Optimization method, optimization device, and optimization program

Abstract

In this optimization method: a first input that includes a first user's departure time, departure location, and destination is obtained; and a second input that includes a second user's departure time, departure location, and destination is obtained. On the basis of both the first input and the second input, a route to be presented to the first user and a route to be presented to the second user are calculated, using a digital twin, so as to optimize a predetermined index. The route to be presented to the first user is presented to the first user, and the route to be presented to the second user is presented to the second user.

Description

[Technical Field]

[0001] The disclosed technology relates to an optimization method, an optimization device, and an optimization program. [Background technology]

[0002] Modeling objects or events that exist in the real world on a computer and feeding back the results of the computer simulation into the real world has been studied for some time. Hereinafter, a computer model of an object or event that exists in the real world will be referred to as a digital twin, or DT.

[0003] For example, Non-Patent Document 1 describes a case where individual wind turbines are customized using a digital twin. [Prior art documents] [Non-patent literature]

[0004] [Non-Patent Document 1] Mitsui Global Strategic Studies Institute, Strategic Studies Institute Report, January 31, 2017, "Four Innovations to Watch in 2017," https: / / www.mitsui.com / mgssi / ja / report / detail / __icsFiles / afieldfile / 2017 / 02 / 09 / 170131tm.pdf Summary of the Invention [Problem to be solved by the invention]

[0005] However, conventional technologies, including Non-Patent Document 1, only consider the digital twin of a single object. Non-Patent Document 1 also mentions that there are attempts to use digital twins to realize urban planning and the like in addition to the wind turbine generators mentioned above, but it only mentions creating digital twins of individual objects.

[0006] Numerous objects exist in the real world, and they influence each other. This means that it is necessary to consider the mutual influence between digital twins. Taking road traffic as an example, roads are used by various objects, including public transportation such as buses, personal cars, and transportation vehicles. In other words, the digital twins of cars and buses influence each other. Naturally, each car can have its own digital twin, and it is conceivable that the digital twin of one car will influence the digital twin of a second car. In order to simultaneously optimize one or more indicators as a whole while taking into account the mutual influence of such diverse digital twins, it is necessary to define the indicators and consider the interactions between the digital twins. This type of control cannot be achieved without considering the interactions between digital twins.

[0007] The disclosed technology has been made in consideration of the above points, and aims to provide an optimization method, optimization device, and optimization program that optimizes specified indicators by taking into account the mutual influence between digital twins, using transportation as an example. [Means for solving the problem]

[0008] A first aspect of the present disclosure is an optimization method characterized in that a computer executes a process including: acquiring a first input including a departure time, departure point, and destination point of a first user; acquiring a second input including a departure time, departure point, and destination point of a second user; calculating a route to be presented to the first user and a route to be presented to the second user on a digital twin that optimizes predetermined indicators based on both the first input and the second input; presenting the route to be presented to the first user to the first user; and presenting the route to be presented to the second user to the second user.

[0009] A second aspect of the present disclosure is an optimization device that includes an acquisition unit that acquires a first input including a departure time, departure point, and destination point of a first user, and acquires a second input including a departure time, departure point, and destination point of a second user, and an optimization unit that calculates a route to be presented to the first user and a route to be presented to the second user on a digital twin, optimizing predetermined indicators based on both the first input and the second input, and presents the route to be presented to the first user to the first user and the route to be presented to the second user to the second user.

[0010] A third aspect of the present disclosure is an optimization program that causes a computer to execute a process of acquiring a first input including a departure time, departure point, and destination point of a first user, acquiring a second input including a departure time, departure point, and destination point of a second user, calculating a route to be presented to the first user and a route to be presented to the second user on a digital twin that optimizes predetermined indicators based on both the first input and the second input, and presenting the route to be presented to the first user to the first user and the route to be presented to the second user to the second user. [Effects of the Invention]

[0011] According to the disclosed technology, it is possible to provide an optimization method, an optimization device, and an optimization program that optimize a specified index by taking into account the mutual influence between digital twins. [Brief explanation of the drawings]

[0012] [Figure 1] FIG. 1 is a diagram illustrating an overview of an embodiment of the present disclosure. [Figure 2] FIG. 1 is a block diagram illustrating a schematic configuration example of an embodiment, which is an example of a functional configuration including an optimization device. [Figure 3] FIG. 1 is a schematic diagram showing an example of each function in the optimization of mobile MaaSDT. [Figure 4] FIG. 2 is a block diagram showing a hardware configuration of the optimization device. [Figure 5]10 is a flowchart showing the flow of optimization processing by the optimization device. [Figure 6] FIG. 10 is a diagram showing a flow of optimization example 1 in the case where a route is optimized using a congestion index value. [Figure 7] FIG. 10 is a diagram showing the flow of optimization example 2 when optimizing a route using index values ​​of congestion levels of transportation means and index values ​​of prices for each user. DETAILED DESCRIPTION OF THE INVENTION

[0013] An example of an embodiment of the disclosed technology will be described below with reference to the drawings. Note that the same or equivalent components and parts in each drawing are given the same reference numerals. Also, the dimensional proportions in the drawings are exaggerated for the convenience of explanation and may differ from the actual proportions.

[0014] First, an outline of an embodiment of the disclosed technology will be described. Fig. 1 is a diagram for explaining an outline of this embodiment.

[0015] The method according to this embodiment is assumed to be used in a city planning digital twin computing (DTC) platform that expands the value provided by linking multiple digital twins. DTC assumes digital twin computing that aims to expand value by linking multiple digital twins.

[0016] Specific applications of digital twin technology include controlling or recommending the movement of people and goods (including cargo and vehicles) in real space. For example, logistics or the movement of people can be predicted using each digital twin. The predicted results can be used to recommend or control the behavior of each digital twin so that specified indicators (load capacity, operating time, waiting time, congestion level, etc.) are met for all digital twins.

[0017] Furthermore, digital twins can be used not only from a local perspective in urban development, but also to envision a hierarchical structure of terminal DTCs. Figure 1 is an illustration of a terminal DTC based on a station or other location. The example in Figure 1 shows an example of a hierarchy with a national terminal DTC, a wide-area (metropolitan area) terminal DTC, and a local (one-mile) terminal DTC. In such a terminal DTC, the value provided by transportation modes such as airplanes, bullet trains, and trains is set as the index value of the digital twin. It is also expected that digital twins related to logistics, transportation, and tourism in each DTC will be optimized collectively or individually. Note that terminals can vary in size and scale, and digital twins can be constructed hierarchically based on area or scale, such as by country, region, prefecture, city or ward, and station or building.

[0018] Specific examples of digital twins include mobility as a service (MaaS) DT, tourism DT, logistics as a service (LaaS) DT, and retail DT. Each will be explained below. As will be described later, one or more indicators can be selected to be optimized depending on the target DT. Examples of indicators include profitability, customer satisfaction, operational efficiency, quality of service provision, environmental impact, social impact, and employee satisfaction. In addition, in cases where users are directly involved, such as BtoC, possible indicators include the degree of match with the user's timing and preferences, the degree of match with the actions the user is about to take, the user's subjective level of comfort, and the user's state of health.

[0019] Mobility MaaSDT targets the provision of value through the last mile, customer distribution, pricing, and subscriptions. The last mile is optimizing transportation methods for the last mile, for example, from the station to home. Customer distribution is dispersing congestion on trains, for example. Pricing is the use of pricing to disperse congestion. Subscription (hereafter abbreviated to "subscription") is the optimization of the allocation of usage at multiple locations by applying subscriptions, including the location of the destination. Indicators include average occupancy, congestion level, comfort, CO2 concentration, safety, the price paid by the end user, and room occupancy rate.

[0020] In tourism DT, targets include best price subscription, remote environment provision, and remote experience optimization. Best price subscription involves optimizing tourism for MaaS, which includes primary and secondary transportation, by considering the congestion level of transit points at transportation means and facilities, securing transit points, and pricing them. Primary transportation refers to transportation between stations, such as traveling from Shinagawa to Kyoto. Secondary transportation refers to transportation from a station to a destination, such as traveling within Kyoto after arriving there. Remote environment provision involves predicting and recreating the environmental conditions of a remote location. Remote experience optimization involves, for example, remote trips operated by remote transportation vehicles, recreating the experience of a remote location. Indicators include facility utilization rate, average occupancy rate, turnover rate, waiting time, congestion level, and environmental impacts such as CO2 concentration.

[0021] LaaSDT targets branch logistics optimization and the provision of value through cloud baggage. Branch logistics optimization is the optimization of logistics in which a user's luggage is transported separately from the user within their daily travel area. Cloud baggage is the optimization of item sharing, such as ensuring that necessary items are available at accommodations even when the user travels to a travel destination without luggage. Indicators include transport volume, loading rate, delivery time, price paid by the end user, environmental impact such as CO2 concentration, number of users, and the suitability of delivery contents for each user. Suitability refers, for example, to the degree of match with the end user's behavior and preferences, and the degree to which the item is necessary at the travel destination.

[0022] Retail DT targets the creation of a single terminal and the provision of value through warehouse-style retail. Single terminal creation is the optimization of logistics, for example, by consolidating products from stores in nearby stations in a store at one station, allowing customers to purchase products without having to go to the nearby station or store. Warehouse-style retail is the optimization of logistics that takes into account supply and demand, such as determining the order in which products should be moved according to user demand when stores provide products via mobile media. Indicators include the size of the target area, sales, coverage rate of retail stores, environmental impact such as CO2 concentration, delivery time, and product lineup.

[0023] (First embodiment) First, the key points of the first embodiment will be described. The first embodiment describes a technology that uses digital twins of at least multiple users to present routes to each user in a way that optimizes an index of the digital twin as a whole, including digital twins of multiple means of transportation. Here, congestion level is used as an index to be optimized.

[0024] Assume there are a first user and a second user, and a first means of transportation, a second means of transportation, and a third means of transportation. The behavior of the first user changes the congestion level of any of the first means of transportation, the third means of transportation, and the third means of transportation. In other words, it can be seen that the behavior of the first user affects the second user and the means of transportation. Therefore, the departure points, destination points, and desired arrival times (or departure times) of the first user and the second user are obtained, and candidate routes for each are calculated. Combinations of route candidates for the first user and route candidates for the second user are created, and the congestion level of each means of transportation is calculated for each combination. Once the combination that minimizes the overall congestion level is determined, the route candidates for the first user and the second user corresponding to that combination are presented to the first user and the second user, respectively.

[0025] Note that one index may be optimized, or multiple indexes may be optimized simultaneously. Examples of indexes include average occupancy rate and room occupancy rate that take profitability into consideration, congestion level, comfort, safety, and price that take customer satisfaction into consideration, and CO2 emissions and power consumption that take the environment into consideration. Optimization may also be performed using methods other than the optimization methods described below. Furthermore, the user may input data from the real world instead of using their own digital twin.

[0026] The optimization device according to the first embodiment of the present disclosure is configured in consideration of the above-described assumptions regarding DTC. In the following description, a case where the routes of multiple users are optimized using at least the degree of congestion as an index value of the digital twin will be described, taking the case where the above-described travel MaaSDT is optimized as an example.

[0027] Fig. 2 is a block diagram showing an example of a schematic configuration of an embodiment, including an example of a functional configuration including an optimization device. Fig. 2 shows an optimization device 10 that optimizes a digital twin system, an information processing device 20 that provides various data to the optimization device 10, and a device 30 that operates based on the digital twin system optimized by the optimization device 10.

[0028] The optimization device 10 is a device that optimizes a digital twin system that includes at least two digital twins, and as one aspect of optimizing a digital twin system, optimizes digital twins related to the respective routes of multiple users.

[0029] The information processing device 20 is a device that provides information to the optimization device 10, and in this embodiment, is a device that controls various means of transportation such as taxis, buses, trains, and shared bikes.

[0030] In this embodiment, the device 30 is a device operated by a user, and may be, for example, a smartphone, a PC, a tablet device, a wearable device, etc. The device 30 performs operations based on information output from the digital twin system optimized by the optimization device 10. The device 30 also includes installed terminals such as digital signage that present information to people, and includes everything that affects people, things, and the environment in the real world.

[0031] In order to optimize a route, the optimization device 10 has at least information on transportation means (such as transportation operation information and congestion information). Furthermore, when a user's index value is used in route optimization, the optimization device 10 has information on the user, such as user information (such as a schedule and a behavioral model). In the optimization device 10, each digital twin can be constructed based on data provided from, for example, the information processing device 20 and the equipment 30. When the optimization device 10 controls transportation means as a result of optimizing the digital twin system, the optimization device 10 outputs information for controlling the transportation means. Furthermore, the optimization device 10 acquires various data. The various data include transportation means information, user information, and facility information. The acquired transportation means information includes, for example, traffic data, vehicle allocation data (routes and allocated vehicles), operation information, and congestion information. The acquired user information is a behavioral model obtained from the user's personal schedule and a personal profile, such as the user's behavioral history and location information. The acquired facility information includes, for example, facility congestion status, inventory status, pricing, etc. To acquire user information and facility information, various sensors (not shown) that sense status can be used, for example. Examples of sensors include biometric sensors that acquire information on vital signs such as human body temperature, pulse, and blood pressure, and sensors that acquire information on devices and equipment used by humans. Other examples of sensors include air conditioners installed in building facilities, sensors installed in facilities such as EVs, sensors equipped in physical infrastructure such as robots and drones, and cameras or sensors installed in buildings.

[0032] The digital twin system may be constructed inside the optimization device 10, or may be constructed inside a device different from the optimization device 10.

[0033] Here, an example of digital twin optimization realized by the optimization device 10 will be described. FIG. 3 is a schematic diagram showing an example of each function in the optimization of mobility MaaSDT. The example shown in FIG. 3 has a user function, a MaaS optimization function, and a mobility control function, where the user function corresponds to the device 30, the MaaS optimization function corresponds to the optimization device 10, and the mobility control function corresponds to the information processing device 20. It is assumed that a user performs a search by entering a departure time, departure point, destination, etc., through a predetermined input from the device 30. At this time, the MaaS route selection function generates route candidates using traffic data, etc. The optimization device 10 links various digital twins to obtain location data and operation status of transportation means from the information processing device 20 and controls the transportation means for the information processing device 20 (linked MaaSDT). The optimization device 10 also presents routes and selects prices for the device 30, and arranges transportation means, etc. (travel support).

[0034] FIG. 4 is a block diagram showing the hardware configuration of the optimization device 10. As shown in FIG.

[0035] 4, the optimization device 10 includes a CPU (Central Processing Unit) 11, a ROM (Read Only Memory) 12, a RAM (Random Access Memory) 13, a storage 14, an input unit 15, a display unit 16, and a communication interface (I / F) 17. Each component is connected to each other via a bus 19 so as to be able to communicate with each other.

[0036] The CPU 11 is a central processing unit that executes various programs and controls each part. That is, the CPU 11 reads the programs from the ROM 12 or the storage 14 and executes the programs using the RAM 13 as a work area. The CPU 11 controls the above-mentioned components and performs various arithmetic processing in accordance with the programs stored in the ROM 12 or the storage 14. In this embodiment, the ROM 12 or the storage 14 stores an optimization program that optimizes the digital twin system. Furthermore, multiple digital twins may be constructed in the storage 14.

[0037] The ROM 12 stores various programs and various data. The RAM 13 temporarily stores programs or data as a working area. The storage 14 is configured with a storage device such as an HDD (Hard Disk Drive) or an SSD (Solid State Drive) and stores various programs including the operating system and various data.

[0038] The input unit 15 includes a pointing device such as a mouse and a keyboard, and is used to perform various inputs.

[0039] The display unit 16 is, for example, a liquid crystal display, and displays various information. The display unit 16 may also function as the input unit 15 by adopting a touch panel system.

[0040] The communication interface 17 is an interface for communicating with other devices. For this communication, for example, a wired communication standard such as Ethernet (registered trademark) or FDDI, or a wireless communication standard such as 4G, 5G, or Wi-Fi (registered trademark) is used. The above is the hardware configuration of the optimization device 10.

[0041] Next, we will explain each functional component of the optimization device 10. Each functional component is realized by the CPU 11 reading out an optimization program stored in the ROM 12 or storage 14, expanding it in the RAM 13, and executing it. As shown in Fig. 2, the optimization device 10 has, as its functional components, an acquisition unit 101, a candidate generation unit 102, a calculation unit 103, and an optimization unit 104.

[0042] The acquisition unit 101 acquires a first input from a first user and a second input from a second user. The user input contents of the first input and the second input include at least the departure time, departure point, and destination for each user. The first user or the second user may be one or more users, and may be multiple users. Furthermore, the first input and the second input may be acquired from any device 30 operated by the first user and the second user, or may be acquired automatically without user operation. Although the drawing shows one device 30, it goes without saying that there may be a device 30 for each user. Examples of the device 30 include mobile devices such as smartphones and tablets, and personal computers. Furthermore, the desired arrival time may be included instead of the departure time.

[0043] The candidate generation unit 102 generates candidate combinations of route candidates for the first user and route candidates for the second user based on the first input and the second input. The candidate combinations may be generated, for example, by extracting route candidates by available modes of transportation for each user based on the departure time, departure point, and destination, and by referencing information about the modes of transportation. The route candidates are generated by generating combinations of multiple possible modes of transportation along the way based on the combination of departure point and destination. Further variations in the combinations of modes of transportation can be generated by varying the departure time and arrival time. For example, these combinations of multiple modes of transportation may include "bus → taxi → bicycle," "train → walking," and "subway → shuttle bus." These modes of transportation may also represent modes of transportation for each layer of the hierarchical structure described above.

[0044] The calculation unit 103 calculates at least the congestion degree of each means of transportation for each time period as an index value for the candidate combination generated by the candidate generation unit 102. The means of transportation are means of transportation included in the route candidates of the candidate combination. The congestion degree is calculated, for example, using at least the time when a first user uses a means of transportation included in the travel route and the time when a second user uses a means of transportation included in the travel route. The time when a means of transportation is used is the time when the user plans to board a means of transportation included in the route candidate. The congestion degree may be calculated based on, for example, the current and past congestion status of the means of transportation at the corresponding time included in congestion information acquired in advance. The time when a user uses a mode of transportation included in a travel route is not set to a specific time, but rather the departure time is shifted within the constraints of each user's behavior, and multiple candidates are generated for each departure time.By generating combinations of each user's travel route, it is also possible to calculate the transition of congestion throughout the day and calculate an index value.

[0045] The optimization unit 104 calculates each of the routes that are optimal combinations among the candidate combinations as a route to be presented to a first user and a route to be presented to a second user based on the congestion index value estimated for each time period of the means of transportation for the candidate combination. Each of the routes calculated here is calculated on a digital twin as a route that optimizes congestion. The optimization unit 104 outputs the calculated route to be presented to the first user to any device 30 operated by the first user, and outputs the calculated route to be presented to the second user to any device 30 operated by the second user. Each of the routes that optimize congestion is an example of a route to be presented to a first user and a route to be presented to a second user that optimizes a desired index of the present disclosure.

[0046] Next, the operation of the optimization device 10 will be described.

[0047] 5 is a flowchart showing the flow of optimization processing by the optimization device 10. The CPU 11 reads out an optimization program from the ROM 12 or storage 14, loads it into the RAM 13, and executes it, thereby performing optimization processing.

[0048] In step S100, the CPU 11 acquires a first input of a first user and a second input of a second user as the acquisition unit 101. The first input and the second input may be acquired from any device 30 operated by the first user and the second user, or may be acquired automatically without user operation.

[0049] In step S102, the CPU 11 functions as the candidate generating unit 102 to generate candidate combinations of route candidates related to the first user and route candidates related to the second user based on the first input and second input acquired in step S100.

[0050] In step S104, the CPU 11, functioning as the calculation unit 103, calculates at least the congestion degree of each mode of transportation by time as an index value for the candidate combination generated in step S102. The mode of transportation is a mode of transportation included in the route candidates of the candidate combination.

[0051] In step S106, the CPU 11, functioning as the optimization unit 104, calculates each of the routes of the optimal combination from among the candidate combinations as a route to be presented to the first user and a route to be presented to the second user. Each of the routes is calculated based on the congestion index value estimated for each mode of transportation for the candidate combination for each time period.

[0052] In step S108, the CPU 11, as the optimization unit 104, outputs the calculated route to be presented to the first user to any device 30 operated by the first user, and outputs the calculated route to be presented to the second user to any device 30 operated by the second user.

[0053] Below, application example 1 and application example 2 of the mobile MaaSDT in the optimization device 10 will be described. In the application examples, it is assumed that the first user is user A, and the second users are users B and C. Application example 1 in Fig. 6 is a case where a route is optimized using a congestion index value. Application example 2 in Fig. 7 is a case where a route is optimized using a congestion index value for a means of transportation and a price index value for each user. The price is the price when using the means of transportation.

[0054] (Application example 1 of mobile MaaSDT) The application example 1 shown in FIG. 6 will be explained. [1] First, input (departure time, departure point, and destination) is acquired from users A, B, and C. [2] From the input, information on transportation modes (operation information) is referenced, and route candidates for user A (i = A1 to An), route candidates for user B (j = B1 to Bn), and route candidates for user C (k = C1 to Cn) are extracted, respectively, and a candidate combination T(Tijk) of route candidates is generated. Note that the number n is a unique number for each user. For route candidates (A1, B1, C1), the candidate combination T can be generated as candidate combination T111(Tijk). Here, each route candidate (Ai, Bj, Ck) may include transportation mode a, transportation mode b, and transportation mode c. It may also include other transportation modes d, e, and f, and the combination of these transportation modes may vary for each user. [3] Next, the hourly usage status of the transportation modes included in the route candidates is predicted, and a congestion index value is calculated. For transportation usage, information on transportation modes is used to perform congestion prediction, usage time prediction, waiting time prediction, etc., and these are combined to calculate a congestion index value. [4] Using the congestion index value calculated for each transportation mode time, ijk is changed independently for each combination of candidate combinations Tijk to calculate an overall index value for the candidate combination Tijk. [5] From the candidate combinations Tijk, the candidate combination T'ijk with the optimal congestion index value is derived. [6] The route of the candidate combination T'ijk is presented to users A, B, and C, respectively. For example, if the candidate combination T'ijk is (A7, B3, C5), route recommendations are made by presenting route A7 to user A, route B3 to user B, and route C5 to user C.

[0055] (Mobile MaaSDT application example 2) Application example 2 of FIG. 7 will be described. Differences between application example 2 and the case of application example 1 will be described. In application example 2, in [4], ijk of candidate combination Tijk is changed independently to calculate the overall congestion index value and the price index value for each of users A to C. [5] From the candidate combination Tijk, a candidate combination T'ijk is derived that optimizes the congestion index value and the overall price index value. Note that, although a case where the congestion index value and the price index value are combined has been described, the present invention is not limited to this. For example, travel time, operating costs, and energy consumption may be taken into consideration when determining the index value, and different index values ​​may be used for each user.

[0056] As described above, according to the optimization device 10 of the first embodiment of the present disclosure, it is possible to optimize a predetermined index while taking into consideration the mutual influence between digital twins.

[0057] Next, a method for mutual optimization when a plurality of index values ​​are combined will be described.

[0058] DTC urban development for offices and commercial buildings is composed of multiple digital twins. Each digital twin has one or more index values ​​that determine whether a specific state (x) should be prioritized or is the best state. For example, in the case of energy control optimization, the index values ​​can be power consumption, comfort, and CO2 concentration. In the case of transportation operation optimization, the index values ​​can be power consumption, waiting time, and congestion level.

[0059] For example, to optimize one digital twin in a city block, it is sufficient to predict the state of each digital twin independently and select the best state. However, in order for multiple digital twins to maintain their optimal state, it is necessary to consider the impact they have on each other and select the mutually optimal state. There are three methods for selecting the mutually optimal state:

[0060] Method 1 is a method of predicting the states of other digital twins from candidate future states that will occur in one digital twin, and selecting the candidate that maximizes the sum of the index values ​​for each candidate state. Method 2 is a method of selecting a candidate by changing the states of other digital twins so that, when the index value of one digital twin is maximized, the states and their index values ​​fall within predetermined constraints on the index value. Method 3 is a method of selecting a candidate by performing multi-agent reinforcement learning to calculate the state in which all states are optimal. Specific examples of methods 1 and 2 are described below.

[0061] In method 1, for example, if DTs a to e exist, the following steps [1] to [3] are performed.

[0062] [1] First, for DT(a), future state candidates and the index value of a for each state are extracted. Future state candidates are extracted, for example, from a model or rule of the state that occurs in conjunction with the date, time, location, and immediately preceding event. For state candidate x, an index value a(x) is calculated. For state candidate y, an index value a(y) is calculated. For state candidate z, an index value a(z) is calculated. [2] Given the state candidates x, y, z and their respective index values ​​for DT(a), calculate the other DT states (DT(b), DT(c), DT(d), and DT(e)) and the index values ​​for those states. For example, for DT(b), the index value b(x') is calculated for state candidate x', the index value b(y') for state candidate y', and the index value b(z') for state candidate z'. The index values ​​for DT(c), DT(d), and DT(e) can be calculated in a similar manner. [3] Then, the sums T(x) to T(z) of the index values ​​of each state occurring in states a(1) to a(n) are calculated as follows, and the state with the maximum sum of the index values ​​is selected. T(x)=a(x)×p+b(x')×q+c(x'')×γ+... T(y)=a(y)×p+b(y')×q+c(y'')×γ+... T(z)=a(z)×p+b(z')×q+c(z'')×γ+... At this time, parameters (p, q, γ, etc.) are assigned according to the importance of each DT. This concludes the explanation of Method 1.

[0063] In method 2, for example, if DTs a to e exist, the following steps [1] to [7] are performed.

[0064] [1] First, calculate the state s where the index value a(s) in DT(a) is maximum. [2] Calculate the index values ​​of other DTs when DT(a) is in state s. The index values ​​of each DT when DT(a) is in state s are calculated as {a(s), b(s'), c(s''), ...}. [3] For each index value, check whether the value falls within the predefined constraints. At this time, record the DT that falls within the constraints in state s. [4] If a DT that does not satisfy the constraints occurs, reinforcement learning is performed between the DT and DT(a) to derive a state (t) that allows both DT(a) and the DT. [5] Set the state of DT(a) to (t) and perform [2] again. [6] Here, [2] to [5] are repeated, and the repetition stops when a state (u) is calculated in which all DTs fall within the constraints. However, if the repetition does not end after repeating [2] to [5] n times, the repetition stops. [7] Select the state (u) at the time of stopping. If all DTs do not fall within the constraints, select the state (u) where the DTs with the highest importance fall within the constraints based on the predetermined importance of each DT. This concludes the explanation of Method 2.

[0065] The constraints on the index values ​​set in the method described here also vary depending on the time of optimization. The above explanation is an example of verifying whether the optimization times for different index values ​​are aligned and whether each falls within the range of the constraints. However, it is also possible that the optimization times are not aligned for all index values. For example, control to ensure comfort temporarily may be performed multiple times throughout the day, and even if this results in a momentary loss of energy savings, energy savings can still be optimized for a day or a month in total. This makes it possible to apply the idea of ​​optimizing each index value when viewed as a time-integrated value.

[0066] In addition, the following considerations can be made regarding the prediction of future states performed by Methods 1 and 2. Specifically, when predicting future index values ​​at a certain time, a finite number of explanatory variables that contribute to changes in the index value are selected, and it is typically assumed that the number and type of explanatory variables will remain constant until the future time being predicted. However, the further the future time being predicted is from the present time, the more likely it is that new explanatory variables will contribute due to disturbances along the way, or the weights of related explanatory variables will change, making it impossible to ensure the accuracy of the future prediction using only the explanatory variables considered at the time of prediction. To address such cases, explanatory variables that contribute to inaccurate predictions can be extracted from factor analysis of other cases, and this know-how can be stored in a database. Then, when the influence of different explanatory variables occurs due to disturbances or changes in weights, the information can be immediately referenced, allowing the prediction model to be reconstructed quickly and the prediction to be corrected. This can also contribute to improving the accuracy of overall optimization.

[0067] Hereinafter, an embodiment will be described in which different index values ​​are used depending on the application scenario of the DTC.

[0068] (Second embodiment) The second embodiment is an embodiment in which DTC is utilized for MaaS subscription. In the second embodiment, the index value of the city block and the index value of the means of transportation are used as the index values ​​of the digital twin. Note that from the second embodiment onwards, each process can be performed with the same configuration and operation as the first embodiment. Note that a MaaS subscription is a contract for using stores, offices, etc. located in a specified area in real space and receiving discounts, including the usage fee for the store or office and the cost of the travel route to the target area. Generally, it is a flat rate, but it may also be a pay-as-you-go system.

[0069] As an example of the second embodiment, a scenario in which a user searches for a work location in a satellite office will be described. First, based on input from one or more users and user information, block a and block b are extracted as candidate satellite offices suitable for the user. The user information includes previously acquired behavior predictions, preference predictions, and satellite office usage trends. For the extracted block a and block b, an index value for block a and an index value for block b are calculated. The index value is also calculated by taking into account the means of transportation included in the route candidates of other users. For example, if users are A, B, C, etc., the route candidates for each user are (A1 to An, B1 to Bn, C1 to Cn). The index value for the means of transportation may be calculated by calculating the congestion level, travel time, energy consumption, etc., or a combination of these. Then, an overall index value obtained by combining these index values ​​is optimized. This enables overall optimization of satellite office usage in remote work, etc.

[0070] Optimization of multiple indicators may be used to optimize multiple DTs. For example, simultaneous optimization of the MaaSDT and LaaSDT is conceivable. In this case, at least one of average occupancy rate, comfort, and congestion level is selected as the indicator for the MaaSDT. Furthermore, at least one of transportation volume, load factor, delivery period, price, and CO2 is selected as the indicator for the LaaSDT. Then, optimization of the selected indicators as a whole is sufficient. This optimization enables smooth logistics and customer delivery while minimizing total energy consumption. Simultaneous optimization of the end user's digital twin and the MaaSDT is also conceivable. For example, one indicator may be selected from at least one of the following: the degree of match between the end user's digital twin and their behavior; and the degree of match between the end user's digital twin and their requirements, which integrates their wants, needs, and preferences. Another indicator may be selected from at least one of the MaaSDT's average occupancy rate, comfort, and congestion level. Optimization of these selected indicators as a whole enables efficient operation of transportation without interfering with end users' behavior.

[0071] (Third embodiment) The third embodiment is an embodiment in which DTC is utilized for tourism subscriptions. A tourism subscription is a contract made in a predetermined unit that includes transportation fees, including primary and secondary transportation, used when sightseeing in a predetermined area in real space, and facility fees, such as tourist facilities and hotels. This is generally a flat rate, but a pay-as-you-go system is also possible. The predetermined unit may be a predetermined area in real space or a facility with a common theme. In the third embodiment, the index value of the destination and the index value of the transportation are used as the index value of the digital twin. As an example of the third embodiment, a case in which a user's travel itinerary pattern is provided will be described. First, destination candidates for the user are extracted from inputs from multiple users (e.g., user M and user N) and user information. The user information can include the past behavioral history of other users' travels. Next, for itinerary patterns including destination candidates, candidate combinations of itinerary patterns for each user are generated. The itinerary pattern includes destination candidates A to N and transportation modes that pass through the destination candidates. Therefore, an index value for each destination candidate and an index value for each mode of transportation that passes through the destination candidate are calculated. For modes of transportation, candidate combinations can be generated and index values ​​for modes of transportation included in the candidate combinations can be calculated (the types of index values ​​are the same as in the second embodiment). Then, an overall index value obtained by combining these index values ​​is optimized. In this way, itinerary patterns are constructed based on the departure points and behavior predictions of multiple users, and an itinerary pattern that includes destinations A to N included in the itinerary pattern and allows efficient travel along routes that are not congested is provided to the user. Furthermore, by using price as the index value for the destination or mode of transportation, it is possible to provide low-cost itinerary patterns.

[0072] (Fourth embodiment) The fourth embodiment is an embodiment in which DTC is utilized for baggage-free trips. In the fourth embodiment, the index value of an item is used as the index value of a digital twin. As an example of the third embodiment, a case will be described in which it is assumed that a user will purchase necessary items at their travel destination. Similar to the third embodiment, candidate user destinations are extracted and itinerary patterns are generated. Here, combinations of itinerary patterns for multiple users and required item lists associated with the itinerary patterns are generated from the itinerary patterns. An item index value is calculated for each item included in the combination of required item lists. Item demand forecasts, etc. may be used to calculate the item index value. Then, the state in which the sum of the overall index values ​​of each item is highest is extracted, and services are provided to the user in that state.

[0073] Furthermore, DTC is not limited to the above embodiments and can be used in various scenarios by linking digital twins together, such as for the creation of a single terminal for retail digital terminals that aggregate products from nearby stores into one store, and for optimizing remote experiences.

[0074] In the above embodiments, the optimization process executed by the CPU after reading the software (program) may be executed by various processors other than the CPU. Examples of such processors include programmable logic devices (PLDs) such as field-programmable gate arrays (FPGAs), whose circuit configuration can be changed after fabrication, and application-specific integrated circuits (ASICs), which are dedicated electrical circuits that are processors with circuit configurations specifically designed to execute specific processes. The optimization process may be executed by one of these processors, or by a combination of two or more processors of the same or different types (e.g., multiple FPGAs, or a combination of a CPU and an FPGA). The hardware structure of these processors is, more specifically, an electrical circuit that combines circuit elements such as semiconductor devices.

[0075] In addition, in each of the above embodiments, the optimization program is described as being pre-stored (installed) in the storage 14, but this is not limiting. The program may be provided in a form stored in a non-transitory storage medium such as a CD-ROM (Compact Disk Read Only Memory), a DVD-ROM (Digital Versatile Disk Read Only Memory), or a USB (Universal Serial Bus) memory. The program may also be downloaded from an external device via a network.

[0076] The following additional notes are provided regarding the above-described embodiments.

[0077] (Additional note 1) Memory and at least one processor coupled to said memory; Including, The processor: obtaining a first input including a departure time, a departure point, and a destination point for a first user; obtaining a second input including a departure time, a departure point, and a destination point for a second user; Calculating, on the digital twin, a route to be presented to the first user and a route to be presented to the second user that optimize a predetermined index based on both the first input and the second input; Presenting a route to be presented to the first user to the first user; Presenting a route to be presented to the second user to the second user The optimization device is configured as follows.

[0078] (Additional note 2) A non-transitory storage medium storing a program executable by a computer to perform an optimization process, obtaining a first input including a departure time, a departure point, and a destination point for a first user; obtaining a second input including a departure time, a departure point, and a destination point for a second user; Calculating, on the digital twin, a route to be presented to the first user and a route to be presented to the second user that optimize a predetermined index based on both the first input and the second input; Presenting a route to be presented to the first user to the first user; Presenting a route to be presented to the second user to the second user Non-transitory storage medium. [Explanation of symbols]

[0079] 10 Optimization Device 20 Information processing equipment 30 equipment 101 Acquisition Department 102 Candidate generation section 103 Calculation Unit 104 Optimization Department

Claims

1. The entire digital twin system is made up of multiple digital twins with different controls. The plurality of digital twins are configured by a combination of a digital twin related to user movement and a digital twin related to logistics, As the plurality of predetermined indices of the interaction with the user, at least one of a degree of conformance with the timing and preferences of the user, a degree of conformance with the action the user is about to take, a subjective degree of comfort of the user, and a degree of health condition of the user is used; Based on both the location information of a first user and the location information of a second user, calculate a route to be presented to the first user and a route to be presented to the second user that optimizes the plurality of predetermined indicators that may differ for one or the plurality of digital twins as a whole in the entire digital twin system; Presenting a route to be presented to the first user to the first user; Presenting a route to be presented to the second user to the second user A method for optimizing a process including the steps of:

2. The calculation is 2. The optimization method according to claim 1, wherein the index value of a means of transportation is the safety of the means of transportation calculated using at least route candidates for the first user, route candidates for the second user, location information of the first user, and location information of the second user.

3. the route candidates for the first user and the route candidates for the second user include a plurality of transportation means for each predetermined layer; 3. The optimization method of claim 2, wherein the calculation uses an estimated safety for each of the plurality of transportation modes.

4. The index values ​​to be optimized include at least an index value of the transportation means and an index value of the user, The calculation is 4. The optimization method according to claim 2 or 3, wherein the optimization is performed so that all of the index values ​​targeted for optimization are optimized, or so that at least one of the index values ​​targeted for optimization is optimized and satisfies a constraint condition that is preset for the index value.

5. The calculation is The optimization method according to claim 4 , wherein different index values ​​are used for the first user and the second user as the index values ​​to be optimized.

6. The entire digital twin system is made up of multiple digital twins with different controls. The plurality of digital twins are configured by a combination of a digital twin related to user movement and a digital twin related to logistics, As the plurality of predetermined indices of the interaction with the user, at least one of a degree of conformance with the timing and preferences of the user, a degree of conformance with the action the user is about to take, a subjective degree of comfort of the user, and a degree of health condition of the user is used; Based on both the location information of a first user and the location information of a second user, calculate a route to be presented to the first user and a route to be presented to the second user that optimizes the plurality of predetermined indicators that may differ for one or the plurality of digital twins as a whole in the entire digital twin system; Presenting a route to be presented to the first user to the first user; an optimization unit that presents a route to be presented to the second user to the second user; An optimization device including:

7. The entire digital twin system is made up of multiple digital twins with different controls. The plurality of digital twins are configured by a combination of a digital twin related to user movement and a digital twin related to logistics, As the plurality of predetermined indices of the interaction with the user, at least one of a degree of conformance with the timing and preferences of the user, a degree of conformance with the action the user is about to take, a subjective degree of comfort of the user, and a degree of health condition of the user is used; Based on both the location information of a first user and the location information of a second user, calculate a route to be presented to the first user and a route to be presented to the second user that optimizes the plurality of predetermined indicators that may differ for one or the plurality of digital twins as a whole in the entire digital twin system; Presenting a route to be presented to the first user to the first user; Presenting a route to be presented to the second user to the second user An optimization program that causes a computer to perform the processing.

Images