Simulation system and simulation method

Abstract

To provide a simulation system capable of displaying numerous scenarios generated by simulation in a format easy for a user to analyze them.SOLUTION: The simulation system retains information representing multiple scenarios of index vectors generated by simulation and multiple clusters in which index vectors are classified in each of multiple time steps. The simulation system is configured to classify the clusters of the final time step of the multiple scenarios into multiple groups, and for each combination of time step and cluster, to calculate a transition probability that an index vector belonging to the cluster at the time step transits to each group at the final time step, and a cluster to which an index vector belonging to the cluster at the time step is capable of transiting to the next cluster at the next step, and to display the transition probability and the cluster which is capable of the transition.SELECTED DRAWING: Figure 1

Description

[Technical field]

[0001] The present invention relates to a simulation system and a simulation method. [Background technology]

[0002] Background of the present invention is International Publication No. 2016 / 194051 (Patent Document 1). This publication states that "the system holds evaluation information including an evaluated parameter set and an evaluation value of an attention index for each evaluated parameter set, selects a first evaluated parameter set set and a corresponding evaluation value of an attention index from the evaluation information for each parameter set of a first parameter set set, calculates a first weight corresponding to the evaluation value of the attention index for each parameter set of the first evaluated parameter set set based on the distance from each evaluated parameter set of the first evaluated parameter set set, and calculates an estimate of the statistic of the attention index of the parameter set of the first parameter set set based on the evaluation value of the selected attention index and the first weight" (see Abstract). [Prior art documents] [Patent documents]

[0003] [Patent Document 1] International Publication No. 2016 / 194051 Summary of the Invention [Problem to be solved by the invention]

[0004] In Monte Carlo simulation as in the technology described in Patent Document 1, many scenarios (time series) of index vectors consisting of many indexes are generated, but a method for displaying the many generated scenarios in a format that is easy for users to analyze is not disclosed. In particular, Patent Document 1 does not disclose a display that makes it easy for users to search for important branching points for obtaining a desired index vector for a certain time step. Therefore, one aspect of the present invention aims to display many scenarios generated by simulation in a format that is easy for users to analyze. [Means for solving the problem]

[0005] In order to solve the above problem, one aspect of the present invention employs the following configuration: A simulation system includes a processor, a memory, and a display device, the memory holds simulation results indicating a plurality of scenarios each indicating index vectors of a plurality of time steps generated by simulation and a plurality of clusters into which the index vectors of the plurality of scenarios are classified at each of the plurality of time steps, the processor accepts settings for classifying clusters of final time steps of the plurality of scenarios into a plurality of groups and executes classification based on the settings, executes a scenario analysis process for calculating, for each combination of the time steps and the clusters, a transition probability that an index vector belonging to the cluster at the time step will transition to each group at the final time step, and a cluster to which an index vector belonging to the cluster at the time step can transition in the next step, and displays, on the display device, an analysis map including a display indicating the transition probability and a display indicating the cluster to which the transition can be made for each combination of the time steps and the clusters. Effect of the Invention

[0006] According to one aspect of the present invention, a large number of scenarios generated by a simulation can be displayed in a format that is easy for a user to analyze.

[0007] Problems, configurations and effects other than those described above will become apparent from the following description of the embodiments. [Brief description of the drawings]

[0008] [Figure 1] 1 is a block diagram showing an example of the configuration of a simulation system according to a first embodiment; [Diagram 2] FIG. 2 is a diagram illustrating an example of a screen configuration of a portal screen in the first embodiment. [Diagram 3] FIG. 4 is a diagram illustrating an example of a data configuration of index data in the first embodiment. [Figure 4] FIG. 4 is a diagram illustrating an example of a data configuration of simulation input data in the first embodiment. [Diagram 5] 4 is a flowchart illustrating an example of a simulation process in the first embodiment. [Figure 6] 4 is an example of simulation result data in the first embodiment. [Figure 7] FIG. 2 is an explanatory diagram illustrating an example of a simulation process in the first embodiment. [Figure 8] FIG. 2 is an explanatory diagram illustrating an example of a simulation process in the first embodiment. [Figure 9] 1 is a flowchart illustrating an example of a scenario analysis process in the first embodiment. [Figure 10] FIG. 2 is a diagram illustrating an example of a data configuration of index change rate data in the first embodiment. [Figure 11] FIG. 4 is a diagram illustrating an example of a data configuration of group data in the first embodiment. [Figure 12] FIG. 2 is a diagram illustrating an example of a data configuration of scenario analysis map drawing data in the first embodiment. [Figure 13] FIG. 13 is a diagram showing an example of a screen configuration of a scenario analysis result display screen (group organization) in the first embodiment. [Figure 14] FIG. 13 is a diagram showing another example of the screen configuration of the scenario analysis result display screen (group formation) in the first embodiment. [Figure 15]11 is a flowchart illustrating an example of a branch point setting process in the first embodiment. [Figure 16] FIG. 4 is a diagram illustrating an example of a data configuration of branch point setting data in the first embodiment. [Figure 17] FIG. 13 is a diagram showing an example of a screen configuration of a scenario analysis result display screen (branch setting) in the first embodiment. [Figure 18] 11 is a flowchart illustrating an example of a cause analysis process according to the first embodiment. [Figure 19] FIG. 11 is a diagram illustrating an example of a data configuration of factor analysis result data in the first embodiment. [Figure 20] FIG. 13 is a diagram showing an example of a screen configuration of a scenario analysis result display screen (group confirmation) in the first embodiment. [Figure 21] FIG. 13 is a diagram showing another example of the screen configuration of the scenario analysis result display screen (group confirmation) in the first embodiment. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0009] Hereinafter, an embodiment of the present invention will be described in detail with reference to the drawings. In this embodiment, the same components are generally designated by the same reference numerals, and repeated explanations are omitted. Note that this embodiment is merely an example for realizing the present invention, and does not limit the technical scope of the present invention. EXAMPLES

[0010] 1 is a block diagram showing an example of the configuration of a simulation system 100. The simulation system 100 is configured by a computer having, for example, a CPU (Central Processing Unit) 101, a memory 102, an auxiliary storage device 103, a communication device 104, an input device 105, and an output device 106.

[0011] The CPU 101 includes a processor and executes programs stored in the memory 102. The memory 102 includes a ROM (Read Only Memory), which is a non-volatile storage element, and a RAM (Random Access Memory), which is a volatile storage element. The ROM stores immutable programs (e.g., a Basic Input / Output System (BIOS)). The RAM is a high-speed, volatile storage element such as a DRAM (Dynamic Random Access Memory), and temporarily stores programs executed by the CPU 101 and data used when the programs are executed.

[0012] The auxiliary storage device 103 is a large-capacity non-volatile storage device such as a magnetic storage device (HDD (Hard Disk Drive)) or a flash memory (SSD (Solid State Drive)), and stores programs executed by the CPU 101 and data used when the programs are executed. That is, the programs are read from the auxiliary storage device 103, loaded into the memory 102, and executed by the CPU 101.

[0013] The input device 105 is a device such as a keyboard or a mouse that receives input from an operator. The output device 106 is a device such as a display device or a printer that outputs the results of program execution in a format that can be viewed by the operator.

[0014] The communication device 104 is a network interface device that controls communication with other devices according to a predetermined protocol. The communication device 104 may also include a serial interface such as a Universal Serial Bus (USB).

[0015] A part or all of the programs executed by the CPU 101 may be provided to the simulation system 100 from a removable medium (CD-ROM, flash memory, etc.) which is a non-transitory storage medium, or from an external computer equipped with a non-transitory storage device via a network, and may be stored in a non-volatile auxiliary storage device 103 which is a non-transitory storage medium. For this reason, the simulation system 100 may have an interface for reading data from the removable medium.

[0016] The simulation system 100 is a computer system configured on one physical computer or on multiple logically or physically configured computers, and may operate in separate threads on the same computer, or may operate on a virtual computer constructed on multiple physical computer resources.

[0017] The CPU 101 includes, for example, a simulation unit 111, a scenario analysis unit 112, a branch point setting unit 113, a factor analysis unit 114, and a screen generation unit 115, all of which are functional units.

[0018] The simulation unit 111 executes a simulation to generate a scenario that is a time series of an index vector consisting of a plurality of indexes. In the following, in this embodiment, an example in which the simulation unit 111 predicts an index vector at each time step using Monte Carlo simulation will be described, but the algorithm used in the simulation is not limited to Monte Carlo simulation.

[0019] The scenario analysis unit 112 analyzes the scenario generated by the simulation unit 111. The branch point setting unit 113 sets branch points and branch destinations, which will be described later. The factor analysis unit 114 analyzes factors leading from the branch point to the branch destination. The screen generation unit 115 generates information to display a screen on the output device 106.

[0020] For example, CPU 101 functions as a simulation unit 111 by operating according to a simulation program loaded into memory 102, and functions as a scenario analysis unit 112 by operating according to a scenario analysis program loaded into memory 102. The relationships between programs and functional units are similar for other functional units included in CPU 101.

[0021] Some or all of the functions of the functional units included in the CPU 101 may be realized by hardware such as an Application Specific Integrated Circuit (ASIC) or a Field Programmable Gate Array (FPGA).

[0022] The auxiliary storage device 103 holds, for example, index data 121, simulation input data 122, simulation result data 123, index change rate data 124, group data 125, scenario analysis map drawing data 126, branch point setting data 127, and factor analysis result data 128.

[0023] The index data 121 includes information about the indexes included in the index vector. The simulation input data 122 includes input information for the simulation by the simulation unit 111. In the simulation input data 122, for example, a directed graph in which each index is a node and causal relationships between the indexes are indicated by edges is defined in a description format such as a CLD (Causal Loop Diagram).

[0024] Simulation result data 123 indicates the results of a simulation by the simulation unit 111. Index change rate data 124 indicates the rate of change of an index value in a scenario generated by the simulation unit 111. Group data 125 indicates a group of a scenario generated by the simulation unit 111.

[0025] The scenario analysis map drawing data 126 is information for drawing an analysis map of a scenario. The branch point setting data 127 indicates the set branch points and branch destinations. The factor analysis result data 128 indicates the result of the factor analysis.

[0026] It should be noted that some or all of the information stored in the auxiliary storage device 103 may be stored in the memory 102 or in an external database connected to the simulation system 100.

[0027] In this embodiment, the information used by the simulation system 100 does not depend on the data structure and may be expressed in any data structure. In this embodiment, the information is expressed in a table format, but the information can be stored in a data structure appropriately selected from, for example, a list, a database, or a queue.

[0028] 2 is a diagram showing an example of the screen configuration of a portal screen. The portal screen 200 is displayed on the output device 106 by the screen generating unit 115. On the portal screen 200, for example, a simulation execution button 201, a scenario analysis button 202, and a cause analysis execution button 203 are displayed.

[0029] When the simulation execution button 201 is selected, a simulation process described below is executed. When the scenario analysis button 202 is selected, a scenario analysis process described below is executed. However, the scenario analysis process can be executed after the simulation process is executed. When the cause analysis execution button 203 is selected, a cause analysis process described below is executed. However, the cause analysis process can be executed after the scenario analysis process is executed.

[0030] 3 is a diagram showing an example of the data configuration of the index data 121. The index data 121 includes, for example, an index ID column 1211, an index name column 1212, an initial value column 1213, an important index flag column 1214, an index category ID column 1215, and an index category name column 1216.

[0031] The index ID column 1211 holds an index ID for identifying an index. The index name column 1212 holds the name of the index. The initial value column 1213 holds the initial value of the index in the Monte Carlo simulation (the value at the 0th step). In the example of FIG. 3, the initial values ​​of all the indexes are normalized to 1, that is, the rate of increase or decrease from the initial value of each index is predicted at each time step (hereinafter simply referred to as a step) of the Monte Carlo simulation, but the initial value of each index value may be set to a non-normalized value itself, and the value of each index itself may be predicted at each step of the Monte Carlo simulation.

[0032] The important indicator flag column 1214 holds an important indicator flag indicating whether each indicator is an important indicator (if it is an important indicator, 1 is stored, if it is not an important indicator, 0 is stored). The indicator category ID column 1215 holds an indicator category ID that identifies the indicator category to which each indicator belongs. The indicator category name column 1216 holds information indicating the name of the indicator category.

[0033] 4 is a diagram showing an example of a data configuration of the simulation input data 122. The simulation input data 122 holds, for example, an edge ID column 1221, an edge output source index ID column 1222, an edge output destination index ID column 1223, a linearity column 1224, a delay column 1225, a linear variation column 1226, and a delay variation column 1227.

[0034] The edge ID column 1221 holds edge IDs for identifying edges in the CLD. The edge output source index ID column 1222 holds index IDs of indices that are output sources of edges. The edge output destination index ID column 1223 holds index IDs of indices that are output destinations of edges.

[0035] The linear field 1224 holds information indicating a linear change rate (hereinafter, simply referred to as a coefficient) of the index of the output destination of the edge when the index of the output source of the edge changes. The linear field 1224 can store not only positive values ​​but also negative values. In other words, the relationship between the index of the edge output source and the index of the edge output destination may be a positive correlation or an inverse correlation. The delay field 1225 holds information indicating a time lag from when the index of the output source of the edge changes to when the index of the output destination of the edge changes.

[0036] The linear variation field 1226 holds a value indicating the variation of the values ​​in the linear field 1224. In the example of FIG. 4, the coefficient indicated in the linear field 1224 for edge ID "001" is "0.5" and the value in the linear variation field 1226 is "0.3", so in this case the coefficient of the edge is selected from the range of 0.2 (=0.5-0.3) or more and 0.8 (=0.5+0.3) or less. That is, for example, when the index of the edge output source index ID "001" corresponding to the edge ID "001" increases by 1%, the index of the edge output destination index ID "002" corresponding to the edge ID "001" increases by any rate (for example, randomly determined) included in the range of 0.2% to 0.8%.

[0037] The delay variation column 1227 holds a value indicating the variation of the coefficient indicated by the linearity column 1224. In the example of FIG. 4, the value of the delay column 1225 for the edge ID "001" is "60" and the value of the delay variation column 1227 is "40", so in this case the delay of the edge is selected from a range of 20 (=60-40) or more and 100 (=60+40) or less. That is, for example, after the index of the edge output source index ID "001" corresponding to the edge with the edge ID "001" changes, when any step (for example, randomly determined) included in the range of 20 steps to 100 steps has passed, the index of the edge output destination index ID "002" corresponding to the edge with the edge ID "001" changes in response to the change.

[0038] To summarize the relationship between the indices taking linearity and delay into consideration, when the edge output source index for the t-th step is denoted as A(t), the edge output destination index for the t-th step is denoted as B(t), the coefficient is denoted as a, and the delay is denoted as b, for example, the value of the edge output destination index for the t+b-th step can be calculated by B(t+b)=a*(A(t)-A(t-1))+B(t+b-1). Note that the coefficient value taking variability into consideration and the lower and upper limits of the delay value taking variability into consideration may be determined in advance.

[0039] 5 is a flowchart showing an example of a simulation process. The simulation unit 111 executes an initial setting of the simulation (S501). In step S501, the simulation unit 111 sets index data 121 and simulation input data 122 according to an input from a user via the input device 105, for example.

[0040] Also, in step S501, the simulation unit 111 sets, for example, according to input from a user via the input device 105, the total number of time steps in the Monte Carlo simulation (hereinafter referred to as N steps), the number of scenarios generated for each scenario at each time step of the Monte Carlo simulation (hereinafter referred to as L), the number of clusters generated at each time step of the Monte Carlo simulation (hereinafter referred to as M), and the number of scenarios obtained from each cluster by resampling at each time step of the Monte Carlo simulation (hereinafter referred to as R).

[0041] The simulation unit 111 sets the number of (time) steps k in the Monte Carlo simulation to 0 (S502). The simulation unit 111 obtains the initial value of the index vector from the initial value field 1213 of the index data 121, and registers it in the simulation result data 123 (S503). Note that the index vector is a vector in which the values ​​of the indexes defined in the index data 121 are arranged (for example, in ascending order of the index ID).

[0042] The simulation unit 111 increments the number of steps k (by 1) (S504). The simulation unit 111 executes a Monte Carlo simulation for each resampled scenario (where, when k=1, there is only one scenario consisting of only the initial value of the index vector) from which index vectors up to the k-1th step have been obtained, and calculates the index vector for the kth step (S505). For example, a model for executing the Monte Carlo simulation may be stored in the auxiliary storage device 103 in advance.

[0043] In step S505, the simulation unit 111 performs the following process in L patterns for each scenario up to the k-1th step, thereby calculating L index vectors for the kth step for each scenario. Specifically, the simulation unit 111 minutely changes (for example, randomly) the value of each index included in the index vector from the value in the k-1th step, and further calculates the change in other indexes due to the minute change in the value of each index based on the structure of the CLD indicated by the index data 121 (note that the coefficients and delays in the CLD are, for example, randomly determined from a range in which variation is taken into consideration), and executes a process to reflect the calculation results. Therefore, in step S505, L scenarios up to the kth step are generated from each of the scenarios up to the k-1th step.

[0044] The simulation unit 111 classifies the values ​​of the index vectors of the kth step of the scenarios generated in step S505 into M clusters (S506), thereby clustering the scenarios. The simulation unit 111 may perform clustering on the scenarios themselves, rather than on the values ​​of the index vectors of the kth step of the scenarios generated in step S505. The K-means method is an example of a clustering algorithm used in step S506. The simulation unit 111 generates cluster numbers (e.g., clusters closer in distance have closer cluster numbers) to identify each cluster generated in step S506.

[0045] The simulation unit 111 resamples the k-th step index vector values ​​for each cluster at R intervals, thereby resampling the scenarios up to the k-th step at R intervals (S507). Note that the simulation unit 111 may perform resampling on the scenario itself, rather than on the k-th step index vector values ​​of the scenario. Note that the resampling in step S507 is performed by, for example, a method similar to the method used in the particle filter method.

[0046] The simulation unit 111 calculates the transition probability from each cluster at the k-1th step to each cluster at the kth step (S508). Specifically, for example, the simulation unit 111 refers to the scenarios up to the kth step that have been resampled, and calculates the ratio of index vectors belonging to the cluster with cluster number j at the kth step among the scenarios having index vectors belonging to the cluster with cluster number i at the k-1th step for all combinations of (i, j) (where i, j are natural numbers between 1 and M), thereby calculating the transition probability. Note that since clustering is not performed at the 0th step, for example, the index vector at the 0th step (i.e., the initial value) may be treated as belonging to a predetermined cluster (for example, a cluster with cluster number 1).

[0047] The simulation unit 111 determines whether the number of steps has reached the final step (i.e., the Nth step) (S509). If the simulation unit 111 determines that the number of steps has not reached the final step (S509: NO), the process returns to step S504.

[0048] When the simulation unit 111 determines that the number of steps has reached the final step (S509: YES), it determines a representative scenario for each cluster in the Nth step, which is the final step (S510). Specifically, for example, for each cluster in the Nth step, the simulation unit 111 determines the scenario having the index vector closest to the center of gravity of the cluster as the representative scenario of the cluster.

[0049] The simulation unit 111 registers the simulation result in the simulation result data 123 (S511) and ends the simulation process. In step S511, for example, the simulation unit 111 assigns a scenario ID for identifying the scenario to each of the finally resampled scenarios, and registers the index vector value and cluster number at each time step of each scenario, and a representative scenario flag in the simulation result data 123 (in scenario data 1230 contained therein, which will be described later). In addition, in step S511, the simulation unit 111 registers, for example, the transition probability calculated in step S508 in the simulation result data 123 (in cluster transition probability data 1235 contained therein, which will be described later).

[0050] 6 is an example of the simulation result data 123. The simulation result data 123 includes, for example, scenario data 1230 and cluster transition probability data 1235 for each number of steps.

[0051] The scenario data 1230 includes, for example, a scenario ID column 1231 , an index vector value column for each step 1232 , a cluster number column for each step 1233 , and a representative scenario flag column 1234 .

[0052] The scenario ID column 1231 holds a scenario ID. The index vector value per step column 1232 holds the index vector value of each scenario in the step. The cluster number per step column 1233 holds the cluster number of the cluster to which the index vector value of each scenario in the step belongs. The representative scenario flag column 1234 holds a representative scenario flag indicating whether each scenario is a representative scenario (1 is stored if it is a representative scenario, 0 is stored if it is not a representative scenario).

[0053] The cluster transition probability data 1235 indicates the probability of transition from each cluster in the k-1th step to each cluster in the kth step, for each of k = 1,...,N. Specifically, for example, the record 1236 indicates that an index vector belonging to a cluster with cluster number 1 in the N-1th step transitions to a cluster with cluster number 2 in the Nth step with a probability of 0.1%.

[0054] Fig. 7 is an explanatory diagram showing an example of a simulation process. In the example of Fig. 7, the index vector is made up of index A and index B. In the 0th step of the simulation process, only one scenario is obtained, which is made up of only the initial value 701 of the index vector, that is, (1,1).

[0055] When the process of step S505 in the first step is executed, an index vector group 702 consisting of L index vectors in the first step is obtained, that is, L scenarios up to the first step are obtained. Note that each of the index vector group 702 in the first step is a small change from the initial value 701, and therefore is concentrated in the vicinity of the initial value 701.

[0056] When the processing of step S506 in the first step is executed, the index vectors included in the index vector group 702 are clustered (here, the index vectors are clustered into three clusters consisting of clusters with cluster numbers 1, 2, and 3).

[0057] The process of step S507 is executed for each cluster in the first step. To explain the process of step S507 by taking the cluster with cluster number 3 as an example, resampling is executed for the index vector group 703 belonging to the cluster with cluster number 3. Note that in this resampling, it is desirable to set the index vector (scenario) closer to the center of gravity of the cluster with cluster number 3 to have a higher probability of being selected (resampling), and in this case, one index vector (scenario) may be selected multiple times, while there may also be an index vector (scenario) that is not selected. Among the points illustrated as the index vector group 703, points with diagonal lines indicate resampled points, and black points indicate index vectors that were not resampled.

[0058] By executing the processing of steps S504 to S509 at each time step, an index vector group 704 (resampling scenario) at the final step (Nth step) is obtained. As the time step progresses, the change from the initial value 701 of the index vector becomes larger, so that, for example, the index vector group 704 at the Nth step has a larger variation and a larger difference between clusters compared to the index vector group 702 at the first step. By executing the processing of step S510 for each cluster of the index vector group 704 at the final step, a representative scenario is selected from each cluster.

[0059] Fig. 8 is an explanatory diagram showing an example of a simulation process. Each point in Fig. 8 indicates an index vector at each step, and the time series of index vectors connected by arrows across time steps indicates a scenario. As explained in Fig. 7, in the 0th step of the simulation process, only one scenario consisting of only the initial value 701 of the index vector, (1,1), is obtained.

[0060] Furthermore, as described above, when the process of step S505 in the first step is executed, L scenarios up to the first step are obtained. As described above, when the process of step S506 in the first step is executed, the index vectors included in the index vector group of the first step are clustered (here, the clustering is performed into M clusters with cluster numbers from 1 to M).

[0061] When the process of step S507 is executed for each cluster in the first step, resampling of index vectors (scenario) is performed. Among the index vectors in the first step, those that are the starting points of the arrows are index vectors that were subject to resampling, and those that are not the starting points of the arrows are index vectors that were not subject to resampling. There may be index vectors that are not subject to resampling, such as index vector 801, there may be index vectors that are subject to resampling and are selected only once as a starting point for the next step (third step), such as index vector 802, and there may be index vectors that are subject to resampling and are selected multiple times as a starting point for the next step (third step), such as index vector 803.

[0062] By executing the above-mentioned process at each time step, the scenario obtained at the final step is also resampled for each cluster. In the above-mentioned simulation process, the above-mentioned clustering process and resampling process are executed, so that it is possible to generate a variety of scenarios with high prediction accuracy while reducing the calculation load of the simulation system 100.

[0063] 9 is a flowchart showing an example of a scenario analysis process. The scenario analysis unit 112 reads the index data 121 and the simulation result data 123 (S901). The scenario analysis unit 112 calculates the change rate from the initial value of each index value at the final step for the representative scenario of each cluster, and registers it in the index change rate data 124 (S902).

[0064] 10 is a diagram showing an example of the data configuration of index change rate data 124. Index change rate data 124 includes, for example, an index ID column 1241 and an index change rate column 1242. The index ID column 1241 holds an index ID. The index change rate column 1242 holds a value indicating the rate of change from the initial value of each index value for each cluster in the final step.

[0065] Returning to the explanation of Fig. 9, the scenario analysis unit 112 sets the group data 125 based on, for example, an input from the user via the input device 105 (S903).

[0066] 11 is a diagram showing an example of the data configuration of the group data 125. The group data 125 includes, for example, a group number column 1251, a cluster number column 1252, a group name column 1253, a details column 1254, and a display color column 1255.

[0067] The group number column 1251 holds a group number that identifies a group. The cluster number column 1252 holds the cluster numbers of the clusters (in the final step) that belong to the group.

[0068] For example, clusters with consecutive cluster numbers or one cluster belong to each group, and the smaller the cluster number, the smaller the group number belongs to. Therefore, in step S903, for example, the scenario analysis unit 112 determines one or more positions for dividing the cluster numbers arranged in ascending order by input from the input device 105, and generates groups by dividing them at the determined positions. As described above, closer cluster numbers are assigned to clusters that are closer to each other, so that by generating groups using this method, clusters that are closer to each other belong to the same group, and the closer the numbers of the groups, the closer the distance between the clusters are to each other.

[0069] The details column 1254 holds information indicating a detailed explanation of the group, etc. The value stored in the details column 1254 is determined in step S903, for example, by input from the input device 105. The display color column 1255 holds information indicating the display color (for example, RGB values) of each group in a scenario analysis map described later. In step S903, the scenario analysis unit 112 determines the number of groups, for example, by input from the input device 105, and determines the display color of each group by arranging the determined number of groups at positions equally spaced apart in the color space (the distance between groups in the color space is set to be equal to or greater than a predetermined value).

[0070] Returning to the explanation of Fig. 9, the scenario analysis unit 112 sets the (time) step number k to N (i.e., the final step) (S904). The scenario analysis unit 112 decrements the step number k (by 1) (S905). The scenario analysis unit 112 sets the cluster number i to 1 (S906). The scenario analysis unit 112 refers to the simulation result data 123 and extracts a scenario in which an index vector belongs to the cluster with cluster number i at the kth step (S907). The scenario analysis unit 112 sets the group number j to 1 (S908).

[0071] The scenario analysis unit 112 calculates the proportion of the number of scenarios among the extracted scenarios in the most recent processing of step S907, whose index vectors in the final step belong to the group with group number j (i.e., the probability that an index vector belonging to the cluster with cluster number i in the kth step will transition to a cluster included in the group with group number j in the final step), and stores the proportion in the scenario analysis map drawing data 126 (S909).

[0072] Specifically, for example, in step S909, the scenario analysis unit 112 refers to the cluster transition probability data 1235 for each of the kth step to the final step (Nth step), and calculates the cluster transition probability starting from the cluster with cluster number j in the kth step, transitioning through any cluster from the k+1th step to the N-1th step, and ending at the cluster included in the group with group number j in the Nth step.

[0073] For example, in step S909, when k=N-1, the scenario analysis unit 112 obtains the transition probability that a cluster with cluster number i in the N-1th step will transition to any cluster included in the group with group number j in the Nth step from the cluster transition probability data 1235, and calculates the sum of the obtained transition probabilities.

[0074] Also, for example, in step S909, when k is equal to or smaller than N-2, the scenario analysis unit 112 acquires, for all cluster numbers h, the transition probability that a cluster with cluster number i at the kth step transitions to a cluster with cluster number h at the k+1th step from the cluster transition probability data 1235. Furthermore, the scenario analysis unit 112 may calculate, for all cluster numbers h, the product of the transition probability that a cluster with cluster number i at the kth step transitions to a cluster with cluster number h at the k+1th step and the cluster transition probability starting from the cluster with cluster number h at the k+1th step (transitioning through any cluster from the k+2th step to the N-1th step) and ending at a cluster included in the group with group number j at the Nth step, and may then calculate the sum of the calculated products.

[0075] Since no clustering is performed in the 0th step, for example, the index vector (that is, the initial value) in the 0th step may be treated as belonging to a predetermined cluster (for example, a cluster with a cluster number of 1).

[0076] The scenario analysis unit 112 determines whether the group number j has reached the number of the final group (S910). If the scenario analysis unit 112 determines that the group number j has not reached the number of the final group (S910: NO), it increments the group number j (S911) and returns to the processing of step S909.

[0077] If the scenario analysis unit 112 determines that the group number j has reached the number of the final group (S910: YES), it determines whether the cluster number i has reached the number of the final cluster (S912). If the scenario analysis unit 112 determines that the cluster number i has not reached the number of the final cluster (S912: NO), it increments the cluster number i (S913) and returns to the processing of step S907.

[0078] If the scenario analysis unit 112 determines that the cluster number i has reached the number of the final cluster (S912: YES), it determines whether the number of steps k has reached 0 (S914). If the scenario analysis unit 112 determines that the number of steps k has not reached 0 (S914: NO), it returns to the processing of step S905.

[0079] When the scenario analysis unit 112 determines that the number of steps k has reached 0 (S914: YES), the screen generation unit 115 displays a scenario analysis result display screen on the output device 106 (S915), including the group set in step S902, the index change rate calculated in step S902, the important indexes set in the index data 121, and an analysis map based on the scenario analysis map drawing data 126, etc. The branch point setting unit 113 executes a branch point setting process (S916) and ends the scenario analysis process. The branch point setting process will be described in detail later.

[0080] Although details will be described later, on the scenario analysis result display screen displayed in step S915, instructions to reset groups (e.g., resetting the number of groups or clusters belonging to a group) can be accepted in accordance with input from the input device 105, and when an instruction to reset groups is accepted, the processing of steps S903 to S915 is executed based on the instruction.

[0081] 12 is a diagram showing an example of the data configuration of scenario analysis map drawing data 126. The scenario analysis map drawing data 126 includes, for example, group transition probability data 1261 for each number of steps. The group transition probability data 1261 indicates the cluster transition probability calculated in step S909 for each k=0,...,N, starting from each cluster at the kth step, (transitioning through any cluster from the k+1th step to the N-1th step), and ending at a cluster included in each group at the Nth step.

[0082] Specifically, for example, record 1262 indicates that an index vector belonging to a cluster with cluster number 1 in the 0th step transitions with a probability of 22.7% to any cluster included in a group with group number 4 in the Nth step. Note that, in the scenario analysis process shown in Fig. 9, the explanation of the calculation process of group transition probability data 1261 for k = N is omitted, but in the group transition probability data 1261 for k = N, for each cluster number, 1 is stored in the transition probability corresponding to the group to which the cluster of the cluster number belongs, and 0 is stored in the transition probability corresponding to other groups.

[0083] 13 is a diagram showing an example of the screen configuration of a scenario analysis result display screen (group formation) in a state where group formation is selected. A scenario analysis result display screen 1300 in a state where group formation is selected includes, for example, buttons 1301 to 1308, a group setting area 1309, a number of groups input area 1310, a button 1311, an index display area 1312, and an analysis map display area 1313.

[0084] When button 1301 is selected, the screen transitions to the portal screen 200 of Fig. 2. When button 1302 is selected, information about the group being set is saved in group data 125. When button 1303 is selected, an area for inputting detailed settings of the group (for example, values ​​to be registered in group name column 1253 and details column 1254 of group data 125) is displayed, for example, by a pop-up or the like.

[0085] When button 1304 is selected, the processes of steps S904 to S916 are executed again based on the information of the group being set, and an analysis map corresponding to the group is displayed in analysis map display area 1313. When button 1305 is selected, a scenario analysis result display screen 1300 (screen in FIG. 13) in which group organization is possible is displayed. When button 1306 is selected, transition is made to a scenario analysis result display screen (branch setting) described later. When button 1307 is selected, transition is made to a scenario analysis result display screen (group confirmation) described later.

[0086] When the button 1308 is selected, it is possible to change the display unit of the time steps in the analysis map displayed in the analysis map display area 1313. For example, when one time step is one month, when the button 1308 is selected, the display unit of the time steps in the analysis map displayed in the analysis map display area 1313 switches to either one month or one year.

[0087] The group setting area 1309 is an area for setting groups. In the group setting area 1309, a plurality of squares indicating clusters (in the final step) and separators 1320 (the number of groups minus 1) according to the number of groups set in the group number input area 1310 are displayed.

[0088] The multiple cells indicate clusters corresponding to the cluster numbers in ascending order from the left, and clusters corresponding to cells separated by separators 1320 (the number of clusters is 10 in the example of FIG. 13) form groups. For example, groups are formed in such a way that a cluster corresponding to a cell located to the left of the leftmost separator 1320 belongs to a group with group number 1, a cluster corresponding to a cell located between the i-th separator 1320 from the left (i is an integer equal to or greater than 1 and less than the set number of groups j) and the i+1-th separator 1320 from the left belongs to a group with group number i+1, and a cluster corresponding to a cell located to the right of the rightmost separator 1320 belongs to a group with group number j.

[0089] That is, in the example of FIG. 13, clusters with cluster numbers 1 to 3 belong to the group with group number 1 (group name "Group A"), clusters with cluster numbers 4 to 6 belong to the group with group number 2 (group name "Group B"), clusters with cluster numbers 7 to 8 belong to the group with group number 3 (group name "Group C"), and clusters with cluster numbers 9 to 10 belong to the group with group number 4 (group name "Group D").

[0090] The clusters included in each group can be changed by moving the positions of the separators 1320 left or right according to input to the input device 105. The squares corresponding to each group are displayed in the display color indicated by the display color column 1255 of the group data 125.

[0091] The group number input area 1310 is an area for receiving input of the number of groups, and as described above, the number of groups and the number of separators 1320 change according to the numerical value input in the group number input area 1310. When a button 1311 is selected, the indices displayed in the index display area 1312 are sorted (for example, so that important indices are displayed at the top).

[0092] The index display area 1312 displays, for example, a check box for an important index flag, an index category name, and an index name indicated by the index data 121, as well as an index change rate for each cluster (at the final step based on the 0th step) indicated by the index change rate data 124. By inputting to the input device 105, a check can be added or removed from the check box for an important index flag, and the result of changing the check in the check box is reflected in the index data 121.

[0093] In the indicator display area 1312, the indicator change rate is displayed in a matrix with each cluster on the horizontal axis and each indicator on the vertical axis, and important indicators are displayed at the top of the matrix, so that the user can appropriately set groups while checking the change rate of the indicator that he or she considers important. In the example of Fig. 13, "elderly population", which is an important indicator belonging to the indicator category name "welfare", shows a change of +10% to +20% in clusters with cluster number 2 and cluster number 7, and a change of -10% to +10% in other clusters.

[0094] An analysis map for analyzing a scenario is displayed in the analysis map display area 1313. The analysis map includes a pie chart corresponding to each cluster of each step (for simplicity, pie charts other than those for the 0th step, 1st step, 2nd step, and Nth step are omitted in FIG. 13, but in reality, pie charts for all steps are displayed).

[0095] The pie chart for each step indicates the group transition probability corresponding to that step indicated by the scenario analysis map drawing data 126. That is, for example, the pie chart 1331 indicates the group transition probability corresponding to cluster number 1 in the first step in the scenario analysis map drawing data 126 (that is, the probability of transitioning to each group in the final step when the cluster with cluster number 1 in the first step is used as the starting point).

[0096] Each group transition probability in the pie chart is displayed in a display color corresponding to the group in the group data 125, and with an area according to the group transition probability. The pie chart in the analysis map display area 1313 allows the user to easily grasp, for example, to which group the index vector is likely to transition in the final step when each cluster in each time step is used as the starting point.

[0097] Although an example has been described in which the area of ​​a pie chart indicates the specific value of the group transition probability, it is also possible to make it possible to recognize the groups to which a transition can be made in the final step without indicating the specific value of the group transition probability (for example, if in the final step a transition is only possible to either group number 1 or group number 2, then the pie chart may be displayed divided into two equal parts, with group number 1 in the display color and group number 2 in the display color, regardless of these group transition probabilities).

[0098] Also, for example, if an index vector corresponding to cluster number 1 in the first step can transition to the cluster with cluster number 1 in the second step, clusters that may transition across time steps are shown in a connected relationship on the analysis map, such as connecting pie chart 1331 and pie chart 1332 with a straight line (for example, of a color not used as a group display color). This allows the user to easily grasp the state of transition of index vectors between time steps.

[0099] Figure 14 is a diagram showing another example of the screen configuration of the scenario analysis result display screen (group formation) in the state where group formation is selected. The scenario analysis result display screen 1300 in Figure 14 is similar to Figure 13 except for the display in the analysis map display area 1313. The analysis map display area 1313 in Figure 14 differs from Figure 13 in that it does not display a pie chart of each cluster in each step, but only displays straight lines indicating the connection relationships of clusters across time steps.

[0100] Furthermore, by displaying the line in the analysis map display area 1313 in FIG. 14 in a mixed color according to the group transition probability, the user can visually grasp the group transition probability. Specifically, for example, if the probability that an index vector belonging to a cluster with cluster number 1 in the first step will transition to a group with group number 1, group number 2, group number 3, or group number 4 in the final step is 40%, 30%, 20%, and 10%, respectively, and the display colors of groups with group numbers 1, 2, 3, and 4 are (r1, g1, b1), (r2, g2, b2), (r3, g3, b3), and (r4, g4, b4), respectively, the straight line connecting the cluster with cluster number 1 in the first step to the cluster in the second step in the analysis map display area 1313 will be displayed in a color that is a mixture of (r1, g1, b1) at 40%, (r2, g2, b2) at 30%, (r3, g3, b3) at 20%, and (r4, g4, b4) at 10%.

[0101] 15 is a flowchart showing an example of the branch point setting process in step S917. For example, when the button 1306 is selected, the screen generation unit 115 displays a screen for branch setting (scenario analysis result display screen (branch setting)) on the output device 106 (S1501). Details of the screen for branch setting will be described later. The branch point setting unit 113 sets the position of the branch point in the branch point setting data 127 (S1502). Specifically, for example, the branch point setting unit 113 accepts an input of the number of time steps and a cluster number (for example, the cluster number of the 10th step is 3, etc.) via the input device 105, and sets the position of the branch point according to the input.

[0102] The branch point setting unit 113 sets a plurality of branch destinations from the branch point in the branch point setting data 127 (S1503). Specifically, for example, the branch point setting unit 113 receives an input of a branch point already set in a time step after the branch point or a group in a final step (i.e., a group set in the group data 125) via the input device 105, and sets each of the plurality of branch destinations according to the input. Note that the plurality of branch destinations from the branch point may all be branch points already set in a time step after the branch point, or may all be groups in a final step, or may include both a branch point already set in a time step after the branch point and a group in a final step.

[0103] The branch point setting unit 113 determines whether or not an input indicating that at least one of a plurality of branch destinations from the branch point is a target of factor analysis has been received via the input device 105 (S1504). When the branch point setting unit 113 determines that a factor analysis target setting has been received (S1504: YES), the branch point setting unit 113 sets a factor analysis flag "1" for the branch destination of the factor analysis target setting in the branch point setting data 127 (S1505).

[0104] If the branch point setting unit 113 determines that the factor analysis target setting has not been received (S1504: NO), the process proceeds to step S1506. The branch point setting unit 113 determines whether a branch point setting end instruction has been received via an input to the input device 105 (S1506). If the branch point setting unit 113 determines that the branch point setting end instruction has not been received (S1506: NO), the process returns to step S1502. If the branch point setting unit 113 determines that the branch point setting end instruction has been received (S1506: YES), the branch point setting unit 113 ends the branch point setting process (the screen generating unit 115 ends the display of the screen for branch setting).

[0105] 16 is a diagram showing an example of the data configuration of the branch point setting data 127. The branch point setting data 127 includes, for example, a branch point number column 1271, a step number column 1272, a cluster number column 1273, and a branch destination column 1274. The branch point number column 1271 holds a branch point number for identifying a branch point. The branch point number is assigned by the branch point setting unit 113 in step S1502, for example.

[0106] The step number column 1272 and the cluster number column 1273 respectively hold the number of time steps and the cluster number for identifying the position of the branch point. The branch destination column 1274 holds information indicating a plurality of branch destinations from a branch point and a factor analysis flag for each branch destination. Note that, although two branch destinations occur from one branch point in the example of Fig. 16, three or more branch destinations may occur from one branch point.

[0107] Fig. 17 is a diagram showing an example of the screen configuration of a scenario analysis result display screen (branch setting) in a state in which branch setting is selected. Display of the screen in Fig. 17 starts in step S1502.

[0108] Scenario analysis result display screen 1700 includes, for example, button 1301 and buttons 1305 to 1307 similar to those of scenario analysis result display screen 1300 in FIG.

[0109] When button 1702 is selected, information about the branch point being set is saved in branch point setting data 127. When button 1708 is selected, it is possible to change the display unit of the time steps in the analysis map displayed in branch point selection area 1710. For example, when one time step is one month, when button 1708 is selected, the display unit of the time steps in the analysis map displayed in branch point selection area 1710 switches to either one month or one year.

[0110] The branch point selection area 1710 displays an analysis map similar to the analysis map display area 1313 in Fig. 13. The branch point selection area 1710 may display an analysis map similar to the analysis map display area 1313 in Fig. 14.

[0111] For example, when a pie graph is selected in the branch point selection area 1710 by an input via the input device 105, the time step and cluster corresponding to the pie graph are selected as a branch point. In Fig. 17, a pie graph 1711 is selected as a branch point, and the selected pie graph 1711 is displayed in an emphasized manner, for example, by being displayed larger than other pie graphs (or by increasing the brightness or changing the display color).

[0112] According to pie chart 1711, a scenario in which an index vector belongs to cluster number 1 at the 50th step can transition to cluster number 1 or cluster number 2 at the next step, the 51st step, and can transition to "Group A" or "Group B" at the final step. According to pie chart 1712, a scenario in which an index vector belongs to cluster number 1 at the 51st step can transition only to "Group A" at the final step, and according to pie chart 1713, a scenario in which an index vector belongs to cluster number 1 at the 51st step can transition to all groups at the final step.

[0113] In other words, depending on whether the index vector of the cluster with cluster number 1 in the 50th step transitions to cluster number 1 or cluster number 2 in the 51st step, whether the cluster can transition only to "Group A" or to all groups in the final step changes. Therefore, the pie chart 1711 can be an important branching point for the user. By displaying such an analysis map in the branching point selection area 1710, the user can easily grasp branching points that are important to the user, such as the cluster indicated by the pie chart 1711.

[0114] In the branch destination selection area 1720, for example, a group information display area 1721, a branch point display 1722, and a factor analysis target setting button 1723 are displayed. In the group information display area 1721, for example, the name of each group, details, clusters included in each group, etc. are displayed. In addition, in the group information display area 1721, the name of an index having a high rate of change (which may be an increase rate or a decrease rate) from an initial value in each group (for example, a predetermined number of indexes in descending order of the rate of change, or an index having a rate of change equal to or greater than a predetermined value) may be displayed, or the name of an index specified by input from a user in each group may be displayed. Note that the rate of change of an index from an initial value in each group is calculated, for example, by averaging the rate of change of the index from the initial value of the scenarios belonging to the group.

[0115] When a branch point is selected by selecting the pie chart in branch point selection area 1710, a branch point display 1722 is displayed at the position of the time step corresponding to the branch point in branch destination selection area 1720. In the example of FIG. 17, branch point display 1722 labeled "Step 50" is the branch point corresponding to pie chart 1711 selected in branch point selection area 1710.

[0116] When the branch point display 1722 is displayed, it becomes possible to set a branch destination (i.e., a branch point already set in a time step after the branch point, or a group in the final step) in the branch destination selection area 1720 via input to the input device 105.

[0117] 17, a group named "Group A" and a group named "Group B" are set as branch destinations for the branch point indicated by the branch point display 1722 which states "Step 50." In the branch destination selection area 1720, the branch point display 1722 and the selected branch destination are connected by a broken line.

[0118] 17, a branch point indicated by a branch point display 1722 indicating "Step 50" and a group named "Group C" are set as branch destinations for the branch point indicated by a branch point display 1722 indicating "Step 31." Displaying the branch destination selection area 1720 makes it easier for the user to set an appropriate branch destination and also makes it easier to understand the connection relationship between the branch destinations and the branch points.

[0119] When a branch destination is set for the branch point display 1722, a factor analysis target setting button 1723 is displayed for each branch destination of the branch point display 1722. When a factor analysis target setting button 1723 is selected via an input to the input device 105 (in the example of FIG. 17, a factor analysis target setting button 1723 shaded with diagonal lines is selected, and a blank factor analysis target setting button 1723 is not selected), a factor analysis flag for the branch destination of the branch point is set to "1".

[0120] 18 is a flowchart showing an example of the factor analysis process. The factor analysis unit 114 selects one branch destination that has not been selected in the factor analysis process from among the branch destinations whose factor analysis flags are "1" indicated by the branch point setting data 127 (S1801). The factor analysis unit 114 identifies a scenario whose index vector belongs to a cluster having a time step number and cluster number corresponding to the selected branch destination in the branch point setting data 127 (S1802).

[0121] The factor analysis unit 114 calculates the influence of each index for transitioning to the branch destination by performing a sensitivity analysis based on the index vector at the time step in the identified scenario, and registers the calculation result in the factor analysis result data 128 (S1803). Specifically, for example, the factor analysis unit 114 minutely changes (e.g., less than a predetermined value) the value of the index of the index vector at the time step in each of the identified scenarios for each index, and calculates the probability that the index vector of the identified scenario transitions to the branch destination corresponding to the factor analysis flag. The factor analysis unit 114 calculates the change amount (or a value obtained by substituting the change amount into a predetermined increasing function) of the probability that the index vector of the identified scenario transitions to the branch destination corresponding to the factor analysis flag when the value of the index is changed slightly and when it is not changed, as the influence of the index.

[0122] In the above example, the factor analysis unit 114 performs sensitivity analysis by slightly changing each index value of each scenario identified in step S1802, but it is also possible to select a scenario having an index vector closest to the center of gravity of the cluster from among the scenarios identified in step S1802, and perform sensitivity analysis by slightly changing each index value of the selected scenario in multiple patterns. Also, the factor analysis unit 114 may calculate the degree of influence only for important indexes.

[0123] The factor analysis unit 114 judges whether all branch destinations whose factor analysis flags are "1" indicated by the branch point setting data 127 have been selected (S1804). When the factor analysis unit 114 judges that there are any unselected branch destinations whose factor analysis flags are "1" (S1804: NO), the process returns to step S1801. When the factor analysis unit 114 judges that all branch destinations whose factor analysis flags are "1" have been selected (S1804: YES), the screen generation unit 115 displays a screen for group confirmation (scenario analysis result display screen (group confirmation)) described later on the output device 106, and ends the factor analysis process.

[0124] 19 is a diagram showing an example of the data configuration of the factor analysis result data 128. The factor analysis result data 128 includes, for example, a branch point number column 1281, a branch destination column 1282, an index ID column 1283, and an impact column 1284. The branch point number column 1281 and the branch destination column 1282 respectively hold information indicating the branch point number and branch destination of the branch point that is the subject of the factor analysis. The index ID column 1283 holds an index ID. The impact column 1284 holds the impact of each index for transitioning to the branch destination at the branch point.

[0125] 20 is a diagram showing an example of the screen configuration of a scenario analysis result display screen (group confirmation) in a state in which group confirmation is selected. A scenario analysis result display screen 2000 includes, for example, button 1301 and buttons 1305 to 1307 similar to those of the scenario analysis result display screen 1300 in FIG.

[0126] In the branch display area 2010, for example, a group information display area 1721 similar to the branch destination selection area 1720 in Fig. 17, and a branch point display 1722 are displayed, and the set branch point and the branch destination are connected by a broken line. In addition, in the branch display area 2010, a factor analysis result display button 2011 is displayed.

[0127] When any of the group information display areas 1721 is selected via an input to the input device 105, an indicator display area 2020 corresponding to the selected group is displayed. In the example of Fig. 20, the group information display area 1721 corresponding to the group named "Group A" is selected.

[0128] The indicator display area 2020 displays, for example, the name and details of the group indicated by the group data 125. The indicator display area 2020 displays a graph showing the rate of change (for example, calculated by the average of the rate of change of the important indicators in the clusters included in the group indicated by the indicator change rate data 124) from the initial value of each important indicator in the group (indicators other than the important indicators may also be displayed) for each indicator category.

[0129] When the factor analysis result display button 2011 is selected via an input to the input device 105, the indicator display area 2020 is hidden, and the factor analysis result display area shown in FIG. 21 is displayed.

[0130] Fig. 21 is a diagram showing another example of the screen configuration of the scenario analysis result display screen (group confirmation) in a state where group confirmation is selected. The scenario analysis result display screen 2000 in Fig. 21 is in a state where a factor analysis result display button 2011 to the left of the branch point of "Step 50" is selected in the branch display area 2010 of the scenario analysis result display screen 2000 in Fig. 20, and differs from Fig. 20 in that the selected factor analysis result display button 2011 is filled in black and that the indicator display area 2020 is hidden and the factor analysis result display area 2030 is displayed.

[0131] The factor analysis result display area 2030 displays the factor analysis result indicated by the factor analysis result data 128 for the branch point and branch destination corresponding to the selected factor analysis result display button 2011. Specifically, for example, the indicator category name, indicator name, influence level, and the like of a predetermined number of important indicators (which may be important indicators whose influence level exceeds a predetermined value) in descending order of influence levels of important indicators corresponding to the branch point and branch destination are displayed.

[0132] As described above, the simulation system 100 of this embodiment can generate appropriate and diverse scenarios while reducing the calculation load by performing clustering and resampling of scenarios at each time step of the simulation process.

[0133] Furthermore, although a large number of scenarios are generated in the Monte Carlo simulation, the simulation system 100 displays, for each cluster at each time step, the connection relationships between clusters to which transitions can be made at the next step, and the group transition probabilities at the final step expressed by pie charts and color combinations on the analysis map, allowing the user to easily set branching points that are important to him or her, and thereby enabling the user to perform a factor analysis related to the branching points and branching destinations.

[0134] The present invention is not limited to the above-mentioned embodiment, and various modifications are included. For example, the above-mentioned embodiment has been described in detail to clearly explain the present invention, and is not necessarily limited to those having all the configurations described. It is also possible to replace a part of the configuration of one embodiment with the configuration of another embodiment, and it is also possible to add the configuration of another embodiment to the configuration of one embodiment. It is also possible to add, delete, or replace a part of the configuration of each embodiment with another configuration.

[0135] In addition, the above-mentioned configurations, functions, processing units, processing means, etc. may be realized in part or in whole by hardware, for example, by designing them as integrated circuits. In addition, the above-mentioned configurations, functions, etc. may be realized in software by a processor interpreting and executing a program that realizes each function. Information such as the program, table, file, etc. that realizes each function can be stored in a memory, a recording device such as a hard disk or SSD (Solid State Drive), or a recording medium such as an IC card, SD card, or DVD.

[0136] In addition, the control lines and information lines shown are those that are considered necessary for the explanation, and not all control lines and information lines in the product are necessarily shown. In reality, it can be considered that almost all components are connected to each other. [Explanation of symbols]

[0137] 100 Simulation system, 101 CPU, 102 Memory, 103 Auxiliary storage device, 104 Communication device, 105 Input device, 106 Output device, 111 Simulation unit, 112 Scenario analysis unit, 113 Branch point setting unit, 114 Factor analysis unit, 115 Screen generation unit, 121 Index data, 122 Simulation input data, 123 Simulation result data, 124 Index change rate data, 125 Group data, 126 Scenario analysis map drawing data, 127 Branch point setting data, 128 Factor analysis result data

Claims

1. 1. A simulation system comprising: A processor, a memory, and a display device, The memory includes: A simulation result is stored, the simulation result indicating a plurality of scenarios each indicating an index vector of a plurality of time steps generated by a simulation, and a plurality of clusters into which the index vectors of the plurality of scenarios are classified at each of the plurality of time steps; The processor, receiving a setting for classifying clusters of final time steps of the plurality of scenarios into a plurality of groups, and performing classification based on the setting; execute a scenario analysis process for calculating, for each combination of the time step and the cluster, a transition probability that an index vector belonging to the cluster at the time step will transition to each group at the final time step, and a cluster to which the index vector belonging to the cluster at the time step can transition at the next step; The simulation system displays, on the display device, an analysis map including an indication of the transition probability for each combination of the time step and the cluster, and an indication of the clusters to which the transition can be made.

2. 2. The simulation system according to claim 1, The processor displays, on the display device, a pie chart showing the transition probabilities by area ratio.

3. 2. The simulation system according to claim 1, the memory holds information indicating a display color corresponding to each of the plurality of groups; The processor, calculating a mixed color obtained by mixing display colors corresponding to each of the plurality of groups based on a transition probability corresponding to each of the plurality of groups for each combination of the time step and the cluster; The simulation system displays, on the display device, a display indicating the transition probability using the mixed color.

4. 2. The simulation system according to claim 1, the processor accepts settings of a branch point based on a combination of the time step and the cluster and a plurality of branch destinations corresponding to the branch point by selection on a display showing the transition probability for each combination of the time step and the cluster; each of the plurality of branch destinations corresponding to the branch point is one of the plurality of groups in a final time step or one of branch points subsequent to the branch point; The processor displays, on the display device, a display indicating the branch point and a display indicating a plurality of branch destinations corresponding to the branch point.

5. The simulation system according to claim 4, The processor, Accepting a selection for the branch point and one of a plurality of branch destinations corresponding to the branch point; Executing a cause analysis process for the selected branch point and branch destination; In the factor analysis process, A sensitivity analysis is performed based on the index vector at the branching point to calculate the influence of the index for transitioning from the branching point to the branching destination; A simulation system that displays a display indicating the calculated degree of influence on the display device.

6. 2. The simulation system according to claim 1, The memory includes: An initial value of the index vector; The indicators included in the indicator vector are regarded as nodes, edges indicating causal relationships between the nodes are held, and input information defined by The processor, Executing the simulation based on the initial values ​​and the input information; storing the results of the performed simulation in the simulation result; At each time step of the simulation, generating a plurality of index vectors for the time step based on index vector values ​​for a time step immediately preceding the time step and the input information; generating the plurality of clusters at the time step by clustering the generated plurality of index vectors based on a predetermined algorithm; A simulation system that resamples a scenario by resampling the index vectors included in each of the generated clusters based on a predetermined algorithm.

7. 2. The simulation system according to claim 1, The processor, receiving a group change instruction for a cluster of a final time step of the plurality of scenarios included in the plurality of groups; execute group reconfiguration to classify final time steps of the plurality of scenarios into the plurality of groups based on the group change instruction; re-executing the scenario analysis process for the plurality of groups set by the group re-setting; A simulation system that displays the analysis map on the display device based on the results of the scenario analysis processing.

8. The simulation system according to claim 7, The memory holds an initial value of each index included in the index vector; The processor, Calculating a value indicating a rate of change from the initial value of each index included in the index vector of the plurality of clusters at the final time step; A simulation system that displays, on the display device, an indication showing the rate of change of each index in each of the plurality of clusters, and an indication for accepting the group change instruction.

9. A simulation method using a simulation system, comprising: The simulation system includes a processor, a memory, and a display device; The memory includes: A simulation result is stored, the simulation result indicating a plurality of scenarios each indicating an index vector of a plurality of time steps generated by a simulation, and a plurality of clusters into which the index vectors of the plurality of scenarios are classified at each of the plurality of time steps; The simulation method includes: The processor receives a setting for classifying clusters of final time steps of the plurality of scenarios into a plurality of groups, and executes classification based on the setting; the processor executes a scenario analysis process for calculating, for each combination of the time step and the cluster, a transition probability that an index vector belonging to the cluster at the time step will transition to each group at the final time step, and a cluster to which the index vector belonging to the cluster at the time step can transition at the next step; A simulation method in which the processor displays, on the display device, an analysis map including an indication of the transition probability for each combination of the time step and the cluster, and an indication of the cluster to which the transition can be made.

Images