first determination unit 1004 calculates the sum of the errors by adding up the calculated errors for each link. Then, when the calculated sum of the errors is equal to or less than the threshold value α, the first determination unit 1004 judges that the error is within the allowable range. On the other hand, when the sum of the errors is greater than the threshold value α, the first determination unit 1004 judges that the error is not within the allowable range. The threshold value α can be set arbitrarily.

[0093] Furthermore, the first determination unit 1004 may determine that the error is within the allowable range when the amount of variation in the calculated total error is equal to or less than a threshold value β. The amount of variation in the total error is specified by the ratio of the total error of the current time to the total error of the previous time when the first traffic simulation is repeatedly executed. On the other hand, the first determination unit 1004 determines that the error is not within the allowable range when the amount of variation in the total error is greater than the threshold value β. The threshold value β can be set arbitrarily, and is set to a value of, for example, about 10%.

[0094] Note that, although the case where the sum of errors is calculated has been described as an example here, this is not limiting. For example, the first determination unit 1004 may determine that the error is within the allowable range when each of the calculated errors for each link is equal to or less than the threshold value γ. On the other hand, the first determination unit 1004 may determine that the error is not within the allowable range when at least one of the calculated errors for each link is greater than the threshold value γ. The threshold value γ can be set arbitrarily.

[0095] Here, if the calculated error is not within the allowable range, the first determination unit 1004 updates the first OD flow based on the calculated error. More specifically, for example, the first determination unit 1004 calculates a correction value for the first OD flow based on the calculated error. Next, the first determination unit 1004 updates the first OD flow using the calculated correction value.

[0096] An example of updating the first OD flow will be described later with reference to FIG.

[0097] Then, the first determination unit 1004 executes a first traffic simulation using the updated first OD flow. Furthermore, if the calculated error is within an allowable range, the first determination unit 1004 determines the first OD flow used in the immediately preceding first traffic simulation as the OD flow for each first time frame.

[0098] A specific example of the determined OD flow for each first time frame will be described later with reference to FIG.

[0099] The calculation unit 1005 calculates, based on the determined OD flow for each first time frame, a second OD flow that departs from a departure point (first point) and arrives at a destination point (second point) in each second time frame, for each second time frame obtained by dividing the first time frame. The second time frame is, for example, a time frame obtained by dividing the first time frame in half.

[0100] Specifically, for example, the calculation unit 1005 calculates the second OD flow by allocating the determined OD flow for each first time frame to each second time frame according to a predetermined ratio. The predetermined ratio may be set based on the ratio of the duration of the first time frame to the duration of the second time frame.

[0101] The predetermined ratio may be set based on the ratio of traffic volume observed for a first link included in a route from a starting point (first point) to a destination point (second point) for each second time frame. For example, the first link is a link connected to a node at the starting point (first point).

[0102] A calculation example of the second OD flow will be described later with reference to FIG.

[0103] The second execution unit 1006 executes a second traffic simulation using the calculated second OD flow. The second traffic simulation is a traffic simulation that estimates the traffic volume of each link included in the route from the departure point (first point) to the destination point (second point) for each second time frame obtained by dividing the first time frame.

[0104] Specifically, for example, the second execution unit 1006 executes the second traffic simulation by providing the road network table (road network), the second OD flow, the route table, and the route selection table to the traffic simulator, thereby obtaining a link flow representing the traffic volume of each link included in the route from the starting point (first point) to the destination point (second point) for each second time frame.

[0105] The second determination unit 1007 repeatedly executes the second traffic simulation while adjusting the second OD flow so as to reduce the error between the estimated traffic volume of the link and the traffic volume observed for the link. In this way, the first determination unit 1004 determines the OD flow for each second time frame for the OD pair.

[0106] Specifically, for example, the second determining unit 1007 calculates an error between the estimated traffic volume (link flow) of a link and the traffic volume observed for the link. More specifically, for example, the second determining unit 1007 uses an observation matrix to extract links whose traffic volume can be observed from the road network.

[0107] Next, the second determining unit 1007 extracts the link flow of the extracted link from the link flows of each link estimated for each second time frame. Then, the second determining unit 1007 calculates the error between the link flow of the extracted link (estimated link flow) and the link flow observed for the link (actual link flow) for each second time frame.

[0108] Next, the second determination unit 1007 judges whether the calculated error is within an allowable range. If the calculated error is not within an allowable range, the second determination unit 1007 updates the second OD flow based on the calculated error. More specifically, for example, the second determination unit 1007 calculates a correction value for the second OD flow based on the estimated link flow of the link and the calculated error. Next, the second determination unit 1007 updates the second OD flow using the calculated correction value.

[0109] An example of updating the second OD flow will be described later with reference to FIG.

[0110] Then, the second determination unit 1007 executes a second traffic simulation using the updated second OD flow. Furthermore, if the calculated error is within an allowable range, the second determination unit 1007 determines the second OD flow used in the immediately preceding second traffic simulation as the OD flow for each second time frame.

[0111] A specific example of the determined OD flow for each second time frame will be described later with reference to FIG.

[0112] The output unit 1008 outputs the determined OD flow for each second time frame in association with the OD pair. An OD pair is a combination of a departure point (first point) and a destination point (second point). The output format of the output unit 1008 may be, for example, storage in a storage device such as the memory 302 or the disk 304, transmission to another computer via the communication I / F 305, display on a display (not shown), or printout on a printer (not shown).

[0113] Specifically, for example, the output unit 1008 transmits an OD flow 2202 (OD estimation result) as shown in Fig. 31 to the client device 202. This allows the user of the client device 202 to refer to the OD flow (e.g., the number of vehicles) for each second time frame for an OD pair (start point, destination point).

[0114] In addition, there may be cases where it is desired to estimate an OD flow with a time interval shorter than the second time frame. In this case, the second determination unit 1007 may determine whether the second time frame is a time frame designated in advance. The designation of the time frame is accepted from the client device 202, for example.

[0115] For example, suppose that the second time frame is a time frame with a 30-minute interval. Also, suppose that the specified time frame is a time frame with a 15-minute interval. In this case, the second determination unit 1007 determines that the second time frame is not the specified time frame. If it is determined that the second time frame is not the specified time frame, the calculation unit 1005 sets the OD flow for each determined second time frame as the OD flow for each first time frame.

[0116] The calculation unit 1005 may then calculate a second OD flow for each second time frame obtained by dividing the first time frame. This allows the OD estimation device 201 to estimate the OD flow for each specified time frame by, for example, repeatedly estimating the OD flow while dividing the time frame into smaller and smaller units.

[0117] (Example of calculation of the second OD flow) Next, a second example of calculating the OD flow will be described with reference to FIG.

[0118] Fig. 11 is an explanatory diagram showing a calculation example of the second OD flow. In Fig. 11, the target period is "6:00 to 11:00". In addition, the time width of the first time frame is "60 minutes", and the time width of the second time frame is "30 minutes".

[0119] In this case, the first time frame is, for example, "6:00 to 7:00" or "7:00 to 8:00". If the first time frame is "6:00 to 7:00", the second time frames are "6:00 to 6:30" and "6:30 to 7:00". If the first time frame is "7:00 to 8:00", the second time frames are "7:00 to 7:30" and "7:30 to 8:00".

[0120] Here, the first OD flow of a certain OD pair from "6:00 to 7:00" is "q1_0", and the first OD flow from "7:00 to 8:00" is "q2_0". In this case, if the second OD flow of the same OD pair from "6:00 to 6:30" is "q1_1", q1_1 is calculated using the following formula (1). Also, if the second OD flow of the same OD pair from "6:30 to 7:00" is "q2_1", q2_1 is calculated using the following formula (2).

[0121] Furthermore, if the second OD flow of the same OD pair from "7:00 to 7:30" is "q3_1", q3_1 is calculated using the following formula (3). If the second OD flow of the same OD pair from "7:30 to 8:00" is "q4_1", q4_1 is calculated using the following formula (4). Here, w1 is a constant greater than 0 and less than 1. w2 is a weight (constant) greater than 0 and less than 1.

[0122] q1_1=q1_0*w1 (1) q2_1=q1_0*(1.0-w1) (2) q3_1=q2_0*w2 (3) q4_1=q2_0*(1.0-w2) (4)

[0123] The weights w1 and w2 are set, for example, based on the time width between the first time frame and the second time frame. In the example of FIG. 11, the ratio of the time width between the first time frame and the second time frame is "2:1." In this case, the weight w1 is "w1=1÷2=0.5." Similarly, the weight w2 is "w2=1÷2=0.5."

[0124] In addition, the weights w1 and w2 may be set, for example, based on the ratio of traffic volumes observed for links included in the route from the starting point (first point) to the destination point (second point) for each second time frame.

[0125] Here, the starting point is node1 in the road network NW1, and the starting point is node4 in the road network NW1 (see FIG. 4). The route from node1 (O) to node4 (D) is "node1-link1-link3-node4".

[0126] For example, suppose that the first OD flow q1_0 in the first time frame "6:00~7:00" is "q1_0=10 cars." Also, suppose that the traffic volume (link flow) in the second time frame "6:00~6:30" observed by a sensor (vehicle counter) installed on link1 is "2 cars" and the traffic volume in the second time frame "6:30~7:00" is "8 cars."

[0127] The weight w1 of the second time frame "6:00-6:30" is, for example, "w1=2 / 10=0.2". In this case, the second OD flow q1_1 of the second time frame "6:00-6:30" is "10×0.2=2 cars" from the above formula (1). Also, the second OD flow q2_1 of the second time frame "6:30-7:00" is "10×(1-0.2)=8 cars" from the above formula (2).

[0128] (Example of OD flow estimation) Next, a description will be given of an example of OD flow estimation in the OD estimation device 201. Here, a description will be given of a case where OD flows are estimated in a finer time frame (second time frame) corresponding to each OD pair in the road network NW1 shown in Fig. 4 based on the road network table 500, departure point table 600, destination point table 700, OD pair table 800, and route table 900 shown in Figs. 5 to 9.

[0129] OD flow estimation on coarse time frames The OD estimation device 201 estimates the OD flow in a rough time frame (first time frame) by a first traffic simulation. Here, the rough time frame (first time frame) is a time frame of every 60 minutes. For convenience of explanation, the target period is "h1:6:00-7:00", and an example is given in which only one time frame (first time frame) is included in the target period. The time frame "h1:6:00-7:00" may be simply referred to as time frame h1.

[0130] The specific processing contents when estimating the OD flow in the coarse time frame (first time frame) will be described below.

[0131] Fig. 12 is an explanatory diagram showing a first operation example of the OD estimation device 201. Fig. 12 shows operations that the OD estimation device 201 repeatedly performs when estimating an OD flow in a coarse time frame (first time frame).

[0132] First, as a preliminary step, the OD estimation device 201 sets an initial OD flow 1201. The initial OD flow 1201 is an OD flow given as an initial value to the first traffic simulation, and indicates an OD flow (for example, a random number or a constant) for each 60 minutes corresponding to each OD pair.

[0133] Fig. 13 is an explanatory diagram showing a specific example of an initial OD flow. In Fig. 13, an initial OD flow 1201 shows the OD flows in time frame h1 corresponding to each of OD pairs 1 to 3. Each OD flow indicates the number of vehicles. For example, "2" indicated by the reference numeral 1301 indicates the number of vehicles (OD flow) departing from node 2 and arriving at node 4 in time frame h1.

[0134] Next, the OD estimation device 201 sets the initial OD flow 1201 as an OD flow q (first OD flow) in the Forward process 1210 (see FIG. 12). Then, the OD estimation device 201 executes a first traffic simulation using the OD flow q.

[0135] Specifically, for example, the OD estimation device 201 executes a first traffic simulation by providing the road network table 500, the OD flow q (initial OD flow 1201), the route table 900, the OD pair table 800, and the route selection table (see FIG. 14) to the traffic simulator.

[0136] Here, a specific example of the routing table will be described with reference to FIG.

[0137] Fig. 14 is an explanatory diagram (part 1) showing a specific example of a route selection table. In Fig. 14, a route selection table 1400 shows the probability (route selection probability) that a route (path) corresponding to each OD pair in the road network NW1 will be selected during the time frame h1: 6:00 to 7:00. # in Fig. 14 is an identifier that identifies a route (path).

[0138] For example, O.D.<node1,node4> The probability that path1 (node1-link1-link3-node4) corresponding to the above will be selected is 1.0. The route selection probability for each route depends on the time of day. For example, even for the same route, the route selection probability may be different between 6am and 12pm.

[0139] In the example of Figure 14, there is only one route corresponding to each OD pair. Therefore, for each OD pair, there is no other route that can be selected when moving from O to D, and the route selection probability of each route is "1.0". However, there may be multiple routes for the same OD pair. In this case, the sum of the route selection probabilities of the multiple routes is "1.0".

[0140] In the first traffic simulation, for example, virtual vehicles depart from the departure point every five minutes between "6:00 and 7:00" and travel toward the destination point. However, it is assumed that vehicles that depart within the same time frame reach the destination point within the same time frame. For example, vehicles that depart from node1 (departure point) between "6:00 and 7:00" all reach node4 (destination point) by "7:00". It is assumed that there is no congestion.

[0141] When the first traffic simulation is completed, an estimation result of link flow (traffic volume) for each link included in the path corresponding to each OD pair in the time frame h1 is obtained. The link flow is, for example, the number of vehicles passing through the link.

[0142] Here, a specific example of the link flow estimated by the first traffic simulation will be described with reference to FIGS.

[0143] Fig. 15 is an explanatory diagram (part 1) showing a specific example of link flow (ratio). In Fig. 15, link flow 1500 is information (ρ) obtained by a first traffic simulation, and represents the link flow (ratio) for each link included in a path corresponding to each OD pair in a time frame h1: 6:00 to 7:00.

[0144] The link flow (ratio) indicates the ratio of vehicles that pass through each link in time frame h1 among vehicles traveling from a departure point to a destination point via a path corresponding to each OD pair. The link flow (ratio) is, for example, a value according to the route selection probability of a path corresponding to each OD pair in the route selection table 1400 shown in FIG.

[0145] For example, "1.0" indicated by reference numeral 1501 indicates the ratio of vehicles passing through link 1 among vehicles traveling from node 1 to node 4 via path 1 in time frame h1. Also, "0.0" indicated by reference numeral 1502 indicates the ratio of vehicles passing through link 2 among vehicles traveling from node 3 to node 4 via path 3 in time frame h1.

[0146] The OD estimation device 201 calculates the link flow (traffic volume) based on the link flow 1500 using, for example, the following formula (5). Here, x indicates the link flow (traffic volume), ρ indicates the link flow (ratio), p indicates the route selection probability, and q indicates the OD flow.

[0147] x = ρpq (5)

[0148] Here, if ρ is the link flow 1500 shown in FIG. 15, then x is calculated as in the following formula (6), resulting in a link flow 1600 as shown in FIG.

[0149]

number

[0150] Fig. 16 is an explanatory diagram (part 1) showing a specific example of link flow (traffic volume). In Fig. 16, link flow 1600 is information (x) estimated by a first traffic simulation, and represents link flow (traffic volume) for each link included in a path corresponding to each OD pair in a time frame h1: 6:00 to 7:00.

[0151] The link flow (traffic volume) indicates the number of vehicles that passed through each link in time frame h1. For example, "2" indicated by the reference numeral 1601 indicates the number of vehicles that passed through link2 in time frame h1. This allows the OD estimation device 201 to estimate the traffic volume (number of vehicles) on all links included in the path corresponding to each OD pair.

[0152] Although the traffic volume of all links can be calculated on the simulator, in the real world, it is difficult to install sensors (such as vehicle counters) on all links (roads) on a road network. For this reason, the OD estimation device 201 extracts the traffic volume of links where the traffic volume can be actually observed.

[0153] Here, it is assumed that sensors (vehicle counters) are installed only on link1 and link3 in the road network NW1. For this reason, the OD estimation device 201 extracts links where traffic volume can be observed from the road network NW1 using an observation matrix 1700 as shown in FIG.

[0154] Fig. 17 is an explanatory diagram (part 1) showing an example of an observation matrix. In Fig. 17, an observation matrix 1700 is a matrix for extracting link1 and link3 whose traffic volume is observable in time frame h1 from among link1, link2, and link3 in road network NW1. For example, "1" indicated by reference numeral 1701 is a flag indicating that the traffic volume (number of vehicles) passing through link1 is observable in time frame h1.

[0155] 17, b11 corresponds to information for extracting link1 in time frame h1, and b12 corresponds to information for extracting link3 in time frame h1.

[0156] The OD estimation device 201 extracts link flows (traffic volume) of link1 and link3, which are observable in time frame h1, from among link1, link2, and link3 in the road network NW1, using, for example, the following formula (7). Here, y indicates the observable link flow (traffic volume), L indicates the observation matrix, and x indicates the link flow (traffic volume) calculated on the simulator.

[0157] y = Lx (7)

[0158] Here, if x is the link flow 1600 shown in FIG. 16 and L is the observation matrix 1700 shown in FIG. 17, y is calculated as shown in equation (8) below, and link flow 1800 as shown in FIG. 18 is extracted.

[0159]

number

[0160] FIG. 18 is an explanatory diagram (part 1) showing a specific example of observable link flows (traffic volume). In FIG. 18, link flow 1800 represents the link flows (traffic volume) of link1 and link3 actually observed during time frame h1: 6:00 to 7:00. For example, "6" indicated by reference numeral 1801 indicates the number of vehicles that passed through link2 during time frame h1. # in FIG. 18 is an identifier that identifies the observed link flows (traffic volume).

[0161] Next, the OD estimation device 201 calculates the error between the link flow (traffic volume) calculated (estimated) on the simulator and the actually observed link flow (traffic volume) using, for example, the following formula (9). Here, Δy indicates the error of the link flow (traffic volume), y' indicates the actually observed link flow (traffic volume), and y indicates the link flow (traffic volume) calculated on the simulator.

[0162] Δy = y' - y (9)

[0163] Here, if y is the link flow 1800 shown in Figure 18, Δy is calculated as shown in Figure 19. y' is identified from the link flow observation data.

[0164] Fig. 19 is an explanatory diagram (part 1) showing a specific example of a link flow (traffic volume) error. In Fig. 19, a link flow (traffic volume) error 1900 represents the link flow (traffic volume) error of link1 and link3, where traffic volume can be observed, during time frame h1: 6:00 to 7:00. Note that a link flow 1901 is an actually observed link flow (traffic volume), and is identified from link flow observation data.

[0165] Here, it is assumed that the error 1900 of the calculated link flow (traffic volume) is not within a predetermined allowable range.

[0166] In this case, the OD estimation device 201 updates the OD flow q in the backward process 1220 (see FIG. 12) based on the link flow (traffic volume) error 1900. The OD flow q before the update at this point is the initial OD flow 1201 shown in FIG.

[0167] Specifically, for example, the OD estimation device 201 first calculates a correction value for the OD flow q using the following formula (10). Here, Δq indicates the correction value for the OD flow q. w indicates a weight (constant). For example, w is a value between 0 and 1. T denotes the transpose of p. T denotes the transpose of ρ. L T denotes the transposed matrix of L. Δy denotes the error in link flow (traffic volume).

[0168] Δq=wp T ρ T L T Δy (10)

[0169] Here, w is "w=0.5", p is the route selection table 1400 shown in Fig. 14, ρ is the link flow 1500 shown in Fig. 15, L is the observation matrix 1700 shown in Fig. 17, and Δy is the link flow (traffic volume) error 1900 shown in Fig. 19. In this case, Δq is calculated as in the following formula (11).

[0170]

number

[0171] Next, the OD estimation device 201 updates the OD flow q, for example, by using the following equation (12).

[0172] q = Δq + q (12)

[0173] Here, if Δq is the value calculated as in equation (11) above, the OD flow q is updated as shown in FIG.

[0174] Fig. 20 is an explanatory diagram (part 1) showing an example of updating an OD flow. In Fig. 20, an OD flow 2000 shows an updated OD flow q obtained by updating the initial OD flow 1201 shown in Fig. 13 using Δq (corrected value 2001 of OD flow q) calculated as in the above formula (10).

[0175] Next, the OD estimation device 201 sets the OD flow 2000 as OD flow q (first OD flow). Then, the OD estimation device 201 returns to the Forward process 1210 (see FIG. 12) and executes a first traffic simulation using the OD flow q.

[0176] The OD estimation device 201 repeats the forward process 1210 and the backward process 1220 until the error of the calculated link flow (traffic volume) falls within an allowable range. Here, it is assumed that the OD flow 1202 is estimated as a result of performing the repetition four times.

[0177] Fig. 21 is an explanatory diagram showing an estimation result of an OD flow in a coarse time frame. In Fig. 21, an OD flow 1202 shows an estimation result of an OD flow in a coarse time frame (first time frame). The coarse time frame (first time frame) is a time frame of every 60 minutes.

[0178] -OD flow estimation on fine time frames Next, the OD estimation device 201 estimates the OD flow in a finer time frame (second time frame) by a second traffic simulation. Here, the finer time frame (second time frame) is a time frame of every 30 minutes. In addition, finer time frames obtained by dividing the coarse time frame "6:00-7:00" at 30-minute intervals may be expressed as "h1:6:00-6:30" and "h2:6:30-7:00". In addition, the time frame "h1:6:00-6:30" may be simply expressed as time frame h1, and the time frame "h2:6:30-7:00" may be simply expressed as time frame h2.

[0179] The specific processing contents when estimating the OD flow in the finer time frame (second time frame) will be described below.

[0180] Fig. 22 is an explanatory diagram showing a second operation example of the OD estimation device 201. Fig. 22 shows operations that the OD estimation device 201 repeatedly performs when estimating an OD flow in a fine time frame (second time frame).

[0181] First, the OD estimation device 201 sets an OD flow 2201 to be given as an initial value to the second traffic simulation, based on the OD flow 1202 shown in FIG. 21 (OD flow in a coarse time frame).

[0182] Specifically, for example, the OD estimation device 201 calculates an OD flow 2201 by allocating the OD flow 1202 (OD flow in a coarse time frame) shown in Fig. 21 to each fine time frame (second time frame). Here, it is assumed that the weight is 0.5 and the OD flow in the coarse time frame is allocated to the fine time frame (second time frame) at a ratio of "1:1".

[0183] Fig. 23 is an explanatory diagram showing an example of OD flow allocation. In Fig. 23, an OD flow 2201 indicates an OD flow calculated by allocating the OD flow 1202 (see Fig. 21) in a coarse time frame to each fine time frame.

[0184] For example, "315.5" indicated by reference numeral 2301 indicates the number of vehicles (OD flow) that depart from node 2 in time frame h1 and arrive at node 4. Also, "315.5" indicated by reference numeral 2302 indicates the number of vehicles (OD flow) that depart from node 3 in time frame h2 and arrive at node 4.

[0185] Next, the OD estimation device 201 sets the OD flow 2201 as an OD flow q (a second OD flow) in a forward process 2210 (see FIG. 22). Then, the OD estimation device 201 executes a second traffic simulation using the OD flow q.

[0186] Specifically, for example, the OD estimation device 201 executes the second traffic simulation by providing the road network table 500, the OD flow q (OD flow 2201), the route table 900, the OD pair table 800, and the route selection table (see FIG. 24) to the traffic simulator.

[0187] Fig. 24 is an explanatory diagram (part 2) showing a specific example of a route selection table. In Fig. 24, a route selection table 2400 shows the probability (route selection probability) that a route (path) corresponding to each OD pair in the road network NW1 will be selected during each time frame h1: 6:00 to 6:30, h2: 6:30 to 7:00. # in Fig. 24 is an identifier that identifies a route (path).

[0188] For example, in the time frame h1:6:00~6:30, OD<node1,node4> The probability that path1 (node1-link1-link3-node4) corresponding to OD is selected is 1.0.<node3,node4> The probability that path3 (node3-link3-node4) corresponding to this will be selected is "1.0".

[0189] In the second traffic simulation, for example, virtual vehicles depart from the departure point every five minutes between "6:00 and 7:00" and travel toward the destination point. However, it is assumed that vehicles that depart within the same time frame reach the destination point within the same time frame. For example, vehicles that depart from node1 (departure point) between "6:00 and 7:00" all reach node4 (destination point) by "7:00". It is assumed that there is no congestion.

[0190] When the second traffic simulation is completed, an estimation result of link flow (traffic volume) for each link included in the path corresponding to each OD pair in each time frame h1, h2 is obtained.

[0191] Here, a specific example of the link flow estimated by the second traffic simulation will be described with reference to FIGS.

[0192] Fig. 25 is an explanatory diagram (part 2) showing a specific example of link flow (ratio). In Fig. 25, link flow 2500 is information (ρ) obtained by the second traffic simulation, and represents the link flow (ratio) for each link included in the path corresponding to each OD pair in each time frame h1: 6:00 to 6:30, h2: 6:30 to 7:00.

[0193] The link flow (ratio) is, for example, a value according to the route selection probability of a path corresponding to each OD pair in the route selection table 2400 shown in FIG.

[0194] For example, "1.0" indicated by reference numeral 2501 indicates the ratio of vehicles passing through link 3 in time frame h1 among vehicles traveling from node 2 to node 4 via path 2 in time frame h1. Also, "0.0" indicated by reference numeral 2502 indicates the ratio of vehicles passing through link 2 in time frame h2 among vehicles traveling from node 1 to node 4 via path 1 in time frame h1.

[0195] The OD estimation device 201 calculates the link flow (traffic volume) based on the link flow 2500 using, for example, the above formula (5). If ρ is the link flow 1500 shown in FIG. 25, then x is calculated as shown in the following formula (13), resulting in the link flow 2600 shown in FIG.

[0196]

number

[0197] Fig. 26 is an explanatory diagram (part 2) showing a specific example of link flow (traffic volume). In Fig. 26, a link flow 2600 is information (x) estimated by a second traffic simulation, and represents the link flow (traffic volume) for each link included in a path corresponding to each OD pair in each time frame h1: 6:00 to 6:30, h2: 6:30 to 7:00.

[0198] For example, "315.5" indicated by reference numeral 2601 indicates the number of vehicles that passed through link 2 in time frame h1. Also, "916.5" indicated by reference numeral 2602 indicates the number of vehicles that passed through link 3 in time frame h2. This allows the OD estimation device 201 to estimate the traffic volume (number of vehicles) on all links included in the path corresponding to each OD pair.

[0199] Next, the OD estimation device 201 extracts the traffic volume of links where the traffic volume can be actually observed. Here, it is assumed that sensors (vehicle counters) are installed only on link1 and link3 in the road network NW1. For this reason, the OD estimation device 201 extracts links where the traffic volume can be observed from the road network NW1 using an observation matrix 2700 as shown in FIG.

[0200] Fig. 27 is an explanatory diagram (part 2) showing an example of an observation matrix. In Fig. 27, an observation matrix 2700 is a matrix for extracting link1 and link3 whose traffic volume can be observed in each time frame h1 and h2 from among link1, link2, and link3 in a road network NW1.

[0201] For example, the "1" indicated by the reference numeral 2701 is a flag indicating that traffic volume (number of vehicles) passing through link1 can be observed in time frame h1. The "1" indicated by the reference numeral 2702 is a flag indicating that traffic volume passing through link3 can be observed in time frame h1. The "1" indicated by the reference numeral 2703 is a flag indicating that traffic volume passing through link1 can be observed in time frame h2. The "1" indicated by the reference numeral 2704 is a flag indicating that traffic volume passing through link3 can be observed in time frame h2.

[0202] 27, b11 corresponds to information for extracting link1 in time frame h1. b12 corresponds to information for extracting link3 in time frame h1. b21 corresponds to information for extracting link1 in time frame h2. b22 corresponds to information for extracting link3 in time frame h2.

[0203] The OD estimation device 201 extracts link flows (traffic volumes) of link1 and link3 whose traffic volumes can be observed in each of time frames h1 and h2 from among link1, link2, and link3 in the road network NW1, using, for example, the above formula (7).

[0204] Here, if x is the link flow 2600 shown in FIG. 26 and L is the observation matrix 2700 shown in FIG. 27, y is calculated as shown in equation (14) below, and link flow 2800 as shown in FIG. 28 is extracted.

[0205]

number

[0206] Fig. 28 is an explanatory diagram (part 2) showing a specific example of observable link flows (traffic volume). In Fig. 28, a link flow 2800 represents the link flows (traffic volume) of link1 and link3 actually observed in each time frame h1: 6:00 to 6:30, and h2: 6:30 to 7:00.

[0207] For example, "916.5" indicated by the reference numeral 2801 indicates the number of vehicles that passed through link3 in time frame h1. # in Fig. 28 is an identifier for identifying the observed link flow (traffic volume).

[0208] Next, the OD estimation device 201 calculates the error between the link flow (traffic volume) calculated (estimated) on the simulator and the actually observed link flow (traffic volume), for example, using the above formula (9).

[0209] Here, if y is the link flow 2800 shown in Figure 28, Δy is calculated as shown in Figure 29. y' is identified from the link flow observation data.

[0210] Fig. 29 is an explanatory diagram (part 2) showing a specific example of a link flow (traffic volume) error. In Fig. 29, a link flow (traffic volume) error 2900 represents the link flow (traffic volume) error of link1 and link3, where traffic volume can be observed, in each of time frames h1: 6:00 to 6:30 and h2: 6:30 to 7:00. Note that a link flow 2901 is an actually observed link flow (traffic volume), and is identified from link flow observation data.

[0211] Here, it is assumed that the error 2900 of the calculated link flow (traffic volume) is not within a predetermined allowable range.

[0212] In this case, the OD estimation device 201 updates the OD flow q in the backward process 2220 (see FIG. 22) based on the link flow (traffic volume) error 2900. The OD flow q before the update at this point is the OD flow 2201 shown in FIG.

[0213] Specifically, for example, the OD estimation device 201 first calculates a correction value of the OD flow q using the above formula (10). Here, w is set to "w=0.5", p is set to the route selection table 2400 shown in Fig. 24, ρ is set to the link flow 2500 shown in Fig. 25, L is set to the observation matrix 2700 shown in Fig. 27, and Δy is set to the link flow (traffic volume) error 2900 shown in Fig. 29. In this case, Δq is calculated as shown in the following formula (15).

[0214]

number

[0215] Next, the OD estimation device 201 updates the OD flow q using, for example, the above equation (12).

[0216] Here, if Δq is the value calculated as in equation (15) above, the OD flow q is updated as shown in FIG.

[0217] Fig. 30 is an explanatory diagram (part 2) showing an example of updating the OD flow. In Fig. 30, the OD flow 3000 shows the OD flow q after updating the OD flow 2201 shown in Fig. 23 using Δq (correction value 3001 of the OD flow q) calculated as in the above formula (14).

[0218] Next, the OD estimation device 201 sets the OD flow 3000 as OD flow q (second OD flow). Then, the OD estimation device 201 returns to the Forward process 2210 (see FIG. 22) and executes a second traffic simulation using the OD flow q.

[0219] The OD estimation device 201 repeats the forward process 2210 and the backward process 2220 until the error of the calculated link flow (traffic volume) falls within an allowable range. Here, it is assumed that the OD flow 2202 is estimated as a result of performing the repetition 10 times.

[0220] Fig. 31 is an explanatory diagram showing the estimation result of the OD flow in a fine time frame. In Fig. 31, an OD flow 2202 shows the estimation result of the OD flow in a fine time frame (second time frame). The fine time frame (second time frame) is a time frame of every 30 minutes.

[0221] This allows the OD estimation device 201 to reduce the processing time required for estimating the OD flow, and estimate the OD flow in a fine time frame (second time frame).

[0222] (OD determination process procedure of the OD estimation device 201) Next, the OD determination process procedure of the OD estimation device 201 will be described with reference to FIG. 32 and FIG.

[0223] 32 and 33 are flowcharts showing an example of an OD determination processing procedure of the OD estimation device 201. In the flowchart of Fig. 32, first, the OD estimation device 201 sets an initial OD flow when executing a first traffic simulation (step S3201).

[0224] Next, the OD estimation device 201 sets the width of the first time frame (time span) in the first traffic simulation (step S3202). The width of the first time frame is set to the width of the time frame of the initial OD flow.

[0225] Then, the OD estimation device 201 sets the initial OD flow to the first OD flow (step S3203), and executes a first traffic simulation using the first OD flow (step S3204). Next, the OD estimation device 201 calculates the link flow (traffic volume) for each first time frame from the link flow (ratio) for each first time frame obtained by the first traffic simulation (step S3205).

[0226] Then, the OD estimation device 201 extracts the link flow (traffic volume) of the link where the traffic volume can be observed from the calculated link flow (traffic volume) (step S3206). Next, the OD estimation device 201 calculates the error between the link flow (traffic volume) of the extracted link and the link flow (traffic volume) actually observed for the link (step S3207). The actually observed link flow (traffic volume) is identified from the link flow observation data.

[0227] Then, the OD estimation device 201 judges whether the error of the calculated link flow (traffic volume) is within the allowable range (step S3208). If the error is not within the allowable range (step S3208: No), the OD estimation device 201 calculates a correction value of the first OD flow based on the calculated error (step S3209).

[0228] Next, the OD estimation device 201 updates the first OD flow using the calculated correction value (step S3210). Then, the OD estimation device 201 returns to step S3203, sets the updated first OD flow as the first OD flow, and executes the subsequent processes.

[0229] Furthermore, in step S3208, if the error of the calculated link flow (traffic volume) is within the allowable range (step S3208: Yes), the OD estimation device 201 determines the first OD flow used in the immediately preceding first traffic simulation as the OD flow for the first time frame (step S3211).

[0230] Then, the OD estimation device 201 calculates a second OD flow by distributing the determined OD flow for each first time frame to each second time frame into which the first time frame is divided (step S3212), and proceeds to step S3301 shown in FIG. 33.

[0231] In the flowchart of Fig. 33, first, the OD estimation device 201 sets an initial OD flow when executing the second traffic simulation (step S3301). The initial OD flow is set to the second OD flow calculated in step S3212 shown in Fig. 32.

[0232] Next, the OD estimation device 201 sets the width of a second time frame in the second traffic simulation (step S3302).Then, the OD estimation device 201 sets the initial OD flow to the second OD flow (step S3303), and executes the second traffic simulation using the second OD flow (step S3304).

[0233] Next, the OD estimation device 201 calculates the link flow (traffic volume) for each second time frame from the link flow (ratio) for each second time frame obtained by the second traffic simulation (step S3305).Then, the OD estimation device 201 extracts the link flow (traffic volume) of the link where the traffic volume can be observed from the calculated link flow (traffic volume) (step S3306).

[0234] Next, the OD estimation device 201 calculates an error between the link flow (traffic volume) of the extracted link and the link flow (traffic volume) actually observed for the link (step S3307).Then, the OD estimation device 201 judges whether the error of the calculated link flow (traffic volume) is within an allowable range (step S3308).

[0235] If the error is not within the allowable range (step S3308: No), the OD estimation device 201 calculates a correction value for the second OD flow based on the calculated error (step S3309). Next, the OD estimation device 201 updates the second OD flow using the calculated correction value (step S3310).

[0236] Then, the OD estimation device 201 returns to step S3303, sets the updated second OD flow as the second OD flow, and executes the subsequent processes.

[0237] Furthermore, in step S3308, if the error in the calculated link flow (traffic volume) is within the allowable range (step S3308: Yes), the OD estimation device 201 determines the second OD flow used in the immediately preceding second traffic simulation as the OD flow for the second time frame (step S3311).

[0238] Then, the OD estimation device 201 outputs the determined OD flow for each second time frame (step S3312), and ends the series of processes according to this flowchart.

[0239] This allows the OD estimation device 201 to estimate the OD flow in a finer time frame (second time frame) by utilizing the OD flow estimated in the coarse time frame (first time frame).

[0240] As described above, the OD estimation device 201 according to the embodiment can execute a first traffic simulation for estimating traffic volume for each link included in a route from a first point to a second point using a first OD flow for each first time frame into which a target period is divided. The first OD flow represents traffic volume (e.g., the number of vehicles) departing from the first point and arriving at the second point in the first time frame. Then, the OD estimation device 201 can determine an OD flow for each first time frame for a pair (OD pair) of a first point and a second point by repeatedly executing the first traffic simulation while adjusting the first OD flow so as to reduce an error between the estimated traffic volume of the link and the traffic volume observed for the link.

[0241] This allows the OD estimation device 201 to estimate OD flows for an OD pair of a departure point (O) and a destination point (D) in a time frame (first time frame) that is coarser than the time frame that is originally desired to be estimated. In OD flow estimation in a coarse time frame, even if the number of iterations required to obtain an optimal solution increases, the processing time required for one iteration is expected to be short because the matrix size (L, ρ, p, q) to be handled is relatively small.

[0242] Furthermore, the OD estimation device 201 can calculate, for each second time frame obtained by dividing the first time frame, a second OD flow that departs from a first point and arrives at a second point during the second time frame, based on the determined OD flow for each first time frame. The calculated second OD flow corresponds to an initial OD flow of the second traffic simulation. Furthermore, the OD estimation device 201 can execute a second traffic simulation that estimates traffic volume for each link for each second time frame, using the calculated second OD flow. The OD estimation device 201 can determine an OD flow for each second time frame for a pair of a first point and a second point by repeatedly executing the second traffic simulation while adjusting the second OD flow so that the error between the estimated traffic volume of the link and the traffic volume observed for the link is reduced.

[0243] In this way, the OD estimation device 201 can estimate the OD flow in the time frame that is originally desired to be estimated (second time frame) for an OD pair of a departure point (O) and a destination point (D) by using the OD flow in the coarse time frame (first time frame). The OD estimation device 201 can reduce the number of iterations required to obtain an optimal solution by using the OD flow estimated in the coarse time frame as the initial value for OD flow estimation in the fine time frame.

[0244] Moreover, according to the OD estimation device 201, the second OD flow can be calculated by allocating the determined OD flow for each first time frame to each second time frame according to a predetermined ratio.

[0245] This allows the OD estimation device 201 to distribute and utilize the OD flow for each first time frame to each of the second time frames divided from the first time frame.

[0246] Furthermore, the OD estimation device 201 can output the determined OD flow for each second time frame in association with a pair (OD pair) of a first location and a second location.

[0247] This allows the OD estimation device 201 to output, as an OD estimation result, an OD flow in a finer time frame (second time frame) for an OD pair of a departure point (O) and a destination point (D).

[0248] Furthermore, according to the OD estimation device 201, the second OD flow can be calculated by allocating the determined OD flow for each first time frame to each second time frame in accordance with a ratio set based on the time width between the first time frame and the second time frame.

[0249] This allows the OD estimation device 201 to distribute and use the OD flow for each first time frame to each second time frame according to the ratio of the time width between the first time frame and the second time frame. For example, when the ratio of the time width between the first time frame and the second time frame is "2:1", the OD flow for the first time frame is divided into two and distributed to each second time frame. This allows the OD estimation device 201 to improve the accuracy of calculation of the OD flow used as the initial value of the second traffic simulation.

[0250] Moreover, the OD estimation device 201 can calculate a second OD flow by allocating the determined OD flow for each first time frame to each second time frame according to a ratio set based on the traffic volume for each second time frame observed on a first link included in a route from a first point to a second point. The first link is, for example, a link connected to the first point.

[0251] This allows the OD estimation device 201 to distribute the OD flows for each first time frame to each second time frame according to the ratio of link flows (traffic volume) actually observed for each second time frame, thereby enabling the OD estimation device 201 to improve the accuracy of calculation of the OD flows used as the initial values ​​of the second traffic simulation.

[0252] Moreover, the OD estimation device 201 can calculate the error between the traffic volume of a link estimated by the first traffic simulation and the traffic volume observed for that link, determine whether the calculated error is within an acceptable range, and if the error is not within the acceptable range, update the first OD flow based on the error and execute the first traffic simulation using the updated first OD flow. And, if the calculated error is within an acceptable range, the OD estimation device 201 can determine the first OD flow used in the immediately preceding first traffic simulation as the OD flow for each first time frame.

[0253] This allows the OD estimation device 201 to repeat the first traffic simulation while updating the first OD flow until the error between the link flow estimated by the first traffic simulation and the actually observed link flow falls within an acceptable range.

[0254] Moreover, the OD estimation device 201 can calculate the error between the traffic volume of a link estimated by the second traffic simulation and the traffic volume observed for that link, determine whether the calculated error is within an acceptable range, and if the error is not within the acceptable range, update the second OD flow based on the error and execute the second traffic simulation using the updated second OD flow. And, if the calculated error is within an acceptable range, the OD estimation device 201 can determine the second OD flow used in the immediately preceding second traffic simulation as the OD flow for each second time frame.

[0255] This allows the OD estimation device 201 to repeat the second traffic simulation while updating the second OD flow until the error between the link flow estimated by the second traffic simulation and the actually observed link flow falls within an acceptable range.

[0256] From these facts, the OD estimation device 201 can estimate OD flows in fine time frames by using OD flows estimated in coarse time frames, which can significantly reduce the processing time required to estimate OD flows in a desired time frame.

[0257] For example, the OD estimation device 201 can reduce the size of the matrix to be handled in a layer with a large time frame width compared to a layer with a small time frame width, and therefore can shorten the overall processing time. Also, in a layer with a small time frame width, the OD estimation device 201 can reduce the number of iterations required until optimization, since the OD flow is already close to a value close to the true value.

[0258] Furthermore, this OD determination method makes it possible to estimate OD flows (traffic volume) over a detailed time frame in a short time with high accuracy. This makes it possible, for example, to provide a service that estimates OD flows on the cloud. In addition, transportation bureaus and companies that provide dispatch services such as taxis can use the OD estimation results to improve transportation policies and services.

[0259] The OD determination method described in this embodiment can be realized by executing a prepared program on a computer such as a personal computer or a workstation. The OD determination program is recorded on a computer-readable recording medium such as a hard disk, a flexible disk, a CD-ROM, a DVD, or a USB memory, and is executed by being read from the recording medium by the computer. The OD determination program may also be distributed via a network such as the Internet.

[0260] The OD estimation device 201 described in this embodiment can also be realized by an application-specific IC such as a standard cell or a structured ASIC (Application Specific Integrated Circuit) or a PLD (Programmable Logic Device) such as an FPGA.

[0261] The following supplementary notes are further disclosed regarding the above-described embodiment.

[0262] (Supplementary Note 1) A first traffic simulation is executed for each first time frame obtained by dividing a target period, in which a traffic volume for each link included in a route from a first point to a second point is estimated using a first OD traffic volume departing from the first point and arriving at a second point during the first time frame; determining an OD traffic volume for each of the first time frames for a pair of the first location and the second location by repeatedly executing the first traffic simulation while adjusting the first OD traffic volume so that an error between the estimated traffic volume of the link and an observed traffic volume of the link becomes small; calculating, for each second time frame obtained by dividing the first time frame based on the determined OD traffic volume for each of the first time frames, a second OD traffic volume departing from the first point and arriving at the second point during the second time frame; performing a second traffic simulation for estimating a traffic volume for each of the links for each of the second time frames using the calculated second OD traffic volume; determining an OD traffic volume for each of the second time frames for a pair of the first location and the second location by repeatedly executing the second traffic simulation while adjusting the second OD traffic volume so that an error between the estimated traffic volume of the link and an observed traffic volume of the link becomes small; A method for determining OD, characterized in that the processing is executed by a computer.

[0263] (Appendix 2) The calculation process is The OD determination method according to claim 1, further comprising the step of: allocating the determined OD traffic volume for each of the first time frames to each of the second time frames according to a predetermined ratio, thereby calculating the second OD traffic volume.

[0264] (Additional Note 3) The determined OD traffic volume for each second time frame is output in association with the pair of the first location and the second location. 2. The OD determination method according to claim 1, characterized in that the processing is executed by the computer.

[0265] (Supplementary Note 4) The OD determination method according to Supplementary Note 2, characterized in that the ratio is set based on the time width between the first time frame and the second time frame.

[0266] (Supplementary Note 5) The OD determination method described in Supplementary Note 2, characterized in that the ratio is set based on the traffic volume observed for the first link included in the route for each of the second time frames.

[0267] (Supplementary Note 6) The OD determination method according to Supplementary Note 5, characterized in that the first link is a link connected to the first point.

[0268] (Appendix 7) The process of determining the OD traffic volume for each first time frame is calculating an error between the traffic volume of the link estimated by the first traffic simulation and the traffic volume observed for the link, and determining whether or not the calculated error is within an allowable range; If the error is not within the allowable range, updating the first OD traffic volume based on the error, and executing the first traffic simulation using the updated first OD traffic volume; If the calculated error is within an allowable range, the first OD traffic volume used in the immediately preceding first traffic simulation is determined as the OD traffic volume for each of the first time frames. 2. The method for determining OD according to claim 1.

[0269] (Appendix 8) The process of determining the OD traffic volume for each second time frame is calculating an error between the traffic volume of the link estimated by the second traffic simulation and the traffic volume observed for the link, and determining whether or not the calculated error is within an allowable range; If the error is not within the allowable range, updating the second OD traffic volume based on the error, and executing the second traffic simulation using the updated second OD traffic volume; If the calculated error is within an allowable range, the second OD traffic volume used in the immediately preceding second traffic simulation is determined as the OD traffic volume for each of the second time frames. 2. The method for determining OD according to claim 1.

[0270] (Appendix 9) For each first time frame into which a target period is divided, a first traffic simulation is performed to estimate a traffic volume for each link included in a route from a first point to a second point by using a first OD traffic volume departing from the first point and arriving at a second point during the first time frame; determining an OD traffic volume for each of the first time frames for a pair of the first location and the second location by repeatedly executing the first traffic simulation while adjusting the first OD traffic volume so that an error between the estimated traffic volume of the link and an observed traffic volume of the link becomes small; calculating, for each second time frame obtained by dividing the first time frame based on the determined OD traffic volume for each of the first time frames, a second OD traffic volume departing from the first point and arriving at the second point during the second time frame; performing a second traffic simulation for estimating a traffic volume for each of the links for each of the second time frames using the calculated second OD traffic volume; determining an OD traffic volume for each of the second time frames for a pair of the first location and the second location by repeatedly executing the second traffic simulation while adjusting the second OD traffic volume so that an error between the estimated traffic volume of the link and an observed traffic volume of the link becomes small; An OD determination program that causes a computer to execute a process. [Explanation of symbols]

[0271] 101 Information processing device 110 First Simulation 120 Second Simulation 200 Information Processing Systems 201 OD estimation device 202 Client device 210 Network 300 Bus 301 CPU 302 Memory 303 Disk Drive 304 Disk 305 Communication I / F 306 Portable Recording Media I / F 307 Portable recording media 500 Road Network Table 600 Departure Point List 700 Destination list 800 OD Pair Table 900 Routing Table 1000 Control section 1001 Acquisition Department 1002 Setting Department 1003 First Executive Division 1004 First decision unit 1005 Calculation section 1006 Second Executive Division 1007 Second Decision Section 1008 Output section 1201 Initial OD flow 1202, 2000, 2201, 2202, 3000 OD Flow 1210,2210 Forward processing 1220,2220 Backward processing 1400,2400 Route Selection Table 1500,1600,1800,1901,2500,2600,2800,2901 Link Flow 1700,2700 Observation matrix 1900,2900 error 2001,3001 Correction value

Claims

1. executing a first traffic simulation for each first time frame obtained by dividing a target period, using a first OD traffic volume departing from a first point and arriving at a second point during the first time frame, to estimate a traffic volume for each link included in a route from the first point to the second point; determining an OD traffic volume for each of the first time frames for a pair of the first location and the second location by repeatedly executing the first traffic simulation while adjusting the first OD traffic volume so that an error between the estimated traffic volume of the link and an observed traffic volume of the link becomes small; calculating, for each second time frame obtained by dividing the first time frame based on the determined OD traffic volume for each of the first time frames, a second OD traffic volume departing from the first point and arriving at the second point during the second time frame; performing a second traffic simulation for estimating a traffic volume for each of the links for each of the second time frames using the calculated second OD traffic volume; determining an OD traffic volume for each of the second time frames for a pair of the first location and the second location by repeatedly executing the second traffic simulation while adjusting the second OD traffic volume so that an error between the estimated traffic volume of the link and an observed traffic volume of the link becomes small; A method for determining OD, the processing of which is carried out by a computer.

2. The calculation process includes: The OD determination method according to claim 1, characterized in that the second OD traffic volume is calculated by allocating the determined OD traffic volume for each of the first time frames to each of the second time frames according to a predetermined ratio.

3. outputting the determined OD traffic volume for each of the second time frames in association with the pair of the first location and the second location; 2. The method for determining OD according to claim 1, wherein the processing is executed by the computer.

4. 3. The OD determination method according to claim 2, wherein the ratio is set based on a time width between the first time frame and the second time frame.

5. executing a first traffic simulation for each first time frame obtained by dividing a target period, using a first OD traffic volume departing from a first point and arriving at a second point during the first time frame, to estimate a traffic volume for each link included in a route from the first point to the second point; determining an OD traffic volume for each of the first time frames for a pair of the first location and the second location by repeatedly executing the first traffic simulation while adjusting the first OD traffic volume so that an error between the estimated traffic volume of the link and an observed traffic volume of the link becomes small; calculating, for each second time frame obtained by dividing the first time frame based on the determined OD traffic volume for each of the first time frames, a second OD traffic volume departing from the first point and arriving at the second point during the second time frame; performing a second traffic simulation for estimating a traffic volume for each of the links for each of the second time frames using the calculated second OD traffic volume; determining an OD traffic volume for each of the second time frames for a pair of the first location and the second location by repeatedly executing the second traffic simulation while adjusting the second OD traffic volume so that an error between the estimated traffic volume of the link and an observed traffic volume of the link becomes small; An OD determination program that causes a computer to execute a process.

Citation Information

Patent Citations

  • Traffic state simulation device, and program

    JP2009259158A