JP6347751B2 - Simulation model generation support method and simulation model system - Google Patents

Description

  The present invention relates to a simulation model generation support method and a simulation model generation support system, and more specifically, a technique for interactively guiding the process from CLD to SD model generation for the user and improving the efficiency of SD model development. About.

  In order to predict changes and trends in business and social systems, system dynamics (hereinafter referred to as SD) technology exists as one of simulation technologies utilizing information systems. In order to develop a simulation model (hereinafter simply referred to as SD model) using SD, CLD (Causal Loop Diagram: causal loop) that models causal structures between event elements in the form of a directed graph network as a tool for visualizing the concept of the model Fig. CLD) is often used. In addition, development of quantitative simulation models generally involves model development and simulation execution using software dedicated to SD model development (hereinafter simply SD software) such as Powersim Studio, Vensim, and iThink / Stella (both are registered trademarks). Is called.

  As a conventional technology related to such technology, for example, a simulation model is defined based on a causal structure that is easy for humans to understand, and the simulation of system dynamics is performed for the purpose of greatly shortening the simulation program development period. The relationship between components used such as stock and flow, and the relational expression between nodes are defined by a visual model diagram SFD (Stock-Flow model, hereinafter referred to as SFD) and an I / F that assists the definition of the relational expression, and simulation is performed. A system (see Patent Document 1) has been proposed.

  Also, a rule definition method (see Non-Patent Document 1) for generating an SFD from a CLD structure based on graph theory has been proposed.

JP 2007-286777 A

“Simplied Translation of CLD's into SFD's” (J. R. Runns, 2001)

  In the development of the SD model described above, even when the relationship between variables included in the simulation is organized using CLD, the development of the SD model requires a lot of time and effort. This is because the person in charge does not cover the notation of SFD itself and the knowledge of the unique functions provided by each SD software, which are necessary for a series of development work, and the man-hours are increased. One. Furthermore, one of the causes is that it takes a lot of time for the person in charge to formulate relational expressions between variables included in the SD model one by one. On the other hand, no means for solving such problems has been presented in the prior art.

The present invention has been made in view of such circumstances, and an object of the present invention is to provide a technique for interactively guiding the process from CLD to SD model generation to improve the efficiency of SD model development.

The simulation model generation support method of the present invention that solves the above problems stores each data of a causal loop diagram showing a predetermined event to be simulated and candidate system dynamics variables corresponding to each component of the causal loop diagram An information processing apparatus including the storage device outputs candidate information related to system dynamics variables to a display device for each component of the causal loop diagram based on the data of the storage device, and system dynamics variables based on the candidate information The process of receiving the designation of the simulation execution condition and the unit of each system dynamics variable by the input device, the simulation execution condition and the unit of each system dynamics variable are applied to a predetermined determination algorithm, and the result of the determination Indicates a predetermined condition A system dynamics variable corresponding to the system dynamics variable identified as the stock as a candidate of a relational expression indicating the relationship between the process of identifying the dynamics variable as a stock variable and the relationship between the specified system dynamics variables designated as described above. A process of receiving attribute information of a flow variable that changes the value of the input by the input device, a process of receiving information of predetermined data representing the state of each system dynamics variable by the input device, and each corresponding system dynamics between the predetermined system dynamics variables The attribute information of the variable and the predetermined data are applied to a predetermined algorithm to estimate a candidate for a relational expression between the corresponding system dynamics variables and output to the display device, and between the corresponding system dynamics variables based on the relational expression candidates Specify the relational expression on the input device. The system dynamics model definition information including at least a relational expression between the process and the system dynamics variable corresponding to each component obtained by each process and the predetermined system dynamics variable. And a process for converting the system dynamics model development program specification.

Further, the simulation model generation support system of the present invention stores a storage device storing each data of a causal loop diagram showing a predetermined event to be simulated and candidate system dynamics variables corresponding to each component of the causal loop diagram And based on the data of the storage device, for each component of the causal loop diagram, candidate information related to system dynamics variables is output to a display device, specification of system dynamics variables based on the candidate information, simulation execution conditions, and each Processing for accepting an input of a unit of system dynamics variable by an input device, and applying the simulation execution condition and the unit of each system dynamics variable to a predetermined determination algorithm, and a system dynamics variable indicating a predetermined condition as a result of the determination stock The value of the corresponding system dynamics variable is changed for the system dynamics variable identified as the stock as a candidate of a relational expression indicating the relationship between the processing identified as the variable and the predetermined system dynamics variable designated as described above Processing for receiving attribute information of flow variables by the input device, processing for receiving information of predetermined data representing the state of each system dynamics variable by the input device, and the attribute information of each corresponding system dynamics variable between the predetermined system dynamics variables And applying the predetermined data to a predetermined algorithm to estimate a candidate for a relational expression between corresponding system dynamics variables and outputting it to a display device, and specifying the relational expression between the corresponding system dynamics variables based on the candidate for the relational expression Processes accepted by the input device and each of the above processes The obtained system dynamics model definition information including at least the system dynamics variable corresponding to each component and the relational expression between the predetermined system dynamics variables is converted into the specification of the system dynamics model development program by a predetermined algorithm. And an arithmetic unit that executes processing for conversion.

  ADVANTAGE OF THE INVENTION According to this invention, the process from CLD to SD model generation can be induced | guided | derived interactively for a user, and efficiency improvement of SD model development can be aimed at.

It is a figure which shows the structural example of the simulation model production | generation assistance system in this embodiment. It is a figure which shows the example of CLD in this embodiment. It is a figure which shows the node data file in the CLD management data of this embodiment. It is a figure which shows the link data file in the CLD management data of this embodiment. It is a figure which shows the structure of SD variable name substitution data in this embodiment. It is a figure which shows the example 1 of a procedure of the simulation model production | generation assistance method of this embodiment. It is a figure which shows the example 2 of a procedure of the simulation model production | generation assistance method of this embodiment. It is a figure which shows the example 3 of a procedure of the simulation model production | generation assistance method of this embodiment. It is a figure which shows the initial screen of SD model definition in this embodiment. It is a figure which shows the definition screen of the SD variable name in the SD model definition in this embodiment. It is a figure which shows the definition screen of the unit of the variable in SD model definition in this embodiment. It is a figure which shows the definition screen of the flow variable in SD model definition in this embodiment. It is a figure which shows the definition screen of the calculation formula of the variable in SD model definition in this embodiment. It is a figure which shows the function conversion data which convert between each function of the software only for SD in this embodiment, and general spreadsheet software.

--- System configuration ---

  Embodiments of the present invention will be described below in detail with reference to the drawings. FIG. 1 is a diagram illustrating a configuration example of a simulation model generation support system 100 according to the present embodiment. A simulation model generation support system 100 shown in FIG. 1 is a computer system that interactively guides the process from CLD to SD model generation to improve the efficiency of SD model development.

  Such a simulation model generation support system 100 has a hardware configuration similar to that of a general computer device, and includes a storage device 101 and a RAM that are configured by appropriate nonvolatile storage elements such as an SSD (Solid State Drive) and a hard disk drive. A memory 103 composed of volatile storage elements, etc., a CPU 102 that performs overall control of the device itself by reading the program 102 held in the storage device 101 into the memory 103 and executing it, and performs various determinations, computations, and control processing, etc. A calculation device 104, an input device 105 that accepts key input and voice input from a user in charge of simulation model generation, and a display device 106 such as a display that displays processing data. The storage device 101 stores at least a program 102 for implementing functions necessary for the simulation model generation support system 100 of the present embodiment and other data 120, 125, 126, 127, and the like. .

  Next, functions provided in the simulation model generation support system 100 of this embodiment will be described. The functional units 2011 to 2014 and 2019 described below are implemented by executing a program 102 included in the simulation model generation support system 100, for example.

  The simulation model generation support system 100 includes a CLD reading unit 2011, a logic definition rule management unit 2012, a data correlation calculation unit 2013, an SD dedicated software conversion unit 2014, and an SD dedicated software execution unit 2019.

  Among these, the CLD reading unit 2011 reads attribute information such as nodes and links from the data of the CLD 120 input via the input device 105 or an appropriate medium reader, and stores this in the CLD management data 125 of the CLD data management unit 2015. To do. Details of the CLD management data 125 will be described later.

The logic definition rule management unit 2012 manages procedures and support information for prompting the definition of the SD model while using the logic definition file held by the logic definition file management unit 2016. In other words, the logic definition rule management unit 2012 outputs candidate information related to SD variables to the display device 106 for the nodes (each component) of the above-described CLD 120 (causal loop diagram data) indicating a predetermined event to be simulated. This corresponds to the function of accepting designation of the SD variable based on the candidate information from the user by the input device 105. In this case, in the process of accepting the designation of the SD variable, the simulation execution condition and the input of each SD variable unit are accepted.

  In addition, the data correlation calculation unit 2013 is a cause-side variable that changes the SD variable for each SD variable specified for the node of the CLD 120 based on the data held by the logic definition file management unit 2016 or the like. Are retrieved from the link data management file 502, the correlation between the cause-side variable and the result-side variable is analyzed, and the relational expression between the variables is estimated. In other words, the data correlation calculation unit 2013 estimates a relational expression candidate indicating the relationship between predetermined SD variables using a predetermined algorithm and outputs the estimated result to the display device 106. Between the corresponding SD variables based on the relational expression candidate, This corresponds to the function of receiving the designation of the relational expression from the user by the input device 105. In this case, the simulation execution condition and the unit of each SD variable are applied to a predetermined determination algorithm, the SD variable whose determination result indicates the predetermined condition is identified as a stock variable, and the relational expression candidate is estimated. Processing for accepting flow variable attribute information for changing the value of the corresponding SD variable by the input device 105 for the SD variable identified as stock, and processing for accepting information on the predetermined data representing the state of each SD variable by the input device 105 In addition, the attribute information and the predetermined data of each corresponding SD variable are applied to a predetermined algorithm between predetermined SD variables, and a relational expression candidate between the corresponding SD variables is estimated.

  In addition, in the process of estimating the above-mentioned relational expression candidates, a process for receiving information on predetermined data representing the state of each SD variable by the input device 105, and a predetermined algorithm for the predetermined data of each SD variable between predetermined SD variables. When the relationship between the calculated SD and the relationship between the predetermined data of each SD variable is less than a predetermined standard, Information for prompting the setting is output to the display device 106.

  The SD dedicated software conversion unit 2014 converts the logic definition data held by the above-described logic definition file management unit 2016 or the like into specifications (eg, notation) of the SD dedicated software. In other words, the SD dedicated software conversion unit 2014 includes an SD model including at least the SD variable corresponding to each component (node) of the CLD 120 and the relational expression between the predetermined SD variables obtained in each process so far. This function corresponds to the function of converting the definition information (logic definition file or the like) into the above-mentioned SD-dedicated software specifications by a predetermined algorithm.

  The SD dedicated software execution unit 2019 executes the SD dedicated software using the data held by the SD dedicated software data management unit 2020.

The CLD reading unit 2011, the logic definition rule management unit 2012, the data correlation calculation unit 2013, the SD dedicated software conversion unit 2014, the CLD data management unit 2015, the logic definition file management unit 2016, and the SD variable name replacement data management unit 2017. The variable state data management unit 2018 may be mounted on general spreadsheet software.
--- Data structure example ---

  Next, an example of the structure of data used by the simulation model generation support system 100 of this embodiment will be described. First, the CLD 120 corresponding to the event to be simulated will be described. FIG. 2 is a diagram illustrating an example of the CLD 120 used in the present embodiment. In general, the CLD 120 is a diagram showing a causal structure of an evaluation object in a business or social system. W. It is used in the system simulation method and system dynamics proposed by Forester. The structure is composed of nodes indicating the evaluation target components and links indicating the causal relationship between the components. These causal relationships between components are categorized by two scales: “whether the cause and result have the same direction of increase or decrease in value” or “the presence or absence of a time gap in which the cause is reflected in the result.” Accordingly, there are multiple types of links.

  For example, if a link connection source component increases (or decreases) and the link connection destination component also increases (or decreases), the corresponding link is a positive correlation link. In addition, a link indicating the opposite causal relationship is an inverse correlation link, and a link that is affected with a time delay is called a positive correlation link (or reverse link) with a delay.

  In the case of the CLD 120 illustrated in FIG. 2, the positive correlation link is a link 4011 that connects the “experience value” node and the “quality” node, and the positive correlation link that has a delay connects the “quality” node and the “order” node. The link 4012 and the reverse link with a delay correspond to the link 4013 connecting the “equipment deterioration level” node and the “quality” node.

  What is generated from the CLD 120 by the CLD reading unit 2011 of the simulation model generation support system 100 and managed as the CLD management data 125 by the CLD data management unit 2015 is illustrated in the node data management file 501 illustrated in FIG. 3 and FIG. This is the link data management file 502.

  Among these, the node data management file 501 shown in FIG. 3 is composed of a collection of records in which the node ID 5011 and the node name 5012 of each node are associated with each other. The link data management file 502 shown in FIG. 4 includes a link ID 5021 of each link, a link type 5022, and a set of records in which each data of the start node 5023 and the end node 5024 of the corresponding link is associated. .

  The link type shown in the above-described link data management file 501 indicates the type of link already described. For example, the positive correlation link is DN (Direct, Normal), the negative correlation is ON (Opposite, Normal), the positive correlation link with delay is DD (Direct, Delay), and the negative correlation link with delay is ON (Opposite, Delay). Identification information corresponding to the link type is set.

Next, FIG. 5 is a diagram showing the structure of the SD variable name replacement data 126 in the present embodiment. The SD variable name replacement data 126 is data managed by the SD variable name replacement data management unit 2017, and is a table that defines the relationship between the nodes of the CLD 120 described above and the corresponding candidates for variables used in the SD model. The SD variable name replacement data 126 in this embodiment is composed of a set of records in which at least the CLD node name 110101 and the SD variable name candidate 110102 are associated with each other.
--- Processing flow ---

  Hereinafter, the actual procedure of the simulation model generation support method in the present embodiment will be described with reference to the drawings. Various operations corresponding to the simulation model generation support method described below are realized by a program 102 that the simulation model generation support system 100 reads into the memory 103 or the like and executes. And this program is comprised from the code | cord | chord for performing the various operation | movement demonstrated below. Here, the simulation model generation support system 100 is referred to as a system 100 for ease of explanation.

FIG. 6 is a flowchart showing a processing procedure example 1 of the simulation model generation support method in the present embodiment, and specifically shows a basic processing flow in the present embodiment. In this case, the CLD reading unit 2011 of the system 100 acquires the data of the CLD 120 input by the user from the input device 105 or from the storage device 101 or an appropriate medium reader, and the corresponding CLD 120 is obtained from the data of the CLD 120 obtained here. Are extracted, the above-described node data management file 501 and link data management file 502 are generated, and stored in the storage device 101 as CLD management data 125 (step s1).
01). The CLD 120 obtained here is the basis of the SD model. As a technique for extracting node and link information from the CLD 120, an existing technique may be adopted.

  Next, the logic definition rule management unit 2012 of the system 100 reads the format of the SD model definition screen held in advance by the logic definition file management unit 2016 of the storage device 101, and the above-described setting field of the CLD node in this format is described above. Information on each node extracted with respect to the CLD 120 in step s101 is set to generate an SD model definition screen 601 and this is output to the display device 105 (step s102).

  The initial screen of the SD model definition screen 601 output in step s102 will be described with reference to FIG. The SD model definition screen 601 includes a simulation execution condition 602 and a logic definition sheet 603. Among these, the simulation execution condition 602 includes a simulation execution condition item 6021, a value 6022 of the item, and a unit 6023 thereof. In addition, the logic definition sheet 603 includes an SD variable indicating a node ID 6031 and a CLD node name 6032 set by reading from the node data management file 501 obtained for the corresponding CLD 120 and a variable name on the SD model corresponding to the corresponding CLD node. It consists of a name 60, its unit 6034, and type 6035. Here, the numerator of the variable unit 6034 is a unit of the amount of the corresponding SD variable, and the denominator corresponds to a unit period (time span) when the corresponding variable is handled. The variable type 6035 corresponds to information such as stock and flow handled in the SD model.

  Next, the logic definition rule management unit 2012 of the system 100 receives the user designation of the relational expression between the nodes on the SD model definition screen 601 described above from the input device 105 (step s103). Here, details of the step s103 will be further described with reference to FIG.

  FIG. 7 is a diagram showing a procedure example 2 of the simulation model generation support method of the present embodiment, and more specifically, a diagram showing a relational expression setting flow. In this case, the logic definition rule management unit 2012 of the system 100 receives the input of the SD variable name from the input device 105 in the SD variable name 6033 column of the SD model definition screen 601 as shown in FIG. 10 (step s301). The user input work here corresponds to the work of defining the name of the SD variable in the SD variable name 6033 column for replacing the quantitative index on the SD model with respect to the CLD node name 6032 on the SD model definition screen 601. However, depending on the CLD node, it is difficult to replace the quantitative index, and depending on the user's skill or the like, it is difficult to deal with it. Therefore, in this step s301, the logic definition rule management unit 2012 of the system 100 uses the CLD node name 6032 as a key. The SD variable name substitution data 126 managed by the SD variable name substitution data management unit 2017 is searched for, and the candidate SD variable names are displayed as a candidate list 60331 in the corresponding SD variable name column so as to be selectable. Accept selection.

  In the example of the SD model definition screen 601 in FIG. 10, for the CLD node name “quality”, as a candidate list 60331 for the SD variable name 6033, “returned goods rate”, “delivery date compliance rate”, “number of complaints”, “unsatisfied” SD variable candidate groups such as “non-defective product rate” and “customer evaluation score” are displayed in a pull-down format, and the user has selected “number of complaints” as an SD variable. When the user directly inputs an SD variable name not included in the above-described candidate list 60331 using the input device 105, the logic definition rule management unit 2012 of the system 100 accepts the SD variable name and sets it in the SD variable name 6033 column. At the same time, the corresponding SD variable name is added to the SD variable name candidate 110102 column in association with the corresponding CLD node name 110101 in the SD variable name replacement data 126.

In addition, the logic definition rule management unit 2012 of the system 100 inputs an input of each value 6022 and each unit 6023 of the simulation period and time step and each variable unit 6034 as the simulation execution condition 602 on the SD model definition screen 601 as an input device. Accept from 105 (step s302). In the example of the SD model definition screen 601 in FIG. 11, the simulation period value 6022 is “5”, the unit 6023 is “year”, the time step value is “1”, and the unit 6023 is “monthly”. Indicates the specified status. Similarly, as the unit 6034 of each variable, for example, regarding the SD variable name “number of complaints”, the unit 6034 indicates a status designated as “numerator“ case ”, denominator“ month ”and the like.

  Next, the logic definition rule management unit 2012 of the system 100 identifies a stock variable from the SD variables specified in association with the CLD node on the SD model definition screen 601 (step s303). In this case, the logic definition rule management unit 2012 of the system 100 uses the denominator of each variable unit 6034 and the time step value 6022 and unit 6024 based on the simulation execution condition 602 and each variable unit 6034 obtained in step s302. And each variable unit 6034> time step value 6022 and unit 6024, or each variable unit 6034 numerator <>% and denominator = Null, the SD variable is identified as stock. The fact is displayed in the cell of the type 6035 in the logic definition 603. FIG. 11 shows an example in which “stock received number of orders” and “(product A) cumulative production amount” among the SD variables are identified as stock, and “stock” is displayed in the respective type 6035 column. . In system dynamics, for such stock variables, at least one of an inflow variable that is an input to the stock variable and an outflow variable that is an output is required.

  Therefore, the logic definition rule management unit 2012 sets the lines 901 and 902 for defining the inflow variable and the outflow variable for changing the value of the stock variable identified in the above step s303 as the logic of the SD model definition screen 601. It is added under the record of the corresponding variable in the definition 603 (step s304). FIG. 12 illustrates the state of the logic definition 603 after adding such rows 901 and 902. In this state, the user is prompted to input the SD variable name 6033 and the SD variable unit 6034 regarding the inflow variable and the outflow variable. Note that both the inflow variable and the outflow variable are not necessarily required. Therefore, it is preferable that the logic definition rule management unit 2012 also displays an interface that receives a deletion instruction for the corresponding line on the screen 601. In addition, for SD variables that are not identified as either stock or flow, the logic definition rule management unit 2012 may display “auxiliary variable” by setting it in the corresponding line.

  Subsequently, the logic definition rule management unit 2012 accepts user input from the input device 105 in each of the rows 901 and 902 for the above-described flow variables (step s305). In the example of the screen 1100 in FIG. 13, the type “In-flow” is set for the SD variable “Number of monthly orders” in the logic definition 603, and the type “Out-flow” is set for the SD variable “Year-end closing process”. The situation is illustrated. In step s305, the logic definition rule management unit 2012 additionally displays each cell of the data association 6036 and the calculation logic 6037 on the logic definition sheet 603.

  Subsequently, the data correlation calculation unit 2013 obtains the file name of the data representing the time-series state change of each SD variable (transition data such as a log of values accumulated for a predetermined period for the event corresponding to the SD variable). This is accepted by the input device 105 (step s306). Further, the data correlation calculation unit 2013 creates a candidate list of calculation formulas for each SD variable (step s307), outputs this to the display device 106 (step s308), and calculates each SD variable from the candidate list. The user designation of the expression is accepted by the input device 105 (step s309).

  Here, the details of steps s306 to s309 described above will be described with reference to FIG. FIG. 13 is a calculation logic definition screen 1100 for calculating SD variables. In this case, the data correlation calculation unit 2013 receives the file name including the data indicating the state change of each SD variable in the cell of the data association 6036 additionally displayed in the above step s305 by the input device 105, and receives this. The file name 60361 is set, and the start time and end time of the target data in the corresponding file are received by the input device 105 and set to the start time 60362 and the end time 60363, respectively. Each data is stored in the variable state data management unit 2018. In the example of FIG. 13, regarding the SD variable “number of complaints”, the file name “number of customer complaints.dat” is set in the data association 6036 and the target is the range of data to be considered in the file. This shows a situation in which time (start) “$ B $ 6” and target time (end) “$ BI $ 6” are set.

  Subsequently, for each SD variable, the data correlation calculation unit 2013 searches the link data management file 502 for a cause-side variable that changes the SD variable, analyzes the correlation between the cause-side variable and the result-side variable, and calculates A pop-up screen 1001 or 1002 of the expression candidate list is displayed. For example, the data correlation calculation unit 2013 searches the end point node 5024 of the link data management file 502 for the CLD node ID “N005” corresponding to the SD variable “number of failures”, and the starting point node, that is, the CLD node ID on the cause side Is specified as “N003”. Further, the data correlation calculation unit 2013 searches the logic definition 603 for the SD variable corresponding to the node “N003” as “monthly total man-hour”, and specifies this as a cause-side variable.

  That is, the data correlation calculation unit 2013 specifies the above-mentioned “number of failures” as the result-side variable y and “monthly total man-hour” as the cause-side variable x, and in the file indicated by each file name 60361 that is exhausted to the variable y and the variable x. The correlation between y and x is analyzed based on the data (range of target times 60362 and 60363). For the correlation analysis algorithm, existing techniques that have existed in the past may be employed. In this process, the process differs depending on whether the SD variable that is the result-side variable y is a flow or auxiliary variable or a stock variable. In the following, each process is shown separately.

i) When the SD variable as the result side variable y is a flow or auxiliary variable:

  When the SD variable that is the result-side variable y is a flow, as illustrated in FIG. 13, the calculation formula candidate list 1001 includes at least a scatter diagram 10011 of x and y, a calculation formula candidate 10012, and its fitness 10013. Composed. The calculation formula candidate 10012 is obtained by calculating the relationship between x and y for the correlation pattern prepared in advance using the above-described correlation analysis algorithm. Here, the correlation pattern indicates a linear function, a polynomial function, a step function, or the like. At this time, there may be two or more cause-side variables x with respect to the result-side variable y. Therefore, the data correlation calculation unit 2013 displays a calculation formula candidate list for each SD variable, accepts a selection input from the user by the input device 105, sets the selected calculation formula in the calculation logic 6037, and outputs a logic definition file. The data is stored in the management unit 2016. At this time, the correlation patterns may be displayed in descending order of fitness.

  In the above processing, the data correlation calculation unit 2013 compares the suitability of each pattern with a predetermined reference value (s401) as shown in the flow of FIG. If it is low (s402: y), it is preferable to display a message prompting the user to change the contents of the data association 6036 (s403) as no candidates, and return the processing to step s306 in the flow of FIG. It is.

ii) When the SD variable as the result side variable y is stock:

  On the other hand, when the SD variable as the result-side variable y is stock, as shown in FIG. 13, the calculation formula candidate list 1002 includes at least the initial value input cell 10021, the calculation formula 10022 of x and y, and the coefficient of the calculation formula. It consists of a candidate list 10027 to be set. The data correlation calculation unit 2013 displays a calculation formula candidate list 1002 for each SD variable, accepts a selection input from the user by the input device 105, sets the selected calculation formula in the calculation logic 6037, and manages the logic definition file. Stored in the unit 2016. At this time, the candidate list 10027 may be provided with a slide-type indicator and displayed in conjunction with the shape of the graph.

  Returning to the description of the flow in FIG. Subsequently, the data correlation calculation unit 2013 of the system 100 repeats the above-described step s103 until the above-described definition regarding all SD variables is completed (step s104: Y).

  On the other hand, the SD software conversion unit 2014 of the system 100 converts the contents of the logic definition file into a function of the SD development software when it is determined that the definition regarding all the SD variables is completed in the above step s104 (step s104: Y). (Step s105), and the converted contents are exported to the SD dedicated software (Step s106).

  Here, step s105 described above will be described in detail. The conversion process to the SD dedicated software function is performed as follows. Here, as an example, a situation is assumed in which Powersim Studio (registered trademark) is used as SD dedicated software and Excel (registered trademark) is used as spreadsheet software. FIG. 14 is a diagram showing function conversion data 127 for converting between functions of the SD dedicated software and general spreadsheet software in the present embodiment. The difference between the function of the SD software and the function of the spreadsheet software, as exemplified in FIG. 14, is “Delay function”, “Condition function”, “Control function”, “ "GRAPH function". Other functions have the same function, except that they include grammar and time concepts.

i) For functions that are common to both spreadsheet software and SD software:

The SD software conversion unit 2014 applies each function included in the logic definition file to the table 1271 of the function replacement data 127 to specify a function defined by the table 1271 and uses only the SD associated with the specified function. The software function is searched in the table 1272 and converted into the specification of the corresponding function. In the example of FIG. 14, if the “NPV” function in the spreadsheet software is NPV (r, p1, p2,..., PN) (r: discount rate, p: payment amount), Powersim In this case, the order of the first argument (r) and the subsequent arguments (p1, p2,..., PN) is changed, and the subsequent arguments are the nth year NPV (p1 , R), and t + 1 year NPV (p2, r),.
ii) For functions not in spreadsheet software

The SD software conversion unit 2014 displays on the “SD model definition screen” a cell for inputting information that is insufficient for the function of the spreadsheet software. For example, in the case of “DELAYPPL” which is one of the “Delay functions”, this function is a function having (input, delaytime) as an argument (input enters after delaytime). Therefore, when the SD software conversion unit 2014 identifies from the information of the CLD management data 125 that it is connected to a link with a delay, the SD software conversion unit 2014 inputs not only “input” but also “delaytime”. To appear automatically.
iii) For functions only available in spreadsheet software

  In this case, use is prohibited because it is not necessary for the implementation of the SD dedicated software.

  Although the best mode for carrying out the present invention has been specifically described above, the present invention is not limited to this, and various modifications can be made without departing from the scope of the invention.

  According to the present embodiment, the SD model developer can significantly reduce the effort and time for SD model development after CLD creation as compared with the prior art. Therefore, the process from CLD to SD model generation can be guided interactively for the user, and the efficiency of SD model development can be improved.

  At least the following will be clarified by the description of the present specification. That is, in the simulation model generation support method of the present embodiment, when the information processing apparatus receives the designation of the system dynamics variable, the information processing apparatus further executes a process of accepting a simulation execution condition and a unit of each system dynamics variable, Applying the simulation execution condition and the unit of each system dynamics variable to a predetermined determination algorithm, executing a process for identifying a system dynamics variable whose result of the determination indicates a predetermined condition as a stock variable, and In the process of estimating a candidate, for the system dynamics variable identified as the stock, the process of receiving attribute information of the flow variable that changes the value of the corresponding system dynamics variable with the input device, and the state of each system dynamics variable A process of accepting information of predetermined data to be expressed by an input device, and a relationship between the corresponding system dynamics variables by applying the attribute information and the predetermined data of the corresponding system dynamics variables to the predetermined algorithm between the predetermined system dynamics variables A process for estimating an expression candidate may be executed.

  According to this, the relationship between the system dynamics variables based on the stock and the flow can be accurately defined and grasped, and the relational expression calculation between the system dynamics variables can be efficiently and accurately performed. As a result, the SD model developer can significantly reduce the time and effort of developing the SD model after the creation of the CLD. Therefore, the process from CLD to SD model generation can be guided interactively for the user, and the efficiency of SD model development can be improved.

  Further, in the simulation model generation support method of the present embodiment, when the information processing apparatus estimates the relational expression candidates, a process of receiving information on predetermined data representing a state of each system dynamics variable by an input device; Applying the predetermined data of each corresponding system dynamics variable between the predetermined system dynamics variables to the predetermined algorithm to calculate a relational expression between the corresponding system dynamics variables, and calculating the relational expression between the calculated system dynamics variable and each system dynamics variable When the degree of conformity with the relationship between each predetermined data is equal to or less than a predetermined standard, a process of outputting information for prompting resetting of the information of the predetermined data to the display device may be further executed.

  According to this, the relational expression between the system dynamics variables can be estimated with good accuracy, and as a result, the SD model developer can significantly reduce the time and effort of developing the SD model after creating the CLD as compared with the conventional case. Therefore, the process from CLD to SD model generation can be guided interactively for the user, and the efficiency of SD model development can be improved.

Further, in the simulation model generation support system of the present embodiment, the arithmetic device further executes a process of receiving an input of a simulation execution condition and a unit of each system dynamics variable when receiving the specification of the system dynamics variable, Applying the simulation execution condition and the unit of each system dynamics variable to a predetermined determination algorithm, executing a process of identifying a system dynamics variable whose determination result indicates the predetermined condition as a stock variable, and candidates for the relational expression In the process of estimating the system dynamics variable identified as the stock, the process of receiving the attribute information of the flow variable that changes the value of the corresponding system dynamics variable by the input device, and the status of each system dynamics variable are displayed. A process of accepting information on predetermined data by an input device, and a relational expression between corresponding system dynamics variables by applying the attribute information and the predetermined data of each corresponding system dynamics variable between the predetermined system dynamics variables to the predetermined algorithm It is good also as what performs the process which estimates this candidate.

  Further, in the simulation model generation support system of the present embodiment, the calculation device, in the process of estimating the candidate of the relational expression, a process of accepting information of predetermined data representing the state of each system dynamics variable by the input device, Applying the predetermined data of the corresponding system dynamics variable to the predetermined algorithm between predetermined system dynamics variables, calculating a relational expression between the corresponding system dynamics variables, and calculating the calculated relational expression and each of the system dynamics variables When the degree of conformity with the relationship between the predetermined data is equal to or less than the predetermined reference, a process of outputting information for prompting resetting of the information of the predetermined data to the display device may be further executed.

100 simulation model generation support system 101 storage device 102 program 103 memory 104 arithmetic device 105 input device 106 display device 120 CLD (causal loop diagram)
125 CLD management data (causal loop diagram data)
126 SD variable name replacement data (system dynamics variable candidate data)
501 Node data management file 502 Link data management file

Claims (4)

An information processing apparatus provided with a storage device storing each data of a causal loop diagram indicating a predetermined event to be simulated, and system dynamics variable candidates corresponding to each component of the causal loop diagram,
Based on the data in the storage device, for each component of the causal loop diagram, candidate information related to system dynamics variables is output to a display device, specification of system dynamics variables based on the candidate information, simulation execution conditions, and system dynamics A process of accepting an input of a variable unit by an input device;
Applying the simulation execution condition and the unit of each system dynamics variable to a predetermined determination algorithm, and identifying the system dynamics variable whose result of the determination indicates the predetermined condition as a stock variable;
As a candidate of a relational expression indicating the relationship between the specified system dynamic variables for which the designation has been made, with respect to the system dynamic variables identified as the stock, the attribute information of the flow variable that changes the value of the corresponding system dynamic variable is input device A process of accepting the information of the system dynamics variable, a process of accepting information of the predetermined data representing the state of each system dynamics variable by the input device, and the attribute information and the predetermined data of each corresponding system dynamics variable between the predetermined system dynamics variables with a predetermined algorithm A process of estimating the relational expression candidates between the corresponding system dynamics variables and outputting them to the display device, and receiving the designation of the relational expressions between the corresponding system dynamics variables based on the relational expression candidates at the input device;
For system dynamics model development, a system dynamics model definition information including at least a system dynamics variable corresponding to each component and a relational expression between the predetermined system dynamics variables obtained in each processing is defined by a predetermined algorithm. The process of converting to the specification of the program,
A simulation model generation support method characterized by executing The information processing apparatus is
In the process of estimating the relational expression candidates, the process of accepting information of predetermined data representing the state of each system dynamics variable by the input device, and the predetermined data of each corresponding system dynamics variable between the predetermined system dynamics variables When applied to a predetermined algorithm to calculate a relational expression between corresponding system dynamics variables, and the degree of conformity between the calculated relational expression and the relation between each predetermined data of each system dynamics variable is below a predetermined reference, Further executing a process of outputting information prompting resetting of the information of the predetermined data to the display device;
The simulation model generation support method according to claim 1. A storage device storing each data of a causal loop diagram indicating a predetermined event to be simulated, and candidate system dynamics variables corresponding to each component of the causal loop diagram;
Based on the data in the storage device, for each component of the causal loop diagram, candidate information related to system dynamics variables is output to a display device, specification of system dynamics variables based on the candidate information, simulation execution conditions, and system dynamics A process of receiving an input of a unit of a variable by an input device, and applying the simulation execution condition and the unit of each system dynamics variable to a predetermined determination algorithm, and a system dynamics variable whose result of the determination indicates a predetermined condition as stock As a candidate of a relational expression indicating the relationship between the specified system dynamics variable and the specified system dynamics variable specified as described above, the value of the corresponding system dynamics variable is changed for the system dynamics variable identified as the stock. Processing for receiving attribute information of flow variables by the input device, processing for receiving information of predetermined data representing the state of each system dynamics variable by the input device, and the attribute information of each corresponding system dynamics variable between the predetermined system dynamics variables And applying the predetermined data to a predetermined algorithm to estimate a candidate for a relational expression between corresponding system dynamics variables and outputting it to a display device, and specifying the relational expression between the corresponding system dynamics variables based on the candidate for the relational expression System dynamics model definition information including at least a process received by the input device, a system dynamics variable corresponding to each component obtained in each process, and a relational expression between the predetermined system dynamics variables, a predetermined algorithm System dynamics model An arithmetic unit for executing a process of converting the specification of Hatsuyo program,
A simulation model generation support system comprising: The arithmetic unit is
In the process of estimating the relational expression candidates, the process of accepting information of predetermined data representing the state of each system dynamics variable by the input device, and the predetermined data of each corresponding system dynamics variable between the predetermined system dynamics variables When applied to a predetermined algorithm to calculate a relational expression between corresponding system dynamics variables, and the degree of conformity between the calculated relational expression and the relation between each predetermined data of each system dynamics variable is below a predetermined reference, A process of outputting information for prompting resetting of the information of the predetermined data to the display device;
The simulation model generation support system according to claim 3.