Plan creation support device, plan creation support system, and plan creation support method

Abstract

To support quick creation of a new business plan based on a changed evaluation index even when the evaluation index in a business plan is changed.SOLUTION: A plan creation support device 50 is configured to: store a first model for identifying values of constituent elements of a business plan for optimizing a value of a first evaluation index of the business plan; identify a causal relation between a value of a second evaluation index of the business plan and the values of the constituent elements on the basis of actual result information of the constituent elements and the value of the second evaluation index of the business plan; calculate, on the basis of the causal relation, changed values from values of the constituent elements as calculated according to the first model, the changed values satisfying a predetermined condition concerning the second evaluation index; and identify new values of the constituent elements for optimizing the values of the first evaluation index and the second evaluation index of the business plan on the basis of a second model for calculating values of the constituent elements with which deviations from the calculated changed values of the constituent elements are minimized, and the first model.SELECTED DRAWING: Figure 2

Description

[Technical field]

[0001] The present invention relates to a planning support device, a planning support system, and a planning support method. [Background technology]

[0002] Each company draws up various business plans to carry out its business. However, the more complex the business is, the more time it takes to draw up the business plans.

[0003] Moreover, in recent years, social demands have led to an increased demand for formulating operational plans from a variety of perspectives. For example, railway operators are now being asked to formulate operational plans (i.e., train timetables) based on various indicators from the perspectives of reducing environmental impact (e.g., reducing greenhouse gas emissions) and reducing energy consumption (e.g., reducing electricity consumption), in addition to reducing train congestion.

[0004] Regarding technology for creating business plans from a specified perspective, for example, Patent Document 1 discloses a timetable creation system that stores evaluation index improvement degree information that records the relationship between changes to each component of a bus schedule and the change in quality from each perspective of the bus schedule due to the changes to the component, selects the component of the bus schedule that will most improve the quality of the bus schedule based on the evaluation index improvement degree information, selects from among multiple timetable creation models capable of creating a bus schedule, the timetable creation model that will most improve the quality of the entire bus schedule by changing the selected component, and creates a new bus schedule based on the selected timetable creation model. [Prior art documents] [Patent documents]

[0005] [Patent Document 1] Patent Publication No. 2021-088205 Summary of the Invention [Problem to be solved by the invention]

[0006] However, in Patent Document 1, when an evaluation index in a business plan (train schedule) is changed, the train schedule needs to be reconstructed. However, when the business plan is complicated as described above, it takes a lot of time to reconstruct the train schedule, which increases the workload of the company.

[0007] The present invention has been made in consideration of such problems, and aims to provide a planning support device, a planning support system, and a planning support method that are capable of supporting the rapid creation of a new plan based on a changed evaluation index, even when an evaluation index in a plan is changed. [Means for solving the problem]

[0008] One aspect of the present invention for solving the above-mentioned problems is a plan creation support device comprising: a storage device that stores a first model that identifies values ​​of each component constituting a business plan so as to optimize the value of a first evaluation index of the business plan, the value being calculated based on each of the components; and a control device that executes a causal relationship search process that identifies, based on a predetermined algorithm, a causal relationship between the value of a second evaluation index of the business plan calculated based on each of the components and the value of each of the components; a component change recommendation process that calculates, based on the identified causal relationship, a change value of each of the components from a value calculated by the first model that satisfies a predetermined condition related to the second evaluation index; a second model that calculates values ​​of each of the components so as to minimize deviation from the calculated change value of each component; and a plan creation process that identifies new values ​​of each component that optimize the values ​​of the first evaluation index and the second evaluation index of the business plan, based on the first model.

[0009] Another aspect of the present invention for solving the above-mentioned problems is a plan creation support system including a plan creation support device that supports the creation of a business plan and a predetermined facility that executes the business plan, in which the plan creation support device is provided with a storage device that stores a first model that identifies values ​​of each component that optimize the value of a first evaluation index of the business plan, which is calculated based on each component, and a control device that executes the following processes: a causal relationship search process that identifies a causal relationship between the value of a second evaluation index of the business plan, which is calculated based on each component, and the value of each component, based on a predetermined algorithm; a component change recommendation process that calculates a change value of each component from a value calculated by the first model that satisfies a predetermined condition related to the second evaluation index, based on the identified causal relationship; a second model that calculates a value of each component that minimizes deviation from the calculated change value of each component; a plan creation process that identifies new values ​​of each component that optimize the values ​​of the first evaluation index and the second evaluation index of the business plan, based on the first model; and a process of generating control signals for executing a new business plan composed of the identified values ​​of each component, and controlling the facility based on the generated business signal. Effect of the Invention

[0010] According to the present invention, even when an evaluation index in a plan is changed, it is possible to assist in quickly creating a new plan based on the changed evaluation index. Configurations and effects other than those described above will become apparent from the following description of the embodiments. [Brief description of the drawings]

[0011] [Figure 1] 1 is a diagram illustrating an example of a configuration of a plan creation support system according to an embodiment of the present invention. [Diagram 2] 2 is a diagram illustrating an example of hardware included in the planning support device and functions of the planning support device 50. FIG. [Diagram 3] FIG. 2 is a diagram illustrating an example of target plan information. [Figure 4] FIG. 11 is a diagram illustrating an example of component information. [Diagram 5] FIG. 11 is a diagram illustrating an example of evaluation index information. [Figure 6] FIG. 13 is a diagram illustrating an example of deviation reduction model information. [Figure 7] FIG. 11 is a flow diagram illustrating an example of a plan creation support process. [Figure 8] FIG. 11 is a diagram showing an example of component change recommendation information. [Figure 9] FIG. 13 is a diagram showing an example of a plan creation result display screen. [Figure 10] FIG. 11 is a flow diagram illustrating details of an evaluation index causal search process. [Figure 11] FIG. 13 is a diagram showing an example of a causal structure setting screen. [Figure 12] FIG. 13 is a diagram showing an example of diagram causal structure information to be created. [Figure 13] FIG. 13 is a diagram illustrating an example of generated evaluation index causal relationship information. [Figure 14] FIG. 11 is a flow diagram illustrating details of a causal relationship evaluation correction process. [Figure 15] FIG. 13 is a diagram illustrating an example of a causal relationship correction screen. [Figure 16] FIG. 11 is a flow diagram illustrating details of a component change recommendation process. [Figure 17] FIG. 13 is a diagram showing an example of a target value setting screen. [Figure 18] FIG. 13 is a diagram showing an example of a component change recommendation display screen. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0012] An embodiment of the present invention will be described with reference to the drawings. 1 is a diagram showing an example of the configuration of a planning support system 1 according to this embodiment. The planning support system 1 is an information processing system that supports the creation of business plans created by operators. In this embodiment, the planning support system 1 is a transportation planning system that supports the creation of train timetables created by railway operators (hereinafter, referred to as users).

[0013] That is, the planning support system 1 includes a simulator 30, an external data acquisition system 40, one or more facilities 10, a performance acquisition system 20, a control system 60, and a planning support device 50. The simulator 30, the external data acquisition system 40, the facilities 10, the performance acquisition system 20, the control system 60, and the planning support device 50 are communicatively connected to each other via a wired or wireless communication network 5 such as the Internet, a LAN (Local Area Network), a WAN (Wide Area Network), a VPN (Virtual Private Network), or a dedicated line.

[0014] The equipment 10 is equipment that is the subject of the plan. The equipment 10 is equipment necessary for railway operation, such as trains, tracks, signals, switches, and railroad crossings. The equipment 10 transmits operation or control history information (for example, information on the running history of trains) to the performance acquisition system 20.

[0015] The performance acquisition system 20 transmits operation history information (operation performance information 200 ) received from each facility 10 to the plan creation support device 50 .

[0016] The simulator 30 is a simulator that simulates virtual operation (motion) of each facility 10, and transmits simulation result information 300 (information corresponding to the operation performance information 200) that is the result of executing the simulation to the plan creation support device 50.

[0017] The external data acquisition system 40 stores external data information, which is information on factors that have affected the operation (performance) of the facility 10, such as past weather. The external data acquisition system 40 transmits the external data information to the plan creation support device 50.

[0018] The plan creation support device 50 is an information processing device that supports the creation of plans for the operation (movement) of each facility 10. The plan creation support device 50 transmits information (plan information) indicating the created plan to the control system 60 and the simulator 30. The control system 60 and the simulator 30 each run (operate) each facility 10 (virtual facility) based on the received plan information.

[0019] The plan creation support device 50 also acquires the operation performance information 200, the simulation result information 300, and the external data information 400 (hereinafter, these pieces of information are collectively referred to as performance information). Then, the plan creation support device 50 creates a current execution plan (i.e., a current train schedule) for each facility 10 based on the performance information, etc., and stores it in target plan information 100, which will be described later. This operation plan (hereinafter, referred to as a target plan) is created based on a predetermined objective function (first model).

[0020] This objective function is a function for calculating the value of a predetermined evaluation index (e.g., the congestion rate of each train) with the components constituting the execution plan (e.g., the stations at which each train stops, and the departure and arrival times of each stop) as explanatory variables. The plan creation support device 50 creates a target plan that optimizes the evaluation index by identifying the values ​​of each component that will result in an optimal (minimum) value for this objective function. Note that each evaluation index may be calculated using not only the values ​​of the components, but also the values ​​of other evaluation indexes.

[0021] Here, the railway operator may consider a new evaluation index (hereinafter referred to as a second evaluation index) in addition to the above evaluation index (hereinafter referred to as a first evaluation index) that was considered in the target plan. Even in such a case, the plan creation support device 50 of this embodiment does not perform cumbersome processing such as reconstructing the first model, but creates a new model (deviation reduction model) that combines the currently stored first model with a newly created second model (details will be described later) related to the second evaluation index, thereby creating an action plan (hereinafter referred to as an improved plan) that changes the target plan. Note that the first evaluation index and the second evaluation index may each be a set of multiple evaluation indexes.

[0022] FIG. 2 is a diagram illustrating an example of hardware included in the planning support device 50 and functions that the planning support device 50 has.

[0023] First, the plan creation support device 50 includes an arithmetic device 51 (control device) such as a central processing unit (CPU), a digital signal processor (DSP), a graphics processing unit (GPU), a field-programmable gate array (FPGA), or an application specific integrated circuit (ASIC), a storage device 52 such as a random access memory (RAM), a read only memory (ROM), a hard disk drive (HDD), or a solid state drive (SSD), a communication device 53 including a network interface card (NIC), a wireless communication module, a universal serial interface (USB) module, or a serial communication module, an input device 54 such as a keyboard, a mouse, or a touch panel, and an output device 55 such as a liquid crystal monitor or a liquid crystal display (LCD). The simulator 30, the external data acquisition system 40, the equipment 10, the performance acquisition system 20, and the control system 60 also include similar hardware.

[0024] The plan creation support device 50 stores the target plan information 100, the operation performance information 200, the simulation result information 300, and the external data information 400 described above.

[0025] The plan creation support device 50 also stores component information 500 that defines the contents of the components, evaluation index information 600 that stores the contents and calculated values ​​of evaluation indexes, diagram causal structure information 700 that defines initial values ​​of the causal relationships between each evaluation index and each component and other evaluation indexes, evaluation index target value information 800 that stores the targets of the evaluation index to be achieved in the improvement plan, mathematical planning model information 900 that stores each model, deviation reduction model information 1000 that stores each objective function related to the improvement plan (hereinafter referred to as deviation reduction model), evaluation index causal relationship information 1100 that saves calculated values ​​of the causal relationships between each evaluation index and each component and other evaluation indexes, component change recommendation information 1200 that stores the changes to each component when changing from a target plan to an improvement plan, and improvement plan information 1300 that stores each component of the improvement plan.

[0026] (Target plan information) 3 is a diagram showing an example of the target plan information 100. The target plan information 100 has data on each train ID 101, each train process ID 102, each train type 103, each train arrival station 104, each train departure station 105, each train arrival time 106 at each arrival station, and each train departure time 107 at each departure station. Note that the improvement plan information 1300 also has a similar data configuration.

[0027] (Component Information) 4 is a diagram showing an example of the component information 500. The component information 500 has data on an ID 501 of each component, a type 502 of each component, and content 503 (entity) of each component.

[0028] In this embodiment, the components include, for example, the number of trains (which is essentially the number of trains running on each section for each train), arrival time (which is essentially the arrival time of each train at each station), departure time (which is essentially the departure time of each train from each station), stop status (which is essentially whether each train passes through or stops at each station), and running intervals (which is essentially the running intervals of each train in each section).

[0029] (Evaluation index information) FIG. 5 is a diagram showing an example of evaluation index information 600. The evaluation index information 600 has data of ID 601 of each evaluation index, type 602 of each evaluation index, and content 603 (entity) of each evaluation index. In the content 603 of the evaluation index, for example, an equation related to components for calculating the value of the evaluation index is set. In this embodiment, the evaluation indexes include, for example, an average congestion rate (the actual value is the congestion rate of each section of each train), power consumption (the power consumption of each section of each train), a total average congestion rate (the actual value is the average value of the congestion rates of all sections of all trains), and a total power consumption (the actual value is the total value of the power consumption of all sections of all trains). In addition, the evaluation index information 600 may store an index value of each evaluation index calculated by the plan creation support device 50.

[0030] (Divergence reduction model information) 6 is a diagram showing an example of deviation reduced model information 1000. The deviation reduced model information 1000 has data of an ID 1001 assigned to each model, a list 1002 of each component, and a deviation reduced model 1003 having those components as explanatory variables. In the deviation reduced model 1003, for example, information of a second model is set.

[0031] As shown in FIG. 2, the plan creation support device 50 also stores the following functional units (programs): a plan evaluation unit 61, a causal relationship search unit 62, a causal relationship evaluation correction unit 63, a component change recommendation unit 64, a model extension unit 65, and a plan creation unit 66.

[0032] The plan evaluation unit 61 calculates the value of a first evaluation index in the target plan (target plan) and stores it in the evaluation index information 600.

[0033] In this embodiment, the objective function of the first model is an objective function F(x) for the first evaluation index, with each component x (x∈S: feasible set) as an explanatory variable. The objective function F(x) is a function whose value is minimized when the first evaluation index is optimal.

[0034] The causal relationship searching unit 62 identifies, based on a predetermined algorithm, a causal relationship between the value of the second evaluation index of the work plan calculated based on each component and the value of each component.

[0035] In this embodiment, the causal relationship searching unit 62 identifies the causal relationship based on the performance information and the value of the second evaluation index calculated from the performance information. Specifically, the causal relationship searching unit 62 identifies the causal relationship based on a model based on a predetermined mathematical formula (structural equation). The details of the structural equation will be described later.

[0036] The causal relationship evaluation correction unit 63 corrects the causal relationship identified by the causal relationship search unit 62. Specifically, the causal relationship evaluation correction unit 63 calculates the accuracy of the causal relationship identified by the causal relationship search unit 62 based on performance information, checks whether the accuracy of the calculated causal relationship is equal to or lower than a predetermined threshold, and if the accuracy of the calculated causal relationship is equal to or lower than the predetermined threshold, receives an input for correction of the causal relationship from the user.

[0037] The component change recommendation unit 64 calculates a change value from the value calculated by the first model for each component of the business plan that satisfies a specified condition regarding the second evaluation index, based on the causal relationship identified by the causal relationship search unit 62 (or corrected by the causal relationship evaluation correction unit 63).

[0038] In this embodiment, the above-mentioned predetermined conditions are the target value of the second evaluation index to be achieved and constraints on the values ​​of each component element.

[0039] The model extension unit 65 creates a second model that calculates the value (amount of change) of each component that minimizes the deviation from the change value of each component calculated by the component change recommendation unit 64. In other words, the second model is a model for calculating the optimal amount of change to the value of each component of the current train schedule when creating an improvement plan taking into account the second evaluation index.

[0040] In this embodiment, the second model is an objective function G(x) for the second evaluation index, with each component x (x∈S: feasible set) as an explanatory variable. The model extension unit 65 stores information specifying the configurations of the objective function F(x) of the first model and the objective function G(x) of the second model in the deviation reduction model information 1000.

[0041] The plan creation unit 66 specifies values ​​of each component element that optimizes the values ​​of the first evaluation index and the second evaluation index of the improvement plan, based on the second model created by the model extension unit 65 and the first model. The plan creation unit 66 stores information on the improvement plan configured by the specified values ​​of each component element in the improvement plan information 1300.

[0042] In this embodiment, the plan creation unit 66 specifies the value of each component based on a model (hereinafter, referred to as a deviation evaluation model) that is a linear combination of the objective function F(x) of the first model and the objective function G(x) of the second model. The details of the deviation evaluation model will be described later.

[0043] Each of the functional units of the plan creation support device 50 described above is realized by the calculation device 51 reading and executing each program stored in the storage device 52. Each program can be recorded on a recording medium and distributed, for example. Note that the plan creation support device 50 may be realized, in whole or in part, by using virtual information processing resources provided by using a virtualization technique, a process space separation technique, or the like, such as a virtual server provided by a cloud system. Also, all or in part of the functions provided by the plan creation support device 50 may be realized by a service provided by a cloud system via an API (Application Programming Interface), for example. Next, the processing performed in the planning support system 1 will be described.

[0044] <Plan creation support processing> 7 is a flow diagram illustrating an example of a process (planning support process) for supporting a railway operator in creating a train schedule. The planning support process is started, for example, when a predetermined input is made to the planning support device 50 by a user.

[0045] The causal relationship searching unit 62 selects a plan (target plan) that is to be updated (s11). For example, the causal relationship searching unit 62 acquires target plan information 100 based on an input from a user.

[0046] The causal relationship searching unit 62 acquires data for calculating the value of the first evaluation index of the target plan (s12). For example, the causal relationship searching unit 62 acquires the component information 500, the evaluation index information 600, and the like.

[0047] The causal relationship searching unit 62 calculates the value of the first evaluation index of the target plan (s13). For example, the causal relationship searching unit 62 calculates the value of each evaluation index of the target plan by inputting the content of the target plan information 100 acquired in s11 into the content 503 of the component for a calculation formula of the first evaluation index specified by the content of the record related to the first evaluation index of the target plan in the evaluation index information 600 and the component information 500.

[0048] The causal relationship searching unit 62 executes evaluation index causal relationship searching process s14 to generate evaluation index causal relationship information 1100 indicating the strength of the causal relationship between each evaluation index and each component element in the evaluation index. The evaluation index causal relationship searching process s14 will be described later in detail.

[0049] The causal relationship evaluation correction unit 63 executes a causal relationship evaluation correction process s15 for correcting the evaluation index causal relationship information 1100 generated in the evaluation index causal search process s14. The causal relationship evaluation correction process s15 will be described in detail later.

[0050] The component change recommendation unit 64 executes a component change recommendation process s16 for generating component change recommendation information 1200, which is information on changes (values ​​from the current state) of components in a train schedule that is optimal when the second evaluation index is taken into consideration, based on the evaluation index causal relationship information 1100 generated (corrected) in the causal relationship evaluation correction process s15. The component change recommendation process s15 will be described in detail later.

[0051] (Component change recommendation information) 8 is a diagram showing an example of component change recommendation information 1200. The component change recommendation information 1200 has data on a change recommendation ID 1201, which is an ID given to a change to each component, an ID 1202 of each component, a content 1203 of each change, and an amount of change 1204 of each component. In the example shown in the figure, in order to turn the current target plan into a train schedule that optimizes the first evaluation index and the second evaluation index, the optimal amount of change is to delay the "arrival time of train A at station X" in the current target plan (calculated by the first model) by "5 seconds."

[0052] Next, as shown in FIG. 7, the plan creation section 66 creates an improvement plan by modifying the target plan selected in s11 based on the component change recommendation information 1200 generated in the component change recommendation process s16 (s17).

[0053] That is, first, the plan creation unit 66 identifies configuration information of a deviation reduced model in which the values ​​of each component indicated by the component change recommendation information 1200 can be changed, based on the deviation reduced model information 1000. Specifically, the plan creation unit 66 acquires the contents of the deviation reduced model information 1000 of a record in which each component set in the component change recommendation information 1200 is set in the list 1002, from the deviation reduced model information 1000.

[0054] In addition, the plan creation unit 66 acquires a model (first model) used in creating the target plan. For example, the plan creation unit 66 acquires the first model (objective function F(x)) used in creating the target plan from a predetermined database.

[0055] Then, the plan creation unit 66 creates a second model for creating an improvement plan. Specifically, the plan creation unit 66 creates, for each record of the component change recommendation information 1200, a term whose value decreases as it approaches the value of the component indicated by the content 120 of the component and the change amount 1204 of the component, based on the configuration of the deviation reduction model. For example, the plan creation unit 66 creates an objective function M×P(x) by multiplying the sum of the created terms by a predetermined coefficient M (M is a positive value sufficiently larger than a value that the objective function F(x) can generally take).

[0056] Then, the plan creation unit 66 creates a deviation-reduced model by combining the first model and the second model. For example, the plan creation unit 66 creates an objective function G(x)={objective function F(x)+objective function M×P(x)} of the deviation-reduced model.

[0057] The plan creation unit 66 creates an improvement plan by solving the created deviation reduction model. Specifically, the plan creation unit 66 specifies the value of each component x that minimizes the value of the new objective function G(x) by using a method such as mathematical programming.

[0058] Then, the plan creating section 66 adds information about the improvement plan created in s17 to the improvement plan information 1300 (s18).

[0059] The plan creation section 66 may display the contents of the improvement plan information 1300 on a predetermined screen (plan creation result display screen).

[0060] (Plan creation result display screen) 9 is a diagram showing an example of a plan creation result display screen 2000. The plan creation result display screen 2000 has an improvement plan display field 2001 that displays each component (for example, the timetable of each train) in the target plan and the improvement plan, a deviation display field 2002 that displays the difference (deviation) between the value of the evaluation index of the target plan and the value of the evaluation index of the improvement plan, a component filter setting field 2003 that accepts the setting of a constraint on the change amount of each component from the user, a re-creation field 2004 that is selected by the user when re-creating the improvement plan, and a confirmation approval field 2005 that is selected by the user when the content of the improvement plan information 1300 is confirmed.

[0061] The plan creation support device 50 may generate a control signal for controlling each facility 10 (trains, etc.) to execute the improvement plan based on the created improvement plan information 1300, and control each facility 10 (trains, etc.) based on the generated control signal. Specifically, the plan creation support device 50 transmits the generated control signal to the control system 60, and the control system 60 controls each facility 10 (trains, etc.).

[0062] Next, each step in the planning support process will be described in detail below.

[0063] <Evaluation index causal search processing> FIG. 10 is a flow diagram illustrating the details of the evaluation index causal search process s14.

[0064] The causal relationship searching unit 62 acquires initial information on the causal relationships between each evaluation index and each component and each other evaluation index. Specifically, the causal relationship searching unit 62 acquires diagram causal structure information 700 (s141).

[0065] The schedule causal structure information 700 is set by the user on a predetermined screen (causal structure setting screen), for example.

[0066] (Causal structure setting screen) 11 is a diagram showing an example of a causal structure setting screen 2100. The causal structure setting screen 2100 has a causal structure editing field 2101 for setting each factor (evaluation index or component) constituting a causal relationship and receiving an input from a user for setting the presence or absence of a causal relationship between each factor, an evaluation index type addition field 2102 selected by a user when adding an evaluation index to the causal structure editing field 2101, a component type addition field 2103 selected by a user when adding a component to the causal structure editing field 2101, a setting completion field 2104 selected by a user when completing editing using the causal structure editing field 2101 and reflecting it in the diagram causal structure information 700, and a cancel field 2105 selected by a user when abandoning editing using the causal structure editing field 2101.

[0067] (Diamond causal structure information) 12 is a diagram showing an example of the created diagram causal structure information 700. The diagram causal structure information 700 is information on the initial values ​​of the causal relationships between each evaluation index 701 and each of the other evaluation indexes and each component 702. In the example shown in the figure, "1" is set when there is a causal relationship, "0" is set when there is no causal relationship, and "-1" is set when the presence or absence of a causal relationship is unknown.

[0068] Next, as shown in FIG. 10, the causal relationship searching unit 62 selects one of all the evaluation indexes including the second evaluation index (s142).

[0069] The causal relationship searching unit 62 identifies the causal relationships between the evaluation index selected in s142 and each of the components and each of the other evaluation indexes (s143).

[0070] For example, the causal relationship searching unit 62 identifies the format of the structural equation (presence or absence of each term) based on the contents 603 of the evaluation index in the evaluation index information 600 and the diagram causal structure information 700. The causal relationship searching unit 62 acquires performance information and assigns the acquired performance information to the identified structural equation to identify each coefficient and constant in the structural equation. Such identification of causal relationships can use general statistical causal search, and LiNGAM, for example, can be adopted.

[0071] A structural equation for a certain evaluation index P includes terms for each component i, which is an explanatory variable, and other evaluation indexes j, and each term has a coefficient. In this embodiment, the structural equation is expressed by the following formula (1).

[0072] Evaluation index P = Σ{Ai × other evaluation index i} + Σ{Bj × component j} + C (Equation 1)

[0073] Here, A and B are coefficients and C is a constant, whose values ​​are identified by the causal search described above.

[0074] The causal relationship searching unit 62 stores data of the structural equation including the selected coefficients and constants identified in s143 in the evaluation index causal relationship information 1100 (s144). This ends the evaluation index causal relationship searching process s14.

[0075] (Evaluation index causal relationship information) 13 is a diagram showing an example of the created evaluation index causal relationship information 1100. The evaluation index causal relationship information 1100 is data including ID 1101 of each evaluation index and data 1102 of the causal relationship related to each evaluation index. The data 1102 of the causal relationship includes a list 1103 of terms of other evaluation indexes constituting a structural equation related to each evaluation index (a list of IDs of the evaluation indexes), a list 1104 of coefficients related to each evaluation index term, a list 1105 of terms of each component (a list of IDs of the component), a list 1106 of coefficients related to each component term (a list of IDs of the component), and a constant 1107.

[0076] <Causal relationship assessment correction process> Next, FIG. 14 is a flow diagram illustrating the details of the causal relationship evaluation correction process s15.

[0077] The causal relationship evaluation correction unit 63 acquires the evaluation index causal relationship information 1100 generated in the evaluation index causal search process s14 (s151).

[0078] The causal relationship evaluation correction unit 63 selects one of the evaluation indexes from the evaluation index causal relationship information 1100 acquired in s151 (s152). Specifically, the causal relationship evaluation correction unit 63 acquires one record from the evaluation index causal relationship information 1100.

[0079] The causal relationship evaluation correction unit 63 calculates the accuracy of the causal relationship for the evaluation index selected in s151 (s153).

[0080] For example, the causal relationship evaluation correction unit 63 calculates the value of each evaluation index by inputting corresponding performance information into the causal relationship data 1102 of the record of the evaluation index causal relationship information 1100 acquired in s152. Also, the causal relationship evaluation correction unit 63 calculates the value of each evaluation index directly from the corresponding performance information. The causal relationship evaluation correction unit 63 compares the value of the evaluation index calculated from the evaluation index causal relationship information 1100 with the value of the evaluation index calculated from the performance information to calculate the degree of deviation. Note that the method of specifying the accuracy of the causal relationship described here is one example, and for example, accuracy information may be acquired from a predetermined database or the like, or accuracy information may be input from a user.

[0081] The causal relationship evaluation correction unit 63 executes the process of s153 for all evaluation indexes in the evaluation index causal relationship information 1100 (s154).

[0082] Thereafter, the causal relationship evaluation correction unit 63 displays the accuracy of the causal relationships for each evaluation index calculated in the processing of s153 on a specified screen (causal relationship correction screen), and accepts input from the user specifying whether or not to correct each causal relationship (s155).

[0083] When the causal relationship evaluation correction unit 63 receives a specification to correct each causal relationship (s155: YES), it reflects the specified correction content in the evaluation index causal relationship information 1100 and repeats the processing from s152 onwards. When the causal relationship evaluation correction unit 63 does not receive a specification to correct each causal relationship (when a specification to end correction is received, etc.) (s155: NO), it executes the processing of s157.

[0084] In addition, the causal relationship evaluation correction unit 63 may check whether there is an evaluation index for which the accuracy of the causal relationship calculated in the processing of s153 is below a predetermined threshold, and if there is an evaluation index for which the accuracy of the causal relationship calculated in the processing of s153 is below a predetermined threshold, perform the processing of s157 for that evaluation index.

[0085] (Causal relationship modification screen) 15 is a diagram showing an example of a causal relationship correction screen 2200. The causal relationship correction screen 2200 has a causal relationship edit field 2201 that displays each factor constituting the causal relationship (explanatory variables in a structural equation; specifically, evaluation indexes and components) and the strength (coefficient) of the causal relationship between the factors, an evaluation index display field 2202 that displays the name of the evaluation index related to the causal relationship displayed in the causal relationship edit field 2201, an accuracy display field 2203 that displays the accuracy of the causal relationship displayed in the causal relationship edit field 2201, and an evaluation index selected by the user when correcting the evaluation index displayed in the causal relationship edit field 2201. It has a target correction column 2204, a component correction column 2205 selected by the user when correcting the component displayed in the causal relationship editing column 2201, a coefficient correction column 2206 selected by the user when correcting the coefficient displayed in the causal relationship editing column 2201, a cancel column 2207 selected by the user when discarding the corrections, a correction completion column 2208 selected by the user when confirming the corrections, and a reevaluation execution column 2209 selected by the user when recalculating the accuracy of the causal relationship based on the corrections.

[0086] 14, the causal relationship evaluation correction unit 63 updates each causal relationship based on the content corrected in s155, and reflects the content corrected in s155 in the evaluation index causal relationship information 1100 (s157). This ends the causal relationship evaluation correction process s15.

[0087] <Component Change Recommendation Processing> Next, FIG. 16 is a flow diagram illustrating details of the component change recommendation process s16.

[0088] The component change recommendation unit 64 obtains a target value of the target evaluation index in the improvement plan and a constraint on the amount of change of the values ​​of each component related to the target evaluation index from the value in the target plan to the value in the improvement plan for a second evaluation index (hereinafter referred to as the target evaluation index) among the evaluation indexes related to the causal relationship determined in the processing up to the causal relationship evaluation correction processing s15 (s161).

[0089] For example, the component change recommendation unit 64 acquires the target value of the target evaluation index and the constraints on the change amount of each component from the evaluation index target value information 800. The target value of the target evaluation index is, for example, "a 10% reduction in the congestion rate of each train." The constraints on the change amount of the component are, for example, "the upper limit of the change amount of the arrival time and departure time at each station is 30 seconds."

[0090] The target value and change amount constraints for the target evaluation index may be set by the user on a specific screen, for example.

[0091] (Target value setting screen) 17 is a diagram showing an example of a target value setting screen 2300 for setting a target value of a target evaluation index and a change amount constraint. The target value setting screen 2300 has a target evaluation index display field 2301 in which the target evaluation index is displayed, a current value display field 2302 in which the value of the target evaluation index in the target plan is displayed, a target value setting field 2303 in which the setting of the target value of the target evaluation index is received from the user, a component display field 2304 in which the component for which a change amount constraint is set is displayed, a change amount target setting field 2305 in which the setting of the change amount constraint is received from the user, a setting completion field 2306 selected by the user when each set value is to be stored, and a cancel field 2307 selected by the user when each set value is to be discarded.

[0092] Next, as shown in FIG. 16, the component change recommendation unit 64 selects one of the target evaluation indexes (s162).

[0093] The component change recommendation unit 64 acquires a structural equation that defines the causal relationship related to the target evaluation index selected in s162 (s163). Specifically, the component change recommendation unit 64 acquires the contents of the record related to the target evaluation index from the evaluation index causal relationship information 1100 generated in the causal relationship evaluation correction process s15.

[0094] The constituent element change recommendation unit 64 selects one of the factors (explanatory variables) in the structural equation acquired in s163 (s164).

[0095] The component change recommendation unit 64 acquires the accuracy of the structural equation related to the explanatory variable selected in s164 (the accuracy calculated in the causal relationship evaluation correction process s15). The component change recommendation unit 64 determines whether the acquired accuracy exceeds a pre-stored threshold value (s165).

[0096] If the acquired accuracy exceeds the pre-stored threshold (s165: YES), the component change recommendation unit 64 executes the process of s166. If the acquired accuracy does not exceed the pre-stored threshold (s165: NO), the component change recommendation unit 64 executes the process of s167.

[0097] In s166, the constituent element change recommendation unit 64 expands (substitutes) the explanatory variables selected in s164 into a structural equation related to the explanatory variables selected in s164. After that, the process of s168 is performed.

[0098] In s167, the constituent element change recommendation unit 64 deletes the explanatory variable selected in s164 from the structural equation related to the explanatory variable selected in s164. After that, the process of s168 is performed.

[0099] In s168, the component change recommendation unit 64 checks whether or not all factors (explanatory variables) in the structural equation obtained in s163 have been selected. If all explanatory variables in the structural equation obtained in s163 have been selected, the component change recommendation unit 64 executes the process of s169, and if there is an unselected explanatory variable in the structural equation obtained in s163, the component change recommendation unit 64 executes the process of s164 to select that explanatory variable.

[0100] Through the above process, the constituent element change recommendation unit 64 configures the structural equation related to the target evaluation index with only constituent elements having a causal relationship strength equal to or greater than a predetermined value.

[0101] In s169, the component change recommendation unit 64 checks whether or not all the target evaluation indexes have been selected. If all the target evaluation indexes have been selected, the component change recommendation unit 64 executes the process of s170, and if there is an unselected target evaluation index, the component change recommendation unit 64 repeats the process of s162 to select that target evaluation index.

[0102] In s170, the component change recommendation unit 64 specifies, for each target evaluation index, an amount of change for each component such that the value of the target evaluation index (the value of the structural equation specified by the processes up to s169) is at least equal to or greater than the target value obtained in s161 and satisfies the constraints of each component obtained in s161. The component change recommendation unit 64 can specify the amount of change for each component based on a solution algorithm for any optimization problem, such as a greedy method.

[0103] For example, when the target evaluation index is the congestion rate and the components of the structural equation related to the congestion rate include the arrival and departure times of each station, the component change recommendation unit 64 calculates the deviation in the arrival and departure times of each train in the improved plan (the difference from the target plan) such that the congestion rate of each train is reduced by 10% compared to the target plan and the deviation in the arrival and departure times of each train at each station is 30 seconds or less compared to the target plan. Note that, when the component change recommendation unit 64 cannot specify the amount of change of each component, it may change the target value or constraint (lower the target value or relax the constraint) and perform recalculation.

[0104] The component change recommendation unit 64 stores the change amount of each component specified in s170 for the target plan in the component change recommendation information 1200.

[0105] The component change recommendation section 64 may display the created component change recommendation information 1200 on a predetermined screen (a component change recommendation display screen).

[0106] (Component change recommendation display screen) 18 is a diagram showing an example of a component change recommendation display screen 2400. The component change recommendation display screen 2400 has a component display field 2401 that displays the contents of each component in the target plan and the improvement plan (for example, the timetable of each train), a target value setting field 2402 that displays a target value of a target evaluation index and accepts the setting from the user, a component filter setting field 2403 that displays a constraint on the change amount and accepts the setting from the user, and a confirmation completion field 2404 that is selected by the user when closing the component change recommendation display screen 2400.

[0107] As described above, the plan creation support device 50 of this embodiment identifies a causal relationship between the value of a new evaluation index (second evaluation index) in an improvement plan and the value of each component based on a predetermined algorithm, and calculates a change value of each component that satisfies a predetermined condition for the second evaluation index from the value calculated by the first model (objective function F(x)) related to the target plan based on the identified causal relationship. Then, the plan creation support device 50 identifies a value of each component that optimizes the values ​​of the first evaluation index and the second evaluation index of the improvement plan based on the above-mentioned first model and a second model (objective function G(x)) that calculates a new value of each component that minimizes the deviation from each calculated change value.

[0108] Specifically, for example, the plan creation support device 50 assumes a transportation operation schedule, such as a train timetable, as the business plan.

[0109] That is, the plan creation support device 50 of this embodiment calculates an optimal change value for each component that satisfies a predetermined condition for the second evaluation index based on the causal relationship between the value of the second evaluation index in the business plan (train schedule, etc.) and the value of each component.The plan creation support device 50 then uses a second model that can calculate a new value for each component close to the optimal change value and a first model of the conventional business plan (train schedule, etc.) to identify the value of each component that optimizes the values ​​of the first evaluation index and the second evaluation index of the business plan (train schedule, etc.).

[0110] In this way, according to the plan creation support device 50 of the present embodiment, even when the second evaluation index is considered in addition to the first evaluation index, it is possible to identify the values ​​of each component element that optimizes the values ​​of the first evaluation index and the second evaluation index without modifying the conventional first model. As a result, even when the evaluation index in the plan is modified, it is possible to support the rapid creation of a new plan based on the modified evaluation index.

[0111] In particular, it is possible to support the rapid creation of new train schedules, etc. based on the changed evaluation index. For example, it is possible to support the rapid creation of train schedules, etc. that not only reduce the congestion rate of trains, etc., but also take the environment into consideration.

[0112] Moreover, the plan creation support device 50 of this embodiment identifies the above-mentioned causal relationship based on performance information of the target plan and the value of the second evaluation index calculated from the performance information.

[0113] This makes it possible to accurately calculate the causal relationship between the value of the second evaluation index and the value of each component element based on the operation record of the train schedule.

[0114] Moreover, the plan creation support device 50 of this embodiment identifies the causal relationship based on information on the components having a causal relationship with the value of the second evaluation index (diagram causal structure information 700).

[0115] In this way, by using the diagram causal structure information 700, it is possible to accurately identify the causal relationship between the value of the second evaluation index and the value of each component element.

[0116] In addition, the plan creation support device 50 of this embodiment calculates the accuracy of the identified causal relationship based on performance information, and if the accuracy of the calculated causal relationship is below a predetermined threshold, accepts input for correcting the causal relationship from the user.

[0117] This can improve the accuracy of the causal relationship.

[0118] In addition, the plan creation support device 50 of this embodiment calculates, based on the causal relationship between the second evaluation index and each component, the target value of the second evaluation index to be achieved and a change value of each component that satisfies the conditions regarding the value of each component.

[0119] In this way, by calculating the change value using the second evaluation index and each of the components related thereto as conditions, it is possible to more reliably create an optimal business plan.

[0120] Moreover, the planning support device 50 of this embodiment outputs information indicating the values ​​of the identified new components to the output device (planning result display screen 2000).

[0121] This allows the user to check the contents of the improvement plan.

[0122] Moreover, the plan creation support system 1 of this embodiment generates a control signal for executing an improvement plan configured by the values ​​of the new components identified above, and controls the equipment 10 based on the generated task signal.

[0123] This makes it possible to realize the implementation of optimal improvement plans (train operations, etc.) that take into account the new second evaluation index.

[0124] The present invention is not limited to the above embodiment, and can be implemented using any components without departing from the scope of the present invention. The above-described embodiments and modifications are merely examples, and the present invention is not limited to these contents as long as the features of the invention are not impaired. In addition, although various embodiments and modifications have been described above, the present invention is not limited to these contents. Other aspects conceivable within the scope of the technical idea of ​​the present invention are also included in the scope of the present invention.

[0125] Furthermore, part of the hardware included in each device of this embodiment may be provided in another device.

[0126] Furthermore, each program of each device may be provided in another device, a program may consist of multiple programs, or multiple programs may be integrated into one program.

[0127] In addition, in this embodiment, the planning support system 1 supports the planning of train timetables for railways, but it can also be applied to transportation means other than railways (buses, monorails, etc.) that operate according to a predetermined operation schedule (timetable).

[0128] Furthermore, the planning support system 1 can be applied to various planning tasks such as the task of manufacturing a product through a predetermined process, or the task of performing a service through a predetermined procedure, in addition to the task of supporting the creation of a transportation schedule. In other words, the planning support system 1 can support the creation of a reproducible plan.

[0129] Furthermore, the structural equation described in this embodiment is just an example, and any form of equation with evaluation indexes and components as explanatory variables can be adopted.

[0130] In addition, although each objective function in this embodiment is a function that is optimal when its value is a minimum value, it may be a function that is optimal when its value is a maximum value, or a function that is optimal when its value is any other value.

[0131] In addition, the method of identifying the causal relationship between the evaluation index and the component (other evaluation index) is not limited to the algorithm described in this embodiment. For example, any causal search method using an equation or model other than the structural equation may be adopted, information on the causal relationship may be obtained from a predetermined database, or information on the causal relationship may be input from a user. [Explanation of symbols]

[0132] 50 Planning support device, 62 Causal relationship search unit, 63 Causal relationship evaluation correction unit, 64 Component change recommendation unit, 66 Planning unit

Claims

1. a storage device that stores a first model that specifies values ​​of each component that optimizes a value of a first evaluation index of the business plan, the value being calculated based on each component that constitutes the business plan; and a causal relationship search process for identifying a causal relationship between a value of a second evaluation index of the business plan calculated based on each of the components based on a predetermined algorithm and a value of each of the components; a component change recommendation process for calculating a change value of each of the components that satisfies a predetermined condition related to the second evaluation index, from a value calculated by the first model, based on the identified causal relationship; a second model that calculates values ​​of each of the components so as to minimize deviation from the calculated change values ​​of each of the components, and a plan creation process that identifies new values ​​of each of the components that optimize the values ​​of the first evaluation index and the second evaluation index of the business plan based on the first model; A planning support device comprising a control device that executes the above.

2. The control device includes: In the causal relationship search process, the causal relationship is identified based on performance information of each component that constitutes the business plan and a value of a second evaluation index calculated from the performance information. The planning support device according to claim 1 .

3. the storage device stores causal structure information, which is information on a component having a causal relationship with the value of the second evaluation index, among the components; The control device includes: In the causal relationship search process, the causal relationship is identified based on values ​​of each component element identified by the causal structure information. The planning support device according to claim 1 .

4. The control device includes: calculating the accuracy of the identified causal relationship based on performance information of each component of the business plan, confirming whether the accuracy of the calculated causal relationship is equal to or lower than a predetermined threshold, and executing a causal relationship evaluation correction process that accepts an input for correcting the calculated causal relationship if the accuracy of the calculated causal relationship is equal to or lower than the predetermined threshold; The planning support device according to claim 1 .

5. the control device calculates, in the component change recommendation process, a target value of the second evaluation index to be achieved and a change value of each of the components that satisfies a condition regarding the value of each of the components, based on the identified causal relationship. The planning support device according to claim 1 .

6. The plan creation support device according to claim 1 , wherein the control device outputs information indicating values ​​of the identified new component elements to an output device.

7. the storage device stores a first model that specifies values ​​of each component element that optimizes a value of a first evaluation index of a transportation schedule, the first evaluation index being calculated based on each component element that constitutes the transportation schedule; The control device includes: In the causal relationship search process, a causal relationship is identified between a value of a second evaluation index of the diamond calculated based on each of the components and a value of each of the components; In the plan creation process, a value of each new component element that optimizes the values ​​of the first evaluation index and the second evaluation index of the schedule is identified. The planning support device according to claim 1 .

8. A planning support system including a planning support device that supports the creation of a business plan and a predetermined facility that executes the business plan, The planning support device comprises: a storage device that stores a first model that specifies values ​​of each component that optimizes a value of a first evaluation index of the business plan, the value being calculated based on each component that constitutes the business plan; and a causal relationship search process for identifying a causal relationship between a value of a second evaluation index of the business plan calculated based on each of the components based on a predetermined algorithm and a value of each of the components; a component change recommendation process for calculating a change value of each of the components that satisfies a predetermined condition related to the second evaluation index, from a value calculated by the first model, based on the identified causal relationship; a second model that calculates values ​​of each of the components so as to minimize deviation from the calculated change values ​​of each of the components, and a plan creation process that identifies new values ​​of each of the components that optimize the values ​​of the first evaluation index and the second evaluation index of the business plan based on the first model; A process of generating a control signal for executing a new business plan configured by the values ​​of each of the identified components, and controlling the equipment based on the generated business signal; A planning support system comprising a control device that executes the above.

9. A plan creation method using an information processing device including a storage device that stores a first model that specifies values ​​of each component that optimizes a value of a first evaluation index of a business plan, the first evaluation index being calculated based on each component that constitutes the business plan, and a control device, The control device, a causal relationship search process for identifying a causal relationship between a value of a second evaluation index of the business plan calculated based on each of the components based on a predetermined algorithm and a value of each of the components; a component change recommendation process for calculating a change value of each of the components that satisfies a predetermined condition related to the second evaluation index, from a value calculated by the first model, based on the identified causal relationship; a second model that calculates values ​​of each of the components so as to minimize deviation from the calculated change values ​​of each of the components, and a plan creation process that identifies new values ​​of each of the components that optimize the values ​​of the first evaluation index and the second evaluation index of the business plan based on the first model; A method to assist in creating a plan.

Images