Measures execution program, measures execution method, and information processing device

Abstract

To improve accuracy of measures verification.SOLUTION: An information processing device generates a digital twin in which a real world is reproduced in a virtual space, and acquires information on a person present in a predetermined area in the real world. The information processing device performs simulation related to movement of a person using acquired information of the person in the generated digital twin. The information processing device generates information indicating verification results of measures to be applied to a predetermined area based on results of the executed simulation, and outputs information indicating the generated verification results to a display screen.SELECTED DRAWING: Figure 1

Description

[Technical Field]

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

[0002] Policy verification is required in a variety of situations, and simulations are used to verify policies and estimate the effects and impacts of policies. [Prior art documents] [Patent documents]

[0003] [Patent Document 1] Japanese Patent Application Publication No. 2023-182560 Summary of the Invention [Problem to be solved by the invention]

[0004] However, because conventional techniques perform simulations using preset formulas, they can only obtain simulation results under specific conditions, making it difficult to say that they are capable of performing simulations that are in line with the real world.Furthermore, it is difficult to say that measures developed based on the simulation results will be highly effective in the real world.

[0005] In one aspect, an object of the present invention is to provide a policy execution program, a policy execution method, and an information processing device that can improve the accuracy of policy verification. [Means for solving the problem]

[0006] In the first proposal, the policy execution program causes a computer to execute the following processes: generate a digital twin that recreates the real world in a virtual space; acquire information about people present in a specified area of ​​the real world; use the acquired information about the people in the generated digital twin to conduct a simulation of the movements of the people; generate information indicating the verification results of the policies to be applied to the specified area based on the results of the conducted simulation; and output the generated information indicating the verification results to a display screen. [Effects of the Invention]

[0007] According to one embodiment, the accuracy of policy verification can be improved. [Brief explanation of the drawings]

[0008] [Figure 1] FIG. 1 is a diagram illustrating the execution of a measure by the information processing device according to the first embodiment. [Figure 2] FIG. 2 is a diagram illustrating an example of generating destination data. [Figure 3] FIG. 3 is a diagram illustrating an example of deterioration in accuracy of destination data. [Figure 4] FIG. 4 is a functional block diagram of the information processing apparatus according to the first embodiment. [Figure 5] FIG. 5 is a diagram illustrating an example of generating trip purpose data. [Figure 6] FIG. 6 is a diagram illustrating an example of generating destination data. [Figure 7] FIG. 7 is a diagram illustrating an example of adding destination data. [Figure 8] FIG. 8 is a diagram illustrating an example of attribute generation. [Figure 9] FIG. 9 is a diagram illustrating an example of generating attributed movement data. [Figure 10] FIG. 10 is a diagram for explaining an example of analysis of a measure. [Figure 11] FIG. 11 is a diagram illustrating an example of displaying the effects of measures. [Figure 12]FIG. 12 is a flowchart showing the flow of the process. [Figure 13] FIG. 13 is a flowchart showing a simulation process using a behavior selection model. [Figure 14] FIG. 14 is a diagram illustrating an example of a hardware configuration. DETAILED DESCRIPTION OF THE INVENTION

[0009] The following describes in detail embodiments of the policy execution program, policy execution method, and information processing device disclosed herein with reference to the accompanying drawings. Note that the present invention is not limited to these embodiments. Each embodiment can be appropriately combined within a consistent range. [Example]

[0010] (Explanation of the policy implementation system) FIG. 1 is a diagram illustrating the implementation of a policy by an information processing device 10 according to a first embodiment. The information processing device 10 shown in FIG. 1 executes a simulation on a digital twin as an effort to digitally reproduce human behavior and explore and verify policies to solve social issues. The information processing device 10 receives as input human movement data (movement trajectory data) with attribute information such as age, income, destination, and transportation method, and verifies the effectiveness of the policies based on people's reactions to the policies in the digital twin.

[0011] Specifically, the information processing device 10 generates a digital twin that recreates the real world in a virtual space and acquires information about people who exist in a specified area in the real world. The information processing device 10 then uses the acquired information about people in the generated digital twin to perform a simulation of the movement of people. The information processing device 10 then generates information indicating the verification results of measures to be applied to the specified area based on the results of the simulation, and outputs the generated information indicating the verification results to a display screen.

[0012] For example, the information processing device 10 acquires data on persons residing in a certain area, including movement data of person A, who has attribute information of "purpose of travel: shopping, income level: low, destination: A," and movement data of person B, who has attribute information of "purpose of travel: commuting, income level: high, destination: B." The information processing device 10 then uses a digital twin to verify how the movement trajectories of person A and person B would change if a policy of "setting up a toll area in a certain area to alleviate traffic congestion in that area" were implemented. Here, the movement trajectories of persons are not limited to walking but also include cars, taxis, buses, and the like. For example, in the example of FIG. 1, the movement route of person B, who has a high income, remains unchanged before and after the setting up of the toll area, but the movement route of person A, who has a low income, changes. In other words, a simulation of the implementation of the policy of setting up a toll area demonstrates the expected effect of reducing congestion in a specified area.

[0013] In this way, the information processing device 10 can improve the accuracy of measure verification by constructing a digital twin using data from the real environment and executing and verifying various measures on the digital twin.

[0014] However, since it is difficult to obtain movement data with attribute information from the perspective of protecting personal information, movement data with attribute information is created based on easily available human movement data (without attribute information) and statistical data.

[0015] For example, an example of destination data to which a destination, which is one piece of attribute information, is assigned will be described. FIG. 2 is a diagram illustrating an example of generating destination data. As shown in FIG. 2, the destination data is generated by calculating the selection probability for each destination candidate using a multinomial logit model with explanatory variables such as the purpose of travel, income level, distance from home to the destination candidate, and number of employed people. For example, the utility of the destination, "u", is calculated using "i" indicating the destination, the coefficient β of the explanatory variable, and each explanatory variable, X, such as distance, purpose of travel, income level, and means of travel, and the selection probability of each destination, "p i" is calculated. Then, the "p i " and the "p" closest to the random variable i " is selected as the destination. Examples of destinations include a company at "i=1", a school at "i=1", a store at "i=3", a hospital at "i=4", etc.

[0016] However, compared to trips for shopping, medical appointments, etc., destination data for commuting to work or school can deviate from the true value (statistical data). Figure 3 is a diagram illustrating an example of a deterioration in destination data accuracy. Figure 3 shows the number of people who fall into the combination of Zone 1, where the home is located, and each zone (Zones 2 to 5) where the workplace is located. This represents the number of people who fall into the combination of Zone 1, where the home is located, and each zone (Zones 2, 3, 4, and 5) in statistical data from the census, etc., and the number of people who fall into Zone 1 and each zone (Zone 2, 3, 4, and 5) in the destination data generated using the method in Figure 2.

[0017] As can be seen from Figure 3, in the statistical data, the order of the number of people is "number of people heading from Section 1 to Section 2," "number of people heading from Section 1 to Section 5," "number of people heading from Section 1 to Section 4," and "number of people heading from Section 1 to Section 3," whereas in the destination data, the order of the number of people is "number of people heading from Section 1 to Section 5," "number of people heading from Section 1 to Section 3," "number of people heading from Section 1 to Section 2," and "number of people heading from Section 1 to Section 4," resulting in a discrepancy between the two sets of data.

[0018] This is thought to be because when the purpose of travel is to commute to work or school, the destination is not necessarily determined solely by distance or income level. Even if measures are verified using digital twins using such destination data, the verification accuracy will decrease, and the verification accuracy of the measures will also deteriorate.

[0019] Therefore, the information processing device 10 according to the first embodiment uses different methods for generating destination data depending on the purpose of travel. For example, when the purpose of travel is commuting to work or school, the information processing device 10 generates destination data using a selection probability calculated from statistical data, and when the purpose of travel is other than that, such as shopping or visiting a hospital, the information processing device 10 generates destination data using a selection probability calculated from attribute information.

[0020] As a result, when evaluating measures using a digital twin, the information processing device 10 can generate destination data that does not deviate from statistical data even if the number of destination data is small, thereby suppressing deterioration in the verification accuracy of the measures.

[0021] (Functional configuration) 4 is a functional block diagram illustrating a functional configuration of the information processing device 10 according to Example 1. As illustrated in FIG. 4, the information processing device 10 is an example of a computer including a communication unit 11, a display unit 12, a storage unit 20, and a control unit 30.

[0022] The communication unit 11 is a processing unit that controls communication with other devices, and is realized by, for example, a communication interface, etc. For example, the communication unit 11 receives information, statistical data, etc. related to the policy to be verified from a management terminal used by the administrator, and transmits the verification results, simulation results, etc. to the management terminal.

[0023] The display unit 12 is a processing unit that displays and outputs various types of information, and is realized by, for example, a display, a touch panel, etc. For example, the display unit 12 displays and outputs verification results, simulation results, etc.

[0024] The storage unit 20 is a processing unit that stores various data and programs executed by the control unit 30, and is realized by, for example, a memory, a hard disk, etc. The storage unit 20 stores a statistical data DB 21, a destination data DB 22, a destination estimation model 23, and a policy data DB 24.

[0025] The statistical data DB21 is a database that stores statistical data collected and made public by prefectures and municipalities through censuses, etc. For example, the statistical data DB21 stores destination data between sections indicating the number of people traveling from a section to a destination section, income data indicating the income of users belonging to each section, worker data indicating the number of workers in each section, etc.

[0026] The destination data DB 22 is a database that stores destination data among the statistical data. For example, the destination data DB 22 stores data on the number of people who have each section (section 2, section 3, section 4, section 5) of a certain area as their destination, from section 1 of the area.

[0027] The destination estimation model 23 is a model for estimating a destination when the purpose is other than commuting to work or school. For example, the destination estimation model 23 is the formula shown in FIG.

[0028] The policy data DB 24 is a database that stores data related to policies to be verified. For example, the policy data DB 24 stores a policy to make a certain road in a certain area a toll area, a policy to consolidate hospitals in a certain area, and the like.

[0029] The control unit 30 is a processing unit that controls the entire information processing device 10, and is realized by, for example, a processor. This control unit 30 has a digital twin execution unit 40, a generation processing unit 50, and a policy processing unit 60. Note that the digital twin execution unit 40, the generation processing unit 50, and the policy processing unit 60 are realized by, for example, electronic circuits included in the processor or processes executed by the processor.

[0030] The digital twin execution unit 40 is a processing unit that generates a digital twin that recreates the real world in a virtual space. Specifically, when generating a digital twin of a certain area, the digital twin execution unit 40 generates a digital twin that virtually recreates the area using real-world environmental data such as road information, traffic information, weather information, and people information for that area. In addition, the digital twin execution unit 40 can virtually represent a situation in which the environment has changed by dynamically changing weather information, traffic congestion information, and the like.

[0031] Furthermore, for example, the digital twin execution unit 40 recreates objects such as roads and buildings in the digital twin based on map data of roads, buildings, and the like in the real world. The digital twin execution unit 40 then recreates the operation status of each of multiple modes of transportation in the digital twin based on, for example, actual operation data of the modes of transportation. The digital twin execution unit 40 also recreates the circumstances of accidents that have occurred on roads and weather conditions in the digital twin based on, for example, sensing data from sensors placed in the real world.

[0032] The generation processing unit 50 has a travel purpose data generation unit 51, a destination data generation unit 52, an attribute generation unit 53, and a travel data generation unit 54, and is a processing unit that generates travel data, which is information for simulating the movement of people in a digital twin and is used to evaluate measures.

[0033] The trip destination data generation unit 51 is a processing unit that generates trip destination data 100 indicating the trip destination of a person. Fig. 5 is a diagram illustrating an example of trip destination data generation. As shown in Fig. 5, the trip destination data generation unit 51 extracts the residence and trip destination of each person from the statistical data stored in the statistical data DB 21.

[0034] For example, the trip purpose data generation unit 51 generates trip purpose data 100 that associates a "person ID" that identifies a person, a "residence ID" that identifies the person's place of residence, and a "trip purpose" that indicates the purpose of the person's trip. The example in FIG. 5 shows that a person with "person ID=A" lives in an area with "residence ID=0001" and traveled from that area with a "trip purpose" of "commute." It also shows that a person with "person ID=B" lives in an area with "residence ID=0002" and traveled from that area with a "trip purpose" of "shopping."

[0035] The destination data generation unit 52 is a processing unit that generates destination data related to destinations traveled by a person and merges it with the travel purpose data 100. Specifically, if the person's travel purpose is commuting to work or school, the destination data generation unit 52 generates destination data based on a selection probability calculated from statistical data. On the other hand, if the person's travel purpose is other than commuting to work or school, the destination data generation unit 52 generates destination data based on a selection probability calculated from the person's attribute information.

[0036] FIG. 6 is a diagram illustrating an example of generating destination data. As shown in FIG. 6, the destination data generation unit 52 extracts the travel purpose of each person from the statistical data stored in the statistical data DB 21. Then, the destination data generation unit 52 generates destination data for people whose travel purpose is other than commuting to work or school, such as shopping, going to the hospital, or being picked up or dropped off, using the selection probability described in FIG. 2. For example, the destination data generation unit 52 generates destination data 101 for destinations other than commuting to work or school, in which a "person ID" that identifies a person, a "travel purpose" that identifies the person's travel purpose, and a "destination ID" that identifies the destination to which the person traveled are associated. The example in FIG. 6 indicates that a person with "person ID=B" traveled to "destination ID=0001" for "travel purpose=shopping," and that a person with "person ID=C" traveled to "destination ID=0002" for "travel purpose=hospital visit."

[0037] On the other hand, for a person whose travel purpose is commuting to work or school, the destination data generation unit 52 determines the destination based on the distribution of the number of people commuting to school or work at each destination in the person's place of residence. For example, for a person in Section 1, the destination data generation unit 52 may select Section 2, which has the highest population distribution, as the destination, or select a destination closest to a random variable as in Figure 2, or aggregate the population distribution by person attribute (such as gender) and select a destination with the highest number of people with the same attribute. For example, the destination data generation unit 52 generates destination data 102 for commuting to work and school, in which "person ID," "purpose of travel," and "destination ID" are associated with each other. The example in Figure 6 shows that a person with "person ID = A" traveled to "destination ID = 0002" with "purpose of travel = commuting."

[0038] Thereafter, the destination data generation unit 52 generates destination data 101 other than commuting or going to school and destination data 102 for commuting or going to school, and merges these with the travel purpose data 100 to generate destination data 103 of "place of residence, travel purpose, destination."

[0039] Fig. 7 is a diagram illustrating an example of adding destination data. As shown in Fig. 7, the destination data generation unit 52 generates destination data 103 of "residence, purpose of travel, destination" by merging trip purpose data 100 of "person ID, residence ID, purpose of travel," destination data 101 other than commuting to work / school of "person ID, purpose of travel, destination ID," and destination data 102 of commuting to work / school of "person ID, purpose of travel, destination ID."

[0040] For example, the destination data generation unit 52 collects data associated with "person ID=A" from each of the trip purpose data 100, the destination data other than commuting / schooling 101, and the trip purpose data 102, and generates "person ID=A, residence ID=0001, trip purpose=commuting, destination ID=0003." Similarly, the destination data generation unit 52 collects data associated with "person ID=B" and generates "person ID=B, residence ID=0002, trip purpose=shopping, destination ID=0001."

[0041] The attribute generation unit 53 is a processing unit that generates attributes other than the destination. Specifically, if attributes such as gender, age, income, and means of transportation are included in the statistical data, the attribute generation unit 53 uses those attributes, and if they are not included in the statistical data, the attribute generation unit 53 generates attributes using the model described in FIG.

[0042] FIG. 8 is a diagram illustrating an example of attribute generation. The attribute generation unit 53 generates attributes for attributes not included in statistical data using the same method as in FIG. 2 shown in FIG. 8. For example, taking the attribute "transportation means" as an example, as shown in FIG. 8, the attribute generation unit 53 calculates the utility "u" of transportation means using explanatory variables such as "distance, income, whether or not a person owns a car," and calculates the selection probability "p" of each transportation means using the utility "u." i Then, the attribute generation unit 53 calculates "p i " and the "p" closest to the random variable i " is selected as the means of transportation. Examples of means of transportation include walking at "i=1", driving at "i=1", bus at "i=3", and bicycle at "i=4".

[0043] Furthermore, it is preferable to use information that affects or is related to the calculation target (the means of transportation in Figure 8) as explanatory variables. In the example of Figure 8, the means of transportation often varies depending on whether a person owns a car and the distance traveled, so these are used as explanatory variables in the calculation. Note that, since it is conceivable that people commute to work after taking their children to and from nursery school or kindergarten, it is also useful to use family composition and other factors as explanatory variables.

[0044] The movement data generation unit 54 is a processing unit that generates movement data for simulating human movement using a digital twin. Specifically, the movement data generation unit 54 generates attributed movement data 104 by combining the "residence, purpose of movement, destination" data 103 generated by the destination data generation unit 52 with the attributes generated by the attribute generation unit 53.

[0045] Fig. 9 is a diagram illustrating an example of generating attributed travel data 104. As shown in Fig. 9, the travel data generation unit 54 generates attributed travel data by adding the attributes "gender = female, age = 28, income = 2 million..." generated for the person with "person ID = A" to "person ID = A, residence ID = 0001, travel purpose = commute, destination ID = 0003" in the "residence, travel purpose, destination" data 103.

[0046] The policy processing unit 60 has a policy implementation unit 61, a policy analysis unit 62, and a visualization unit 63, and is a processing unit that verifies the policy by using the attributed movement data generated by the generation processing unit 50 to run a simulation on the digital twin regarding the movement of people when the policy is implemented.

[0047] The policy implementation unit 61 is a processing unit that executes a simulation on the digital twin regarding the movement of people when a policy is implemented in a certain area of ​​a certain region. Specifically, the policy implementation unit 61 executes a simulation on the digital twin using the attributed movement data 104 to move agents corresponding to each of multiple people included in the attributed movement data 104.

[0048] For example, as an example of a measure to eliminate traffic congestion, a case will be described in which an area with heavy traffic or a high number of traffic accidents, which is an example of a predetermined area in a certain area A, is set as a toll area. First, the measure implementation unit 61 executes a movement simulation before the measure. Specifically, the measure implementation unit 61 generates the area A on the digital twin, places each person who resides in the area A among the people included in the attributed movement data 104, and recreates a virtual area A that is identical to the real environment.

[0049] In this situation, the policy implementation unit 61 uses a digital twin to generate agents corresponding to each person using attributed movement data 104 of people placed in virtual area A, and performs a simulation of the movement of each agent to identify the movement trajectory of each person before the policy was implemented.

[0050] Next, the policy implementation unit 61 sets the area A on the digital twin as a charging area. Then, the policy implementation unit 61 generates an agent corresponding to each person using the attributed movement data 104 of people placed in the virtual area A where the charging area has been set, and identifies the movement trajectory of each person after the policy is implemented by performing a simulation in which each agent moves.

[0051] Here, the policy implementation unit 61 can synchronize the time between the real environment and the virtual environment (digital twin) and run a simulation during the same time period as the real environment. Furthermore, the policy implementation unit 61 can set any conditions to be verified, such as weather information such as rain or snow, information on event days such as New Year's Day or concert dates, and traffic information such as road closures and one-way streets, in the digital twin, thereby running multiple realistic simulations corresponding to each expected situation.

[0052] The policy analysis unit 62 is a processing unit that generates verification results of policies related to people's movements to be applied to a predetermined area based on the results of the simulation by the policy implementation unit 61. Specifically, the policy analysis unit 62 analyzes movement trajectories before and after the policies, which are the results of the simulation, and identifies the situation in the predetermined area and the congestion state in the predetermined area based on the analysis results.

[0053] Fig. 10 is a diagram illustrating an example of policy analysis. As shown in Fig. 10, the policy analysis unit 62 identifies changes in the movement trajectory of person P with attribute information "purpose of travel: shopping, income level: low, destination: A" and changes in the movement trajectory of person Q with attribute information "purpose of travel: commuting, income level: high, destination: B" regarding the movement routes to destinations A and B in the same area before and after the setting of a charging area.

[0054] For example, the policy analysis unit 62 displays the charging area on the digital twin, and also displays the movement routes of person P and person Q before the policy obtained by the policy implementation unit 61 with dotted lines, and displays the movement routes of person P and person Q after the policy obtained by the policy implementation unit 61 with solid lines. Then, the policy analysis unit 62 determines that person P's movement trajectory has changed to a trajectory that does not pass through the charging area, and that there is no change in person Q's movement trajectory.

[0055] The policy analysis unit 62 then analyzes the effect of the policy from the change in the trajectory and the attribute information. In this example, since the purpose of travel and income of person P and person Q are different, the policy analysis unit 62 analyzes that if the purpose of travel is shopping, where cost takes priority over time, or if the income is low, the person will react by detouring the charging area.

[0056] 10 illustrates an example of analysis using the movement trajectories of two people, but this is merely an example. For example, the policy analysis unit 62 can also analyze changes in the number of vehicles within a toll area or changes in the number of congested roads.

[0057] In this way, the policy analysis unit 62 generates verification results such as changes in people's movement trajectories before and after the policy, changes in the number of vehicles in the charging area, and changes in the number of congested roads, and outputs them to the visualization unit 63. Note that the policy analysis unit 62 can also store data changes during the simulation (for example, changes in movement trajectories and changes in the number of vehicles) and output them to the visualization unit 63.

[0058] The visualization unit 63 is a processing unit that visualizes the verification results generated by the policy analysis unit 62. Specifically, the visualization unit 63 generates a display screen showing verification results such as changes in people's movement trajectories before and after the policy, changes in the number of vehicles in the charging area, changes in the number of congested roads, etc., and outputs the results to the display unit 12 or transmits them to the administrator terminal.

[0059] FIG. 11 is a diagram illustrating an example of a display of the effect of a policy. As shown in FIG. 11, the visualization unit 63 generates a display screen 200 including an area 201 showing the policy content, an area 202 showing simulation information of the policy target, and an effect 203 of the policy implementation. In the example of FIG. 11, the visualization unit 63 displays the policy content "impose a fee on sections where congestion frequently occurs" in the area 201. The visualization unit 63 also displays the area to be simulated and the results of the simulation in the area 202, and displays the time series changes in the simulation by accepting the operation of the button 204. The visualization unit 63 also displays a graph showing the time series changes in the number of vehicles in the charging area before and after the policy and a graph showing the time series changes in the number of congested roads before and after the policy in the area 203.

[0060] (Processing flow) Fig. 12 is a flowchart showing the flow of processing. As shown in Fig. 12, when an administrator or the like issues an instruction to start processing (S101: Yes), the information processing device 10 generates a plurality of travel purpose data from statistical data (S102).

[0061] Next, the information processing device 10 selects one trip purpose data (S103) and determines whether the trip purpose corresponds to commuting to work or school (S104). If the trip purpose corresponds to commuting to work or school (S104: Yes), the information processing device 10 generates destination data from statistical data (S105). On the other hand, if the trip purpose is neither commuting to work nor school (S104: No), the information processing device 10 generates destination data from an estimation model (S106).

[0062] Thereafter, the information processing device 10 adds (combines) the destination data to the trip purpose data (S107), generates and further adds attribute information (S108), and generates attributed trip data (S109).

[0063] Here, if there is unprocessed travel destination data (S110: Yes), the information processing device 10 repeats S103 and subsequent steps. On the other hand, if there is no unprocessed travel destination data (S110: No), the information processing device 10 executes a simulation of the measure using the digital twin (S111), analyzes the effect of the measure (S112), and outputs the analysis result (S113).

[0064] (effect) As described above, the information processing device 10 can verify measures using a digital twin, so that measures can be verified not only in the actual environment but also in an expected environment, thereby improving the accuracy of measure verification.

[0065] Furthermore, the information processing device 10 can generate destination data for commuting to work or school, and destination data for other purposes, using methods appropriate for each, and can therefore generate highly accurate destination data and attributed travel data that do not deviate from statistical data.

[0066] Furthermore, the information processing device 10 can generate attributes that are not included in the statistical data and combine them with the destination data, so that highly accurate attributed travel data that does not deviate from the statistical data can be generated.

[0067] Furthermore, the information processing device 10 executes a simulation of a person's movement using highly accurate destination data and attributed movement data, thereby improving the accuracy of the simulation. Furthermore, the information processing device 10 can verify measures using highly accurate simulations, thereby improving the accuracy of verifying the measures. [Example]

[0068] Although the embodiments of the present invention have been described above, the present invention may be embodied in various different forms other than the above-described embodiments.

[0069] (Numbers, etc.) The attribute names, models, numerical values, graphs, etc. used in the above embodiments are merely examples and can be changed as desired. Furthermore, the process flow described in each flowchart can also be changed as appropriate within a consistent range.

[0070] (Simulation using a behavioral choice model) The information processing device 10 can perform a simulation and display the verification result by using a behavior selection model that determines whether or not a person will act in accordance with a policy.

[0071] For example, the control unit 30 of the information processing device 10 acquires a behavior selection model that determines whether a person will act in accordance with a measure. Next, the control unit 30 of the information processing device 10 executes a simulation in the digital twin to determine whether an agent will comply with a first measure, for example, using the person's attribute information and the behavior selection model. Then, the control unit 30 of the information processing device 10 executes a simulation regarding the movement of the person, for example, using the execution result of the simulation that determines whether the agent will comply. Thereafter, the control unit 30 of the information processing device 10 generates information indicating the verification result of the measure to be applied to a specified area, for example, based on the result of the executed simulation.

[0072] 13 is a flowchart showing a simulation process using a behavior selection model. First, the information processing device 10 generates a digital twin that recreates the real world in a virtual space (step S201). Next, the information processing device 10 places agents corresponding to people existing in the real world in the digital twin (step S202).

[0073] The information processing device 10 identifies a first measure to be applied to a predetermined area in the real world (step S203). The information processing device 10 acquires a behavior selection model that determines whether a person will act in accordance with the first measure (step S204). Then, the information processing device 10 executes a simulation in the digital twin using attribute information of the person corresponding to the agent and the behavior selection model to determine whether the agent will follow the first measure (step S205). The information processing device 10 generates a verification result of the first measure applied to the predetermined area by performing a simulation regarding the movement of the person using the result of the executed simulation (step S206). The information processing device 10 displays the generated verification result on a display screen (step S207).

[0074] More specifically, the information processing device 10 acquires, via communication, terminal data from a terminal used by a person present in a predetermined area in the real world. Next, based on the acquired terminal data, the information processing device 10 places an agent corresponding to the person on a digital twin in which the virtual space and the real world are time-synchronized. The information processing device 10 then links the placed agent with the person's movement data accompanied by attribute information. At this time, the information processing device 10 identifies the agent's attribute information by associating the placed agent with the person's attribute information.

[0075] The information processing device 10 also acquires a behavior selection model, which is a machine learning model for determining whether a person will act in accordance with a measure. Next, the information processing device 10 inputs attribute information of an agent existing in the digital twin into the behavior selection model, thereby executing a simulation for determining whether the agent will comply with the first measure. Then, the information processing device 10 identifies the situation in a specified area based on the results of the executed simulation.

[0076] The information processing device 10 identifies a first measure to be applied to a predetermined area in the real world. The information processing device 10 receives the measure from, for example, an administrator's terminal. Then, the information processing device 10 identifies a behavior selection model for the measure to be applied to the predetermined area in the real world. At this time, the information processing device 10 identifies a behavior selection model for the first measure to be applied to the predetermined area in the real world.

[0077] More specifically, the information processing device 10 inputs attribute information of agents present in the digital twin into the acquired behavior selection model to determine whether each of the multiple agents will comply with the first measure. Next, based on the determination result, the information processing device 10 executes a simulation to determine whether each of the multiple agents will move through a predetermined area. Then, based on the execution result of the simulation to determine whether or not the agents will move through the predetermined area, the information processing device 10 identifies the degree of congestion in the predetermined area.

[0078] For example, the information processing device 10 inputs human movement data with attribute information of a first person's agent into a machine learning model. At this time, for example, the information processing device 10 determines to act in accordance with a first measure based on the output result of the machine learning model, and does not move the first person's agent from a predetermined area. On the other hand, the information processing device 10 inputs human movement data with attribute information of a second person's agent into the machine learning model. At this time, the information processing device 10 determines not to act in accordance with the first measure based on the output result of the machine learning model, and moves the second person's agent from the predetermined area. Then, the information processing device 10 identifies the degree of congestion in the predetermined area based on the movement results of each of the multiple agents. Thereafter, the information processing device 10 generates a verification result of the measure applied to the predetermined area using the identified degree of congestion.

[0079] When introducing social policies, which cannot afford to fail, into the real world, prior verification of the policies is required. In this case, the information processing device 10 can improve the accuracy of the policy verification. Furthermore, by using a simulation that utilizes a behavioral selection model, the load on the computer processing when performing the simulation can be reduced.

[0080] The behavior selection model is trained based on a ground truth label of information regarding whether or not a person will act in accordance with the first measure and training data including attribute information of the person. For example, the information processing device 10 updates the parameters of the neural network of the machine learning model based on the output result of the machine learning model when human movement data with person attribute information is input to the machine learning model and the ground truth label indicating information regarding whether or not the person will act in a predetermined area in accordance with the first measure. This allows the information processing device 10 to generate a machine learning model that determines whether or not a person will act in accordance with the measure.

[0081] (Destination data) In the above embodiment, the information processing device 10 executes a simulation using attributed movement data, but the present invention is not limited to this. For example, the information processing device 10 can execute a simulation using only destination data. In this case, since the amount of information is small, the information processing device 10 can execute a high-speed simulation that is narrowed down to the movement trajectory from the departure point to the destination, thereby reducing the time required for policy verification while maintaining the accuracy of the verification.

[0082] Furthermore, when generating destination data, the information processing device 10 can also generate the destination data using a pre-trained model. For example, the information processing device 10 prepares estimation models that input attributes of a person known from statistical data such as gender and place of residence and output destinations for each purpose of travel, such as an estimation model for commuting to work or school, an estimation model for shopping, and an estimation model for hospital. The information processing device 10 can then generate destinations using an estimation model corresponding to the purpose of travel, and generate destination data using the generated destinations.

[0083] (simulation) In the above embodiment, the information processing device 10 executes a simulation of human movement, but the present invention is not limited to this. For example, the information processing device 10 can also execute a traffic simulation of road traffic such as automobiles, buses, and taxis. Note that each simulation can employ various commonly used calculation formulas and simulation methods.

[0084] (system) The information including the processing procedures, control procedures, specific names, various data and parameters shown in the above documents and drawings may be changed arbitrarily unless otherwise specified.

[0085] Furthermore, the specific form of distribution and integration of the components of each device is not limited to that shown in the figure. For example, the digital twin execution unit 40, generation processing unit 50, and policy processing unit 60 may be integrated. In other words, all or some of the components may be functionally or physically distributed or integrated in any unit depending on various loads, usage conditions, etc. Furthermore, all or any part of the processing functions of each device may be realized by a CPU and a program analyzed and executed by the CPU, or may be realized as hardware using wired logic.

[0086] Furthermore, all or any part of the processing functions performed by each device may be realized by a CPU and a program analyzed and executed by the CPU, or may be realized as hardware using wired logic.

[0087] (Hardware) Fig. 14 is a diagram illustrating an example of a hardware configuration. As shown in Fig. 14, an information processing device 10 includes a communication device 10a, a hard disk drive (HDD) 10b, a memory 10c, and a processor 10d. The components shown in Fig. 14 are connected to each other via a bus or the like.

[0088] The communication device 10a is a network interface card or the like, and communicates with other devices. The HDD 10b stores programs and DBs that operate the functions shown in FIG.

[0089] The processor 10d reads out from the HDD 10b or the like a program that executes the same processing as each processing unit shown in FIG. 4 and expands it into the memory 10c, thereby operating a process that executes each function described in FIG. 4 or the like. For example, this process executes the same functions as each processing unit possessed by the information processing device 10. Specifically, the processor 10d reads out from the HDD 10b or the like a program that has the same functions as the digital twin execution unit 40, the generation processing unit 50, the policy processing unit 60, etc. Then, the processor 10d executes a process that executes the same processing as the digital twin execution unit 40, the generation processing unit 50, the policy processing unit 60, etc.

[0090] In this way, the information processing device 10 operates as an information processing device that executes a policy verification and estimation method by reading and executing a program. The information processing device 10 can also realize functions similar to those of the above-described embodiment by reading the program from a recording medium using a medium reading device and executing the read program. Note that the program in these other embodiments is not limited to being executed by the information processing device 10. For example, the above-described embodiment may also be applied in the same way to cases where another computer or server executes the program, or where these execute the program in cooperation with each other.

[0091] This program may be distributed via a network such as the Internet. Alternatively, this program may be recorded on a computer-readable recording medium such as a hard disk, a flexible disk (FD), a CD-ROM, a magneto-optical disk (MO), or a digital versatile disk (DVD), and may be read out from the recording medium and executed by a computer. [Explanation of symbols]

[0092] 10. Information processing equipment 11 Communications Department 12 Display section 20 Memory section 21 Statistical Data DB 22 Destination Data DB 23 Destination estimation model 24 Policy Data DB 30 Control Unit 40 Digital Twin Executive Department 50 Generation processing section 51 Travel purpose data generation unit 52 Destination data generation unit 53 Attribute generator 54 Movement data generation unit 60 Policy Processing Department 61 Policy Implementation Department 62 Policy Analysis Department 63 Visualization section

Claims

1. On the computer, We create a digital twin that recreates the real world in a virtual space. acquiring information about people present in a predetermined area of ​​the real world; In the generated digital twin, a simulation of the movement of the person is performed using the acquired information of the person; generating information indicating a verification result of the measures to be applied to the predetermined area based on the results of the implemented simulation; outputting the generated information indicating the verification result to a display screen; A policy execution program characterized by executing a process.

2. The process of performing the simulation includes: Using the digital twin, a traffic simulation is performed using destination data related to destinations in the movement of the person; The process of generating information indicating the verification result includes: generating information indicating a verification result of the measures regarding the movement of the people to be applied to the predetermined area based on the results of the traffic simulation; 2. The policy execution program according to claim 1, wherein:

3. generating destination data for the movement of the person using attribute information of the person; The process of performing the simulation includes: Using the digital twin, a simulation is performed regarding the movement of the person using destination data regarding destinations in the movement of the person; The process of generating information indicating the verification result includes: generating information indicating a verification result of the measures regarding the movement of the person to be applied to the predetermined area based on the results of the simulation; 3. The policy execution program according to claim 1 or 2.

4. Acquire the purpose of travel of the person; further causing the computer to execute a process of identifying a model associated with the person's purpose of travel from among a plurality of models using the person's purpose of travel that has been acquired; The process of generating destination data includes: generating the destination data using the identified model; 4. The policy execution program according to claim 3.

5. If the purpose of travel of the person is commuting to work or school, the destination data is generated based on a selection probability calculated from statistical data; If the purpose of travel of the person is other than commuting to work or school, the destination data is generated based on a selection probability calculated from attribute information of the person.

4. The policy execution program according to claim 3.

6. causing the computer to further execute a process of generating first combined data by combining the destination data generated from the statistical data with attribute information of the person, and second combined data by combining the destination data generated from the attribute information of the person with attribute information of the person; The process of performing the simulation includes: using the digital twin to perform a simulation of the movement of the person using the first combined data and the second combined data; 6. The policy execution program according to claim 5,

7. The process of acquiring information about the person includes: Acquire destination data in which a destination is associated with each person, The process of executing the simulation includes: Using the acquired destination data, a simulation is performed on the digital twin in which agents corresponding to each of a plurality of people are moved; The process of generating information indicating the verification result includes: Identifying the situation of the predetermined area based on the results of the simulation.

2. The policy execution program according to claim 1, wherein:

8. The measures to be applied to the predetermined area are measures to alleviate traffic congestion, The process of generating information indicating the verification result includes: Identifying the congestion situation in a predetermined area based on the results of the simulation; generating information indicating a verification result of a measure to be applied to the predetermined area using the identified congestion situation; 8. The policy execution program according to claim 7,

9. The process of performing the simulation includes: obtaining a behavior selection model that determines whether a person will act in accordance with the policy; using the acquired behavior selection model and the acquired information on the person, in the digital twin, executing a simulation to determine whether or not an agent will comply with a first measure; performing a simulation regarding the movement of the person using the execution result of the simulation for determining whether the agent will follow; 2. The policy execution program according to claim 1, wherein:

10. the person information is attribute information of the person, The process of performing the simulation includes: determining whether each of a plurality of agents will comply with a first measure based on the attribute information of agents present in the digital twin and the acquired behavior selection model, thereby performing a simulation to determine whether each of the plurality of agents will move within a predetermined area; Identifying a congestion situation in the predetermined area based on a result of a simulation of whether or not to move through the predetermined area; generating information indicating a verification result of the first measure to be applied to the predetermined area using the identified congestion situation; 10. The policy execution program according to claim 9,

11. The process of generating the digital twin includes: acquires terminal data from a terminal used by a person present in a predetermined area of ​​the real world through communication; Based on the acquired terminal data, an agent corresponding to the person is placed on a digital twin in which the virtual space and the real world are time-synchronized; and Associating attribute information of the person with the agent placed on the digital twin; The process of performing the simulation includes: accepting a first policy to be applied to the predetermined area; acquiring a behavior selection model for the first measure, which is a machine learning model for determining whether the person will act in accordance with the measure; running a simulation in the digital twin to determine whether the agent will comply with the first measure based on the acquired behavior selection model for the first measure and attribute information of the person; 11. The policy execution program according to claim 10.

12. The computer We create a digital twin that recreates the real world in a virtual space. acquiring information about people present in a predetermined area of ​​the real world; In the generated digital twin, a simulation of the movement of the person is performed using the acquired information of the person; generating information indicating a verification result of the measures to be applied to the predetermined area based on the results of the implemented simulation; outputting the generated information indicating the verification result to a display screen; A policy execution method characterized by executing a process.

13. We create a digital twin that recreates the real world in a virtual space. acquiring information about people present in a predetermined area of ​​the real world; In the generated digital twin, a simulation of the movement of the person is performed using the acquired information of the person; generating information indicating a verification result of the measures to be applied to the predetermined area based on the results of the implemented simulation; outputting the generated information indicating the verification result to a display screen; An information processing device comprising a control unit.

Images