Traffic simulation program, traffic simulation method, and information processor

Abstract

To upgrade precision in traffic simulation.SOLUTION: A computer searches for plural moving routes with one or more different pieces of transportation means along which a person moves from a departure point to a destination using one or more pieces of transportation means out of plural pieces of transportation means, uses a behavior model for predicting a selection behavior on the basis of feature quantities representing conditions of plural pieces of transportation means to predict a first moving route selected by a person from among the plural moving routes, uses the first moving route to simulate a first movement along a time base, and uses a result of the first movement to update the feature quantities.SELECTED DRAWING: Figure 1

Description

[Technical Field]

[0001] The present invention relates to a traffic simulation program, a traffic simulation method, and an information processing device. [Background technology]

[0002] Computers may run traffic simulations to simulate the use of multiple modes of transportation in a given area. Some people choose multimodal routes, which involve transferring between two or more modes of transportation from one origin to another. The results of traffic simulations may be used in urban planning, such as public transportation scheduling, road improvements, parking facilities, and transportation pricing changes.

[0003] A route search method has been proposed that presents multimodal travel routes to users by taking into account real-time information such as current traffic volume and parking lot conditions. A modeling method has also been proposed that generates a predictive model that predicts demand at specified stops by learning the relationship between public transport demand and topography.

[0004] In addition, a traffic control method has been proposed that predicts future traffic volume from the current traffic volume of a transportation facility, simulates the future traffic volume if certain corrective measures are taken, and searches for corrective measures that will improve future traffic volume.In addition, a server device has been proposed that searches for a normal route from a departure point to a destination by private car and a special route in which the private car is parked along the way and transferred to public transportation, and displays the normal route and the special route on a user terminal. [Prior art documents] [Patent documents]

[0005] [Patent Document 1] US Patent Application Publication No. 2016 / 0334235 [Patent Document 2] US Patent Application Publication No. 2017 / 0109764 [Patent Document 3] International Publication No. 2020 / 065148 [Patent Document 4] Japanese Patent Application Publication No. 2023-121091 Summary of the Invention [Problem to be solved by the invention]

[0006] The route selection by people may be affected by the dynamic state of transportation facilities, such as congestion levels. Therefore, there is room for improving the accuracy of traffic simulations by appropriately handling the dynamic state of transportation facilities. Therefore, in one aspect, the present invention aims to improve the accuracy of traffic simulations. [Means for solving the problem]

[0007] In one aspect, a traffic simulation program is provided that causes a computer to execute the following processes: search for multiple travel routes in which a person travels from a departure point to a destination using one or more of multiple forms of transportation, where the multiple travel routes differ by one or more forms of transportation; predict a first travel route that the person will select from the multiple travel routes using a behavioral model that predicts selection behavior based on features that indicate the status of the multiple forms of transportation; simulate a first travel of the person along a time axis using the first travel route; and update the features using the results of the first travel. [Effects of the Invention]

[0008] On the one hand, it improves the accuracy of traffic simulations. [Brief explanation of the drawings]

[0009] [Figure 1] FIG. 1 is a diagram illustrating an information processing apparatus according to a first embodiment. [Figure 2] FIG. 10 illustrates an example of hardware of an information processing apparatus according to a second embodiment. [Figure 3] FIG. 1 is a diagram illustrating an example of a travel route from a departure point to a destination. [Figure 4] FIG. 10 is a diagram illustrating an example of a first simulation cycle. [Figure 5] FIG. 10 is a diagram illustrating an example of a second simulation cycle. [Figure 6] FIG. 10 is a diagram illustrating an example of OD data. [Figure 7] FIG. 10 is a diagram illustrating an example of status data related to a parking lot. [Figure 8] FIG. 10 is a diagram showing an example of status data relating to stations and trains. [Figure 9] FIG. 10 is a diagram illustrating an example of calculation of a selection probability of a travel route. [Figure 10] FIG. 2 is a block diagram illustrating an example of functions of the information processing device. [Figure 11] FIG. 10 is a diagram illustrating an example of components of state data. [Figure 12] 10 is a flowchart showing an example of a procedure for a traffic simulation. [Figure 13] 10 is a flowchart illustrating an example of a procedure for OD record processing. [Figure 14] 10 is a flowchart showing an example of a procedure for warehousing processing. DETAILED DESCRIPTION OF THE INVENTION

[0010] Hereinafter, this embodiment will be described with reference to the drawings. (a) First embodiment FIG. 1 is a diagram illustrating an information processing device according to a first embodiment. The information processing device 10 of the first embodiment executes a traffic simulation that simulates the usage of multiple modes of transportation in a certain area. The information processing device 10 may be a client device or a server device. The information processing device 10 may also be called a computer or a traffic simulation device. The traffic simulation of the first embodiment described below is a technology for improving computer functions.

[0011] The information processing device 10 includes a storage unit 11 and a processing unit 12. The storage unit 11 may be a volatile semiconductor memory such as a random access memory (RAM), or may be a non-volatile storage such as a hard disk drive (HDD) or a flash memory.

[0012] The processing unit 12 is, for example, a processor such as a central processing unit (CPU), a graphics processing unit (GPU), or a digital signal processor (DSP). However, the processing unit 12 may also include an electronic circuit such as an application specific integrated circuit (ASIC) or a field programmable gate array (FPGA). The processor executes a program stored in a memory such as a RAM. A collection of processors may be called a multiprocessor or simply a "processor."

[0013] The storage unit 11 stores feature quantities indicating the status of a plurality of transportation modes. A transportation mode is a method of travel for people to move around, and may be called a transportation means or a means of transportation. The transportation modes may include walking, bicycles, private cars, buses, trains, ships, and airplanes. The feature quantities indicate the dynamic status of the transportation modes, and may include congestion levels or delay times.

[0014] For example, a feature related to walking may include the congestion level of a pedestrian road. A feature related to bicycles may include the congestion level of a bicycle road or a bicycle parking lot. A feature related to private cars may include the congestion level of a roadway or a parking lot. A feature related to buses may include the congestion level of buses or bus stops. A feature related to trains may include the congestion level of trains or stations. A feature related to ships may include the congestion level of ships or ports. A feature related to airplanes may include the congestion level of airplanes or airports.

[0015] At the start of the traffic simulation, the feature quantities of the multiple transportation modes may be initialized to initial values. For example, the storage unit 11 stores a feature quantity 14a of the transportation mode 13a, a feature quantity 14b of the transportation mode 13b, and a feature quantity 14c of the transportation mode 13c.

[0016] The memory unit 11 also stores a behavior model 16. The behavior model 16 is a model for predicting a person's selection behavior when selecting a travel route. The behavior model 16 may be a machine learning model trained through machine learning, or may be a linear or nonlinear function including predetermined coefficients. The behavior model 16 may calculate the selection probability of each of a plurality of travel routes. For example, the behavior model 16 calculates the selection probability that a travel route will be selected based on the time, cost, and feature quantities of the travel route.

[0017] The processing unit 12 executes a traffic simulation. First, the processing unit 12 searches for multiple travel routes for a person to travel from a departure point to a destination. The departure point and the destination may be given in advance as a traffic demand. The traffic demand may be generated randomly during the traffic simulation. The processing unit 12 may search for multiple travel routes for each of multiple people whose departure points and / or destinations are different.

[0018] The travel route includes one or more transportation modes among a plurality of transportation modes. The plurality of travel routes use different transportation modes. The processing unit 12 may use timetable data of public transportation modes to search for travel routes using buses or trains. The processing unit 12 may also use road map data to search for travel routes using bicycles or private cars.

[0019] A travel route may include transfers between different modes of transportation and may be called a multimodal travel route. For example, a certain travel route may depart from a departure point in a private car, park the car in a parking lot, and transfer to a bus or train. When searching for possible travel routes, the processing unit 12 may determine whether or not transfers are possible using characteristics of the transportation modes, such as congestion levels. As an example, travel route 15a uses transportation modes 13a and 13b. Travel route 15b uses transportation mode 13c.

[0020] The processing unit 12 predicts the travel route that the person will select from among the plurality of travel routes, using the feature amounts of the plurality of transportation modes and the behavior model 16. The processing unit 12 may randomly select one of the travel routes according to the selection probability calculated by the behavior model 16.

[0021] As an example, the processing unit 12 calculates the time, cost, and feature amount of the travel route 15a. The time of the travel route 15a may be the sum of the required time of the transportation facility 13a and the required time of the transportation facility 13b. The cost of the travel route 15a may be the sum of the fare of the transportation facility 13a and the fare of the transportation facility 13b. The feature amount of the travel route 15a may be calculated from the feature amounts 14a and 14b, and may indicate the average congestion degree of the transportation facilities 13a and 13b.

[0022] The processing unit 12 may calculate the time and cost of the travel route 15a using timetable data and road map data. The processing unit 12 may also calculate the time using the feature amounts 14a and 14b. The degree of road congestion and the degree of parking lot congestion may affect the required time. The processing unit 12 may also calculate the cost using the feature amounts 14a and 14b. The degree of parking lot congestion may affect the parking fee.

[0023] The processing unit 12 inputs the time, cost, and feature values ​​of the travel route 15a into the behavior model 16 to calculate the selection probability of the travel route 15a. The processing unit 12 also inputs the time, cost, and feature values ​​of the travel route 15b into the behavior model 16 to calculate the selection probability of the travel route 15b. The selection probability of a travel route that includes a highly congested transportation facility may be low. The behavior model 16 may further use the person's profile, such as age and gender, to calculate the selection probability. The person's profile may be randomly generated together with traffic demand during traffic simulation. The processing unit 12 may select the travel route 15a according to the selection probabilities of the travel routes 15a and 15b.

[0024] The processing unit 12 simulates the movement of the person along a time axis using the predicted movement route. The processing unit 12 may calculate the position of the person after a certain time period using timetable data and road map data. The processing unit 12 may also calculate the position of the person after a certain time period using feature quantities of transportation means included in the predicted movement route. The processing unit 12 identifies the transportation means used by the person within the certain time period.

[0025] The processing unit 12 uses the results of the above movement to update the feature values ​​of the transportation means included in the predicted movement route. If the use of the transportation means starts within a certain time, the processing unit 12 updates the feature values ​​so that the congestion level of the transportation means increases. If the use of the transportation means ends within the certain time, the processing unit 12 updates the feature values ​​so that the congestion level of the transportation means decreases. The processing unit 12 may calculate the congestion level of each transportation means from the results of the movement of multiple people.

[0026] In the traffic simulation, the processing unit 12 may repeat a cycle including searching for a travel route, selecting a travel route, and executing travel a plurality of times. At this time, the processing unit 12 may carry over and use the feature values ​​of transportation means updated in one cycle in the next cycle.

[0027] Different cycles may correspond to different time periods. In this case, the state of the transportation facility at the end of one time period is carried over to the start of the next time period. Thus, a person's choice of travel route in one time period will affect the travel route selection of other people in the next time period. Different cycles may also correspond to multiple trials for the same person and travel in the same time period. In this case, the state of the transportation facility predicted in one trial is carried over to the next trial. Thus, the state of the transportation facility may converge over multiple cycles.

[0028] The processing unit 12 outputs the results of the traffic simulation. The results of the traffic simulation may include the movement results of one or more people along a time axis, or may include changes in the characteristics of each transportation mode over time. The results of the traffic simulation may also include an index value indicating the efficiency of people's movement, such as carbon dioxide emissions. The processing unit 12 may store the results of the traffic simulation in non-volatile storage, display them on a display device, or transmit them to another information processing device.

[0029] As described above, the information processing device 10 of the first embodiment searches for multiple travel routes in which a person travels from a departure point to a destination using one or more different modes of transportation. The information processing device 10 predicts a first travel route to be selected by the person from the multiple travel routes using a behavioral model 16 that predicts selection behavior based on features indicating the status of the multiple modes of transportation. The information processing device 10 simulates a first movement of the person along a time axis using the first travel route. The information processing device 10 updates the features using the results of the first movement.

[0030] This allows the information processing device 10 to perform a multimodal traffic simulation that allows people to move from a departure point to a destination while changing between different modes of transportation. Therefore, the information processing device 10 can simulate the usage of multiple modes of transportation in a certain area, providing information useful for urban planning. Furthermore, the information processing device 10 can reflect the interdependence between people's travel route selection and the dynamic state of the transportation modes in the traffic simulation, thereby improving the accuracy of the traffic simulation.

[0031] (b) Second embodiment FIG. 2 is a diagram illustrating an example of hardware of an information processing device according to a second embodiment. The information processing device 100 according to the second embodiment executes a multimodal traffic simulation. The information processing device 100 may be a client device or a server device. The information processing device 100 corresponds to the information processing device 10 according to the first embodiment. The traffic simulation according to the second embodiment described below is a technology for improving computer functions.

[0032] The information processing device 100 has a CPU 101, a RAM 102, a HDD 103, a GPU 104, an input interface 105, a medium reader 106, and a communication interface 107, all connected via a bus. The CPU 101 corresponds to the processing unit 12 in the first embodiment. The RAM 102 or the HDD 103 corresponds to the storage unit 11 in the first embodiment.

[0033] The CPU 101 is a processor that executes program instructions. The CPU 101 loads programs and data stored in the HDD 103 into the RAM 102 and executes the programs. The information processing device 100 may have multiple processors.

[0034] The RAM 102 is a volatile semiconductor memory that temporarily stores programs and data. The programs are executed by the CPU 101, and the data is used for calculations by the CPU 101. The information processing device 100 may have a type of volatile memory other than RAM.

[0035] The HDD 103 is a non-volatile storage that stores software programs and data. The software includes an operating system (OS), middleware, and application software. The information processing device 100 may also have other types of non-volatile storage, such as a solid state drive (SSD).

[0036] The GPU 104 performs image processing in cooperation with the CPU 101 and displays the image on a display device 111 connected to the information processing device 100. The display device 111 is, for example, a CRT (Cathode Ray Tube) display, a liquid crystal display, an organic EL (Electro Luminescence) display, or a projector.

[0037] The GPU 104 may be used as a general purpose computing on graphics processing unit (GPGPU). The GPU 104 may execute a program in response to an instruction from the CPU 101. The information processing device 100 may include a volatile semiconductor memory other than the RAM 102 as a GPU memory.

[0038] The input interface 105 receives an input signal from an input device 112 connected to the information processing device 100. The input device 112 is, for example, a mouse, a touch panel, or a keyboard. A plurality of input devices may be connected to the information processing device 100.

[0039] The medium reader 106 is a reading device that reads programs and data from the recording medium 113. The recording medium 113 is, for example, a magnetic disk, an optical disk, or a semiconductor memory. Magnetic disks include flexible disks (FDs) and HDDs. Optical disks include compact discs (CDs) and digital versatile discs (DVDs). The medium reader 106 copies the programs and data read from the recording medium 113 to the RAM 102 or the HDD 103.

[0040] The read program may be executed by the CPU 101. The recording medium 113 may be a portable recording medium. The recording medium 113 may be used to distribute programs and data. The recording medium 113 and the HDD 103 may be referred to as computer-readable recording media.

[0041] The communication interface 107 communicates with other information processing devices via the network 114. The communication interface 107 may be a wired communication interface that connects to a router or a switch via a wired cable, or may be a wireless communication interface that connects to a base station or an access point via a wireless link.

[0042] Next, we will explain the multimodal traffic simulation. The information processing device 100 executes a human behavior simulation in which multiple agents representing multiple people move according to certain rules in a virtual space that accurately reproduces the real space of a target area. This virtual space is sometimes called a digital twin. The virtual space includes virtual transportation facilities reproduced from real transportation data, such as timetable data and road map data.

[0043] The information processing device 100 randomly generates multiple agents representing multiple people by referencing data on residents of the target area. Each agent has a transportation demand and a profile. The information processing device 100 moves the multiple agents in the virtual space according to the transportation demand and profile. The transportation modes available to the agents include walking, private cars, bicycles, buses, trains, ships, and airplanes. The information processing device 100 outputs the movement results of each agent, the dynamic state of each transportation mode, and other evaluation index values. The other evaluation index values ​​are indicators that indicate the efficiency of the transportation design in the target area, such as the total carbon dioxide emissions of multiple transportation modes.

[0044] The information processing device 100 may also perform a human behavior simulation on a digital twin in which the real world and a virtual space are time-synchronized. More specifically, the information processing device 100 generates a digital twin in which the states of multiple transportation systems in the real world are reproduced in a virtual space, where the real world and the virtual space are time-synchronized. The information processing device 100 then simulates the behavior of multiple people in the generated digital twin by moving multiple agents corresponding to multiple people in the real world. For example, the digital twin reproduces the states of multiple real transportation systems in a virtual space based on sensing data from sensors placed in the real world and current transportation operation data. The information processing device 100 then performs a human behavior simulation in which multiple agents representing multiple people are placed on the digital twin in which the states of multiple transportation systems are reproduced, and the placed agents move according to certain rules.

[0045] Figure 3 is a diagram showing an example of a travel route from a departure point to a destination. In a multimodal traffic simulation, an agent has a travel demand to travel from a departure point 31 to a destination 32. There are multiple travel routes between the departure point 31 and the destination 32.

[0046] According to the first movement route, the agent drives a car from the departure point 31 to a parking lot 33 and leaves the car in the parking lot 33. The agent transfers to a bus 36 and travels by bus 36 to a bus stop 35. The agent walks from the bus stop 35 to the destination 32. According to the second movement route, the agent drives a car from the departure point 31 to the destination 32. According to the third movement route, the agent drives a car from the departure point 31 to a parking lot 34 and leaves the car in the parking lot 34. The agent transfers to a train 37 and travels by train 37 to the destination 32.

[0047] In this way, the information processing device 100 searches for multiple travel routes for each of multiple people to travel from a departure point to a destination. The information processing device 100 calculates the utility of each travel route for that person using a utility function. The utility is affected by the time and cost of the travel route, the dynamic state of the transportation used, and the person's profile. The dynamic state includes the degree of congestion. The information processing device 100 calculates the selection probability of each travel route from the utility and randomly selects one of the multiple travel routes according to the selection probability.

[0048] When searching for a route, the information processing device 100 may use public transportation data such as GTFS (General Transit Feed Specification). The public transportation data includes station locations, train schedules, fares between stations, bus stop locations, bus schedules, and fares between bus stops. The information processing device 100 may also use road map data such as OSM (Open Street Map). The road map data indicates roads, parking lot locations, and bicycle parking lot locations. The information processing device 100 may also use a route search program such as OTP (Open Trip Planner).

[0049] The multimodal traffic simulation includes multiple cycles. Different cycles may correspond to different time periods. Different cycles may also correspond to different trials for the same simulation period. In each cycle, the information processing device 100 feeds back the movement results of multiple people and updates status data indicating the dynamic status of transportation facilities. Transportation facilities that are used by many people tend to become more crowded. The information processing device 100 carries over and uses the status data updated in one cycle to the next cycle.

[0050] 4 shows an example of the first simulation cycle. The multimodal traffic simulation includes origin-destination (OD) data generation 141, route search 142, decision making 143, movement calculation 144, state update 145, and indicator output 146.

[0051] OD data generation 141 generates OD data from data on residents of a target area. The OD data includes multiple OD records corresponding to multiple people. The resident data may be the results of a survey conducted in the target area. The OD record includes a departure point, a destination, and a departure time. The OD record also includes a profile such as age and gender.

[0052] For example, the OD data generation 141 randomly selects a departure point within the target area according to the distribution of residents' addresses in the target area. The OD data generation 141 also randomly selects a destination within the target area according to the distribution of residents' workplaces. The OD data generation 141 also randomly selects a departure time according to the distribution of residents' commute times. The OD data generation 141 also randomly selects the age and gender of each OD record according to the distribution of residents' ages and genders.

[0053] For each OD record, the route search 142 searches for multiple travel routes that depart from a departure point at a departure time and reach a destination. The route search 142 uses public transportation data, road map data, and status data. For example, the route search 142 calculates the required walking time from the length of the road indicated by the road map data. The route search 142 also calculates the required bicycle travel time from the length of the road and the degree of congestion at the bicycle parking lot indicated by the status data.

[0054] The route search 142 also calculates the travel time by private car from the length of the road, the degree of road congestion indicated by the status data, and the degree of parking lot congestion indicated by the status data. The route search 142 also calculates the travel time by bus from the bus schedule indicated by the public transportation data and the degree of road congestion. The route search 142 also calculates the travel time by train from the train schedule indicated by the public transportation data.

[0055] The decision making unit 143 calculates the utility of each of multiple travel routes for each OD record using a utility function, and calculates the selection probability of each travel route based on the relative relationship between the utilities of the multiple travel routes. The decision making unit 143 selects one of the travel routes for each OD record according to the selection probability. The utility function calculates the utility from the time, cost, and congestion level of the travel route. The decision making unit 143 may change the weights of the time, cost, and congestion level according to the profile, such as age and gender, included in the OD record.

[0056] The decision making 143 identifies the travel route time from the result of the route search 142. The decision making 143 also calculates the total cost of the transportation modes included in the travel route as the cost of that travel route. For example, the cost of walking and cycling is zero. The cost of a private car is the sum of fuel costs based on the length of the road and parking fees. Parking fees may vary depending on the congestion level of the parking lot. The bus cost is the bus fare based on the travel distance. The train cost is the train fare based on the travel distance.

[0057] Furthermore, the decision making 143 determines the congestion level of the travel route from the congestion levels of one or more transportation modes included in the travel route. This congestion level is inversely proportional to the comfort level. The congestion level of the travel route is, for example, the maximum or average value of the congestion levels of one or more transportation modes. For example, the congestion level for bicycles includes the congestion level of bicycle parking lots. The congestion level for private cars includes the congestion level of parking lots. The congestion level for buses includes the congestion level of the bus vehicle itself and the congestion level of bus stops. The congestion level for trains includes the congestion level of the train vehicle itself and the congestion level of stations.

[0058] The movement calculation 144 moves people in virtual space along the selected movement route for each OD record. This determines the position of the person at each time. The status update 145 updates the status data using the movement results of multiple people corresponding to multiple OD records. For example, the status update 145 calculates the congestion level of a bicycle parking lot from the position of a bicycle after a certain time.

[0059] Furthermore, the status update 145 calculates the degree of congestion of roads and parking lots from the position of private cars after a certain time. The status update 145 also calculates the degree of congestion of buses and bus stops from the position of people after a certain time. The status update 145 also calculates the degree of congestion of trains and stations from the position of people after a certain time.

[0060] The index output 146 calculates evaluation index values ​​such as carbon dioxide emissions from the usage of transportation modes over the entire target period. For example, the carbon dioxide emissions from walking and cycling are zero. The index output 146 calculates carbon dioxide emissions from private cars from the total distance traveled by multiple people in private cars. The index output 146 also calculates carbon dioxide emissions from buses from the total distance traveled by buses over the target period indicated by the public transportation data. The index output 146 also calculates carbon dioxide emissions from trains from the total distance traveled by trains over the target period indicated by the public transportation data.

[0061] In the multimodal traffic simulation, the information processing device 100 first executes OD data generation 141. Next, the information processing device 100 repeats a cycle n times, each cycle including route search 142, decision making 143, movement calculation 144, and state update 145. Here, the information processing device 100 divides the entire target period into n periods. The time span of one cycle is, for example, 1 second, 10 seconds, or 1 minute.

[0062] As an example, the information processing device 100 executes cycles 147-1 to 147-n in order. Cycle 147-1 simulates people's movements from 7:00 to 7:01. Prior to cycle 147-1, the information processing device 100 initializes state data indicating the dynamic state of transportation facilities. The information processing device 100 executes route search 142 and decision making 143 in cycle 147-1 using the initialized state data. State update 145 in cycle 147-1 calculates the dynamic state of transportation facilities as of 7:01.

[0063] Cycle 147-2 simulates the movement of people from 7:01 to 7:02. The information processing device 100 executes route search 142 and decision making 143 in cycle 147-2 using state data updated by state update 145 in cycle 147-1. State update 145 in cycle 147-2 calculates the dynamic state of transportation facilities as of 7:02.

[0064] Cycle 147-n simulates people's movements from 7:59 to 8:00. The information processing device 100 performs route search 142 and decision making 143 in cycle 147-n using state data updated by state update 145 in the previous cycle. State update 145 in cycle 147-n calculates the dynamic state of transportation facilities as of 8:00. Finally, the information processing device 100 performs index output 146.

[0065] FIG. 5 is a diagram showing an example of a second simulation cycle. Here, the information processing device 100 performs n trials of travel simulation using the same OD data and the same target period. Because the decision-making 143 uses random numbers, the route selection results will differ between the n trials. Furthermore, the information processing device 100 carries over the state data updated in one trial to the next trial. In this way, the information processing device 100 averages out the influence of the random numbers and converges the dynamic state of the transportation facility to an appropriate state.

[0066] As an example, the information processing device 100 executes cycles 148-1 to 148-n in order. Cycle 148-1 simulates people's movements from 7:00 to 8:00. Prior to cycle 148-1, the information processing device 100 initializes state data indicating the dynamic state of transportation facilities. The information processing device 100 executes route search 142 and decision making 143 in cycle 148-1 using the initialized state data. State update 145 in cycle 148-1 updates the dynamic state of transportation facilities from 7:00 to 8:00.

[0067] Cycle 148-2 again simulates people's movements from 7:00 to 8:00. The information processing device 100 executes route search 142 and decision making 143 in cycle 148-2 using the state data updated by state update 145 in cycle 148-1. State update 145 in cycle 148-2 updates the dynamic state of transportation facilities from 7:00 to 8:00.

[0068] Cycle 148-n again simulates people's movements from 7:00 to 8:00. The information processing device 100 executes route search 142 and decision making 143 in cycle 148-n using state data updated in state update 145 of the previous cycle. State update 145 in cycle 148-n updates the dynamic state of transportation facilities from 7:00 to 8:00.

[0069] The index output 146 may calculate an evaluation index value such as the amount of carbon dioxide emissions using the travel results of the last cycle 148-n. The index output 146 may also calculate an average evaluation index value from the travel results of cycles 148-1 to 148-n. The index output 146 may also output the dynamic state of the transportation facility in the last cycle 148-n, or may output the average dynamic state of cycles 148-1 to 148-n. The information processing device 100 may also execute a combination of FIGS. 4 and 5.

[0070] 6 is a diagram showing an example of OD data. Table 151 stores OD data including multiple OD records. The OD record includes an OID, departure time, departure point, destination, purpose of travel, gender, age, whether or not the user has a driver's license, and whether or not the user owns a private car.

[0071] The OID identifies the OD record. The origin and destination are expressed using latitude and longitude. The departure time, origin, and destination indicate transportation demand. The purpose of travel may be commuting to work or returning home. The purpose of travel, gender, and age are input into a utility function and affect the person's travel route selection behavior. For example, the information processing device 100 calculates weights for time, fare, and congestion level based on a combination of purpose of travel, gender, and age according to predetermined rules. Furthermore, for example, when a combination of purpose of travel, gender, and age meets a specific condition, the information processing device 100 sets to zero the probability of selecting a travel route whose congestion level exceeds a threshold.

[0072] Whether or not a person has a driver's license and whether or not they own a car are used to search for travel routes from a departure point to a destination. For OD records in which a person has a driver's license and owns a car, the information processing device 100 also searches for travel routes that include a car as a mode of transportation. On the other hand, for OD records in which a person does not have a driver's license or does not own a car, the information processing device 100 does not search for travel routes that include a car as a mode of transportation.

[0073] FIG. 7 is a diagram showing an example of status data related to a parking lot. Table 152 stores status data indicating the dynamic status of one parking lot. Table 152 includes a parking lot ID, parking capacity, number of parked cars, location, status, fee, and congestion level. The parking lot ID identifies the parking lot. The parking capacity is the number of parking lots the parking lot has. The number of parked cars is the number of parking lots in use. The location of the parking lot is expressed using latitude and route.

[0074] The status is open or closed. The fee is a parking fee per hour and may vary depending on the congestion level. The congestion level is low (light), medium, or high (heavy). For example, if the ratio of the number of parked vehicles to the parking capacity is less than a first threshold, the information processing device 100 determines the congestion level to be low. If the ratio is equal to or greater than the first threshold and equal to or less than a second threshold, the information processing device 100 determines the congestion level to be medium. If the ratio exceeds the second threshold, the information processing device 100 determines the congestion level to be high.

[0075] Table 153 is associated with table 152. Table 153 stores status data indicating details of the usage status of the parking lot indicated by table 152. Table 153 stores multiple records, each of which includes a lot ID, a status, a vehicle ID, and a parking time. The lot ID identifies the parking lot. The status is parked or available. The vehicle ID identifies the vehicle parked in that parking lot. The parking time is the time of entry.

[0076] FIG. 8 is a diagram showing an example of status data related to stations and trains. Table 154 stores status data indicating the dynamic status of one station. Table 154 includes the station ID, status, last update time, location, whether it is barrier-free, appropriate capacity, maximum capacity, current number of people, congestion level, and next train. The station ID identifies the station.

[0077] The status is open or closed. The last updated time is the time in the simulation when the current number of people, congestion level, and next train were updated. The station location is expressed using latitude and route. Barrier-free status indicates whether the station has barrier-free facilities. The appropriate capacity is the number of people that are expected to stay within the station premises in the design. The maximum capacity is the upper limit of the number of people that can stay within the station premises. The current number of people is the number of people currently staying within the station premises.

[0078] The congestion level is low, medium, or high. For example, the information processing device 100 determines the congestion level based on the ratio of the current number of passengers to the appropriate capacity, or the ratio of the current number of passengers to the maximum capacity. The next train identifies the train scheduled to arrive at the station next. If multiple lines pass through the station, there is a next train for each of the multiple lines.

[0079] Table 155 is associated with table 154. Table 155 stores status data indicating the dynamic status of one train. Table 155 includes train ID, status, last update time, location, whether it is barrier-free, next station, whether it has a women-only car, number of seats and appropriate number of passengers, maximum number of passengers, current number of passengers, and congestion level. The train ID identifies the train.

[0080] The status can be normal, delayed, etc. The last updated time is the time in the simulation when the next station, current number of passengers, and congestion level were updated. The train's location is expressed using latitude and route. Barrier-free status indicates whether the train has barrier-free facilities. The next station identifies the station where the train is scheduled to stop next.

[0081] The appropriate number of passengers is the number of passengers on a train that is assumed in the design. The appropriate number of passengers is the sum of the number of seats and the number of standing passengers that is assumed in the design. The maximum number of passengers is the upper limit of the number of passengers that can be tolerated on a train. The current number of passengers is the number of passengers currently on the train. The congestion level is low, medium, or high. For example, the information processing device 100 determines the congestion level from the ratio of the current number of passengers to the appropriate number of passengers, or the ratio of the current number of passengers to the maximum number of passengers.

[0082] 9 is a diagram showing an example of calculation of the selection probability of a travel route. Table 156 associates options, variables, values, and selection probabilities. The information processing device 100 searches for multiple travel routes as options for each of multiple OD records. In the example of table 156, the information processing device 100 has found three travel routes for a certain OD record.

[0083] The information processing device 100 calculates the utility V of the travel route k using a utility function such as Equation (1). k In formula (1), for simplicity, the utility function is a linear function. k is the cost of travel route k, t k is the time for travel route k, m k is the congestion level of travel route k. β c is the cost weight, β t is the time weight, β m is the congestion weight.

[0084]

number

[0085] Cost C k , time t k and congestion level m k is the utility V k This is a variable for calculating m k A numerical value representing the degree of congestion is substituted for m. The higher the degree of congestion, k The utility function may also include other variables. c ,β t ,β m may be determined in advance by machine learning or may be specified by a user. In addition, the information processing device 100 determines the weight β according to the profile included in the OD record. c ,β t ,β m may be selected.

[0086] The information processing device 100 calculates a plurality of utilities corresponding to a plurality of travel routes using a utility function. Then, the information processing device 100 calculates the selection probability P of the travel route k from the plurality of utilities using a logit function shown in Equation (2). k Calculate the selection probability P k is a number between 0 and 1. The sum of the selection probabilities of all travel routes is 1.

[0087]

number

[0088] In the example of table 156, the information processing device 100 calculates the utility V1 of travel route 1 from the cost c1, time t1, and congestion level m1 of travel route 1. The information processing device 100 also calculates the utility V2 of travel route 2 from the cost c2, time t2, and congestion level m2 of travel route 2. The information processing device 100 also calculates the utility V3 of travel route 3 from the cost c3, time t3, and congestion level m3 of travel route 3. The information processing device 100 calculates the selection probabilities P1, P2, and P3 of travel routes 1, 2, and 3 from the utilities V1, V2, and V3.

[0089] In the example of table 156, the selection probability P1 is 56%, the selection probability P2 is 33%, and the selection probability P3 is 11%. Therefore, using random numbers, the information processing device 100 selects the travel route 1 with a probability of 56%, the travel route 2 with a probability of 33%, and the travel route 3 with a probability of 11%. Next, the functions and processing procedures of the information processing device 100 will be described.

[0090] 10 is a block diagram showing an example of functions of an information processing device. The information processing device 100 has an OD data storage unit 121, a traffic data storage unit 122, and a status data storage unit 123. These storage units are implemented using, for example, RAM 102 or HDD 103. The information processing device 100 also has a simulation control unit 124, a route search unit 125, a decision making unit 126, a movement calculation unit 127, a status update unit 128, and an index output unit 129. These processing units are implemented using, for example, a CPU 101 and a program.

[0091] The OD data storage unit 121 stores OD data. The OD data includes multiple OD records representing multiple people. The transportation data storage unit 122 stores transportation data indicating the static design of transportation facilities. The transportation data includes public transportation data and road map data. The public transportation data indicates the locations of bus stops, bus schedules, fares between bus stops, locations of stations, train schedules, and fares between stations. The road map data indicates roads, locations of bicycle parking lots, and locations of parking lots.

[0092] The status data storage unit 123 stores status data indicating the dynamic status of transportation facilities. The status data includes road usage status, parking lot usage status, bicycle parking lot usage status, station usage status, train usage status, bus stop usage status, and bus usage status.

[0093] The simulation control unit 124 controls the multimodal traffic simulation. The simulation control unit 124 displays a user interface on the display device 111 and accepts simulation conditions from the user. The simulation conditions include a target area and a target period. The simulation control unit 124 generates OD data that is suitable for the target area and stores it in the OD data storage unit 121. The number of OD records included in the OD data, i.e., the number of moving people, may be specified by the user.

[0094] The simulation control unit 124 also displays the simulation results on the display device 111. The simulation results may show the movement of multiple people along a time axis. The simulation results may also show changes over time in the dynamic state of transportation. The simulation results may also include evaluation index values ​​such as carbon dioxide emissions.

[0095] The route search unit 125 searches for multiple travel routes from the departure point to the destination for each of the multiple OD records included in the OD data stored in the OD data storage unit 121. The route search unit 125 uses the traffic data stored in the traffic data storage unit 122 and the status data stored in the status data storage unit 123.

[0096] The decision-making unit 126 uses a utility function created in advance for each of the multiple OD records to select one of the multiple travel routes searched by the route search unit 125. Here, the decision-making unit 126 inputs the cost and time of the travel route, as well as the congestion level indicated by the status data, into the utility function to calculate the utility of the travel route. The decision-making unit 126 calculates the selection probability of each travel route from the utilities of the multiple travel routes. The decision-making unit 126 probabilistically selects one of the travel routes according to the selection probability.

[0097] The movement calculation unit 127 moves the agent in the virtual space according to the movement route selected by the decision-making unit 126 for each of the multiple OD records. For example, the movement calculation unit 127 advances the time by a fixed amount (for example, one second) and calculates the position of each of the multiple people at each time. The state update unit 128 counts the number of users of each means of transportation from the movement results of the movement calculation unit 127 and determines the dynamic state. As a result, the state update unit 128 updates the state data stored in the state data storage unit 123.

[0098] The index output unit 129 calculates an evaluation index value such as the amount of carbon dioxide emission from the movement result of the movement calculation unit 127. The index output unit 129 notifies the calculated evaluation index value to the simulation control unit 124. Note that the information processing device 100 may store the simulation result in the HDD 103, or may transmit it to another information processing device.

[0099] 11 is a diagram showing examples of components of the state data. The state data stored in the state data storage unit 123 includes road usage data 131, parking lot usage data 132, bicycle parking lot usage data 133, station usage data 134, train usage data 135, bus stop usage data 136, and bus usage data 137.

[0100] Road usage data 131 includes the number of private cars for each road section. Parking lot usage data 132 includes the number of private cars for each parking lot. Bicycle parking lot usage data 133 includes the number of bicycles for each bicycle parking lot. Station usage data 134 includes the number of users for each station. Train usage data 135 includes the number of passengers for each train. Bus stop usage data 136 includes the number of users for each bus stop. Bus usage data 137 includes the number of passengers for each bus.

[0101] Fig. 12 is a flowchart showing an example of the procedure for a traffic simulation. In the following explanation, it is assumed that time progresses in the virtual space by repeating a cycle, as shown in Fig. 4. In step S10, the simulation control unit 124 accepts simulation conditions such as a target area and a target period from the user. In step S11, the simulation control unit 124 randomly generates OD data showing multiple people moving through the target area based on the simulation conditions.

[0102] In step S12, the simulation control unit 124 initializes state data indicating the dynamic state of the transportation facilities. The initial state data may indicate that the congestion levels of all transportation facilities are low. In step S13, the route search unit 125 identifies the start time and end time of the current cycle. The route search unit 125 extracts, from the OD data in step S11, OD records whose departure times fall within the period of the current cycle.

[0103] In step S14, the route search unit 125 searches for multiple travel routes from the departure point to the destination for each OD record extracted in step S13. At this time, the route search unit 125 uses the latest status data in addition to the traffic data indicating the static design of the transportation facility.

[0104] In step S15, the decision-making unit 126 calculates the utility of each of the multiple travel routes from step S14 for each OD record. At this time, the decision-making unit 126 uses the profile and latest status data included in the OD record. The decision-making unit 126 inputs the cost, time, and congestion level of the travel route into a utility function. The decision-making unit 126 relativizes the utilities of the multiple travel routes and calculates the selection probability of each of the multiple travel routes.

[0105] In step S16, the decision-making unit 126 randomly selects one of the travel routes for each OD record according to the selection probability. In step S17, the travel calculation unit 127 identifies people who have not yet arrived at the destination. The people identified here include people who are departing in the current cycle as well as people who have already departed in the previous cycle. The travel calculation unit 127 moves the people who have not yet arrived at the destination along the selected travel route for a certain period of time.

[0106] In step S18, the state update unit 128 calculates the congestion level of each means of transportation according to the movement result of step S17 and updates the state data. In step S19, the simulation control unit 124 determines whether the current time in the virtual space has reached the end point of the target period. If the current time has reached the end point of the target period, the process proceeds to step S20. If the current time has not reached the end point of the target period, the simulation control unit 124 advances the time by a fixed time. Then, the process returns to step S13.

[0107] In step S20, the index output unit 129 aggregates the movement results of step S17 and calculates an evaluation index value. The simulation control unit 124 outputs the simulation results. The simulation results indicate the movement of multiple people along a time axis, the change in congestion level over time for each of multiple transportation modes, and the evaluation index value.

[0108] FIG. 13 is a flowchart showing an example of the procedure for OD record processing. Here, a traffic simulation will be described focusing on one OD record. In step S30, the simulation control unit 124 generates an OD record that indicates a person's transportation demands and profile. In step S31, the route search unit 125 determines whether the departure time in the OD record is included in the period of the current cycle. If the departure time is included in the period of the current cycle, the process proceeds to step S32. If the departure time is not included in the period of the current cycle, step S31 is repeated to wait for the cycle to which the departure time belongs.

[0109] In step S32, the route search unit 125 uses the latest state data to search for multiple travel routes from the departure point to the destination. In step S33, the decision-making unit 126 calculates the utility of each travel route using the latest state data and calculates the selection probability using the utility. In step S34, the decision-making unit 126 randomly selects one travel route from the multiple travel routes according to the selection probability.

[0110] In step S35, the movement calculation unit 127 moves the person indicated by the OD record along the selected movement route for a certain period of time. In step S36, the status update unit 128 reflects the use of transportation simulated in the current cycle in the status data. In step S37, the simulation control unit 124 determines whether the person has arrived at the destination or whether the current time has reached the end of the target period. If the condition is met, processing of the OD record ends. If the condition is not met, processing returns to step S35.

[0111] FIG. 14 is a flowchart showing an example of the procedure for the entry process. Here, a transfer from a private car to public transportation will be described as an example of movement calculation for moving a person along a selected movement route. In step S40, the decision-making unit 126 selects a movement route. In step S41, the movement calculation unit 127 determines whether the selected movement route uses a parking lot. If the movement route uses a parking lot, the process proceeds to step S42. If the movement route does not use a parking lot, the entry process ends.

[0112] In step S42, the movement calculation unit 127 moves the person along the movement route for a certain period of time. If the person is waiting to park, step S42 includes staying at that location for a certain period of time. In step S43, the movement calculation unit 127 determines whether the person has arrived at the parking lot. If the person has arrived at the parking lot, the process proceeds to step S44. If the person has not yet arrived at the parking lot, the process returns to step S42.

[0113] In step S44, the movement calculation unit 127 refers to the latest status data and determines whether there is an available lot in the parking lot. If there is an available lot, the process proceeds to step S48. If there is no available lot, the process proceeds to step S45. In step S45, the movement calculation unit 127 determines whether to wait for parking based on the person's profile. The movement calculation unit 127 may determine whether to wait for parking based on the purpose of the trip, age, and gender. The profile may also include a preference for whether to wait for parking.

[0114] In step S46, the movement calculation unit 127 determines whether or not to wait for parking. If to wait for parking, the process returns to step S42. If not to wait for parking, the process proceeds to step S47. In step S47, the route search unit 125 executes the route search again to search for a route from the current location to the destination that does not require parking the car at the current location. Then, the process returns to step S40. In step S48, the movement calculation unit 127 parks the car in an empty lot in the parking lot. Then, the parking process ends.

[0115] As described above, the information processing device 100 of the second embodiment can execute a multimodal traffic simulation that allows people to move from a departure point to a destination while changing between different modes of transportation. Therefore, the information processing device 100 can simulate the usage of multiple modes of transportation in a certain area and provide information useful for urban planning. Furthermore, the information processing device 100 can reflect the interdependence between the selection of travel routes by multiple people and the dynamic state of the modes of transportation in the multimodal traffic simulation, thereby improving the accuracy of the multimodal traffic simulation. [Explanation of symbols]

[0116] 10. Information processing equipment 11 Storage section 12 Processing section 13a, 13b, 13c Transportation 14a, 14b, 14c Features 15a, 15b Travel route 16 Behavioral Model

Claims

1. Searching for a plurality of travel routes in which a person travels from a departure point to a destination using one or more transportation means among a plurality of transportation means, the one or more transportation means being different; predicting a first travel route that the person will select from among the plurality of travel routes using a behavioral model that predicts selection behavior based on feature amounts that indicate the states of the plurality of transportation facilities; simulating a first movement of the person along a time axis using the first movement path; updating the feature amount using the result of the first movement; A traffic simulation program that executes processing on a computer.

2. the first movement is a movement of the person during a first time period; and causing the computer to further execute a process of predicting a second movement route selected by another person using the updated feature amount, and simulating a second movement of the other person in a second period after the first period using the second movement route.

2. The traffic simulation program according to claim 1.

3. and further causing the computer to execute a process of re-predicting the first movement path using the updated feature amount, and re-simulating the first movement using the re-predicted first movement path.

2. The traffic simulation program according to claim 1.

4. the first travel route includes transfers between different modes of transportation; the feature amount includes a congestion degree of a transfer location where the different transportation modes are transferred; 2. The traffic simulation program according to claim 1.

5. the simulation includes a human behavior simulation that simulates the behavior of a plurality of people in an area using map data showing the area including the plurality of transportation facilities; 2. The traffic simulation program according to claim 1.

6. The simulation includes a process of generating a first digital twin that is a digital twin that reproduces the states of the plurality of transportation systems in the real world in a virtual space, in which the real world and the virtual space are time-synchronized, and simulating the behavior of the plurality of people by moving, in the generated first digital twin, each of a plurality of agents corresponding to each of the plurality of people present in the real world.

2. The traffic simulation program according to claim 1.

7. Searching for a plurality of travel routes in which a person travels from a departure point to a destination using one or more transportation means among a plurality of transportation means, the one or more transportation means being different; predicting a first travel route that the person will select from among the plurality of travel routes using a behavioral model that predicts selection behavior based on feature amounts that indicate the states of the plurality of transportation facilities; simulating a first movement of the person along a time axis using the first movement path; updating the feature amount using the result of the first movement; A traffic simulation method in which processing is performed by a computer.

8. a storage unit that stores feature quantities indicating the states of a plurality of transportation facilities; a processing unit that searches for a plurality of travel routes in which a person travels from a departure point to a destination using one or more of the plurality of transportation modes, the one or more travel routes being different, predicts a first travel route to be selected by the person from among the plurality of travel routes using a behavior model that predicts selection behavior based on the feature amount, simulates a first travel of the person along a time axis using the first travel route, and updates the feature amount using a result of the first travel; An information processing device having the above.

Images