System and method for generating a fault tree

Abstract

To narrow down and extract cause candidate of failure or the like to a reasonable number in a fault tree analysis.SOLUTION: A fault tree generation system 100 comprises a processor and a memory device. The storage device has a mathematical expression DB for storing a mathematical expression relating to a physical phenomenon, a variable DB for storing a variable extracted from a technical document, a variable knowledge DB for storing knowledge relating to the variable, and a variable relation DB for storing a variable relation representing a relation between a variable on the left side and a variable on the right side of the mathematical expression relating to the physical phenomenon. The processor develops: the causal relationship of a failure event; generates a mathematical formula-based causal model which is a model expressing the causal relationship; generates and outputs a failure tree combining the causal relationship related to an arbitrary failure event; generates a variable list of a mathematical formula related to the physical phenomenon, the variable, and a variable condition which is a numerical value that the variable can take; extracts a variable relation between a variable on the left side and a variable on the right side in the mathematical formula; and reduces the number of branches of the failure tree when generating the failure tree.SELECTED DRAWING: Figure 1

Description

[Technical Field]

[0001] The present invention relates to computer techniques for generating fault trees. [Background technology]

[0002] When a breakdown or malfunction (hereinafter referred to as "failure, etc.") occurs in operating equipment or machinery (hereinafter referred to as "equipment, etc."), it is essential to accurately understand the situation and state and determine the correct cause in order to properly deal with the breakdown, etc. One of the main analysis methods used for such purposes is Fault Tree Analysis (FTA). Fault tree analysis is a method of systematically organizing and analyzing failure modes, which indicate the situation and state of the breakdown, and failure mechanisms, which indicate the causes of the breakdown, for the equipment, etc., using a tree-structured model.

[0003] Nowadays, fault tree analysis and its related technologies are widely used in a variety of situations, such as comprehensively predicting potential faults in new equipment during the design and manufacturing stages, and investigating the causes of faults that have occurred in existing equipment (e.g., Patent Document 1). [Prior art documents] [Patent documents]

[0004] [Patent Document 1] Japanese Patent Application Publication No. 7-262019 Summary of the Invention [Problem to be solved by the invention]

[0005] It is desirable for fault tree analysis to be able to extract candidate causes of failures, etc. (hereinafter referred to as "candidate causes") for the equipment in question without any omissions or overlaps (Mutually Exclusive, Collectively Exhaustive; MECE). If the extracted candidate causes are incomplete, various problems may arise, such as the investigation work being prolonged because the correct cause cannot be identified, or there may be omissions in the investigation work during the design and manufacturing stages. Also, if the extracted candidate causes are duplicated, the same candidate causes will be examined repeatedly, resulting in waste in the investigation work.

[0006] On the other hand, it is preferable to extract an appropriate number of candidate causes in a fault tree analysis. If a large number of candidate causes are extracted, the task of identifying the correct cause becomes complicated, and the investigation into the cause of the failure, etc. takes a long time. This has become a particularly urgent issue in the industrial world today, where the analysis of failures, etc. is becoming more difficult due to the increase in complexly configured facilities, etc.

[0007] However, in fault tree analysis, it is difficult to extract all possible causes of failures and other issues without omissions or duplications, while narrowing them down to an appropriate number using existing techniques, and the development of new techniques has been awaited.

[0008] In view of the above-mentioned problems, an object of the present invention is to provide a technique that can extract candidate causes of failures and the like without omission or duplication in fault tree analysis, while narrowing down the number of candidates to an appropriate number. [Means for solving the problem]

[0009] A fault tree generation system according to the present invention is a system for generating a fault tree and includes at least a processor and a storage device. The storage device includes at least a mathematical formula database for storing mathematical formulas related to physical phenomena, a variable database for storing variables extracted from technical documents, a variable knowledge database for storing variable knowledge related to variables, and a variable relation database for storing variable relations representing the relationship between variables on the left side of mathematical formulas related to the physical phenomena and variables on the right side of the mathematical formulas. The processor develops the causal relationships of the malfunction events based on the mathematical formula stored in the mathematical formula database, and generates a mathematical formula-based causal model that is a model representing the causal relationships; based on the generated mathematical formula-based causal model, it generates and outputs a fault tree that combines causal relationships related to any malfunction events from the developed causal relationships of the malfunction events; based on the variable database and the variable knowledge database, it generates a variable list of the mathematical formula related to the physical phenomenon, the variables, and variable conditions that are possible numerical values ​​for the variables; extracts variable relationships between variables on the left side and variables on the right side of the mathematical formula from the mathematical formula related to the physical phenomenon and the generated variable list and variable conditions, and stores them in the variable relationship database; and when generating the fault tree, it constructs causal relationships related to the malfunction events based on the variable relationships stored in the variable relationship database, thereby reducing the number of branches in the fault tree.

[0010] Other problems and solutions disclosed in the present application will be made clear in the detailed description and drawings. [Effects of the Invention]

[0011] According to the present invention, in a fault tree analysis, it is possible to extract candidate causes of a fault or the like without omission or duplication, while narrowing down the number of candidates to an appropriate number. [Brief explanation of the drawings]

[0012] [Figure 1] FIG. 1 is a diagram illustrating an example of the overall configuration of a system including a fault tree generation system according to an embodiment. [Figure 2]FIG. 2 is a diagram illustrating an example of the configuration of a mathematical formula database. [Figure 3] FIG. 10 is a diagram illustrating an example of a configuration of a variable database. [Figure 4] FIG. 2 is a diagram illustrating an example of a configuration of a variable knowledge database. [Figure 5] FIG. 2 is a diagram illustrating an example of the configuration of an expression format mathematical expression database. [Figure 6] FIG. 10 is a diagram illustrating an example of a configuration of a variable relation database. [Figure 7] FIG. 10 is a diagram illustrating an example of a configuration of a variable synonym dictionary database. [Figure 8] FIG. 10 is a diagram illustrating an example of a causal model. [Figure 9] 10 is a flowchart showing an example of the flow of a mathematical-formula-based causal model generation process. [Figure 10] FIG. 10 is a diagram illustrating a process for generating a mathematical expression-based causal model. [Figure 11] FIG. 10 is a diagram illustrating a process for generating a mathematical expression-based causal model. [Figure 12] FIG. 10 is a diagram illustrating a process for generating a mathematical expression-based causal model. [Figure 13] FIG. 10 is a diagram illustrating an example of a mathematical formula-based causal model. [Figure 14] FIG. 10 is a diagram illustrating an example of a mathematical formula-based causal model. [Figure 15] FIG. 10 is a diagram illustrating an example of a mathematical formula-based causal model. [Figure 16] FIG. 1 illustrates an example of a formula-based fault tree. [Figure 17] FIG. 10 is a diagram illustrating an example of a correspondence relationship between a mathematical expression and a fault tree generated based on the mathematical expression. [Figure 18] FIG. 10 is a diagram illustrating an example of a fault tree in which branches have been appropriately reduced by the fault tree generation system according to the embodiment. [Figure 19] FIG. 2 is a diagram illustrating an example of the configuration of a mathematical formula decoding unit and a mathematical formula processing unit. [Figure 20] 10 is a flowchart showing an example of the flow of a variable condition generation process executed by a mathematical expression decoding unit. [Figure 21]10 is a flowchart showing an example of the flow of processing executed by a mathematical processing unit. [Figure 22] FIG. 10 is a diagram illustrating an example of a setting user interface for a mathematical processing unit. [Figure 23] FIG. 10 is a diagram illustrating an example of a setting user interface for a mathematical processing unit. [Figure 24] FIG. 24 is a diagram illustrating an example of a fault tree in which the settings shown in FIG. 23 are reflected. [Figure 25] FIG. 10 is a diagram illustrating an example of a method for constructing a variable knowledge database. DETAILED DESCRIPTION OF THE INVENTION

[0013] In the following description, an "interface apparatus" may refer to one or more interface devices. The one or more interface devices may be at least one of the following: One or more I / O (Input / Output) interface devices. The I / O (Input / Output) interface devices are interface devices for at least one of the I / O device and a remote display computer. The I / O interface device for the display computer may be a communications interface device. The at least one I / O device may be a user interface device, for example, either an input device such as a keyboard and a pointing device, or an output device such as a display device. One or more communication interface devices. The one or more communication interface devices may be one or more homogeneous communication interface devices (e.g., one or more NICs (Network Interface Cards)) or two or more heterogeneous communication interface devices (e.g., a NIC and an HBA (Host Bus Adapter)).

[0014] In the following description, "memory" refers to one or more memory devices, which are an example of one or more storage devices, and may typically be a primary storage device. At least one memory device in the memory may be a volatile memory device or a non-volatile memory device.

[0015] In the following description, a "persistent storage device" may refer to one or more persistent storage devices, which are an example of one or more storage devices. A persistent storage device may typically be a non-volatile storage device (e.g., an auxiliary storage device), and more specifically, may be, for example, a hard disk drive (HDD), a solid state drive (SSD), a non-volatile memory express (NVME) drive, or a storage class memory (SCM).

[0016] In the following description, the term "storage device" may refer to at least one of memory and persistent storage device.

[0017] Furthermore, in the following description, a "processor" may refer to one or more processor devices. The at least one processor device may typically be a microprocessor device such as a CPU (Central Processing Unit), but may also be another type of processor device such as a GPU (Graphics Processing Unit). The at least one processor device may be single-core or multi-core. The at least one processor device may also be a processor core. The at least one processor device may also be a processor device in a broader sense, such as a circuit that is a collection of gate arrays written in a hardware description language that performs some or all of the processing (for example, an FPGA (Field-Programmable Gate Array), a CPLD (Complex Programmable Logic Device), or an ASIC (Application Specific Integrated Circuit)).

[0018] In the following description, functions may be described using the expression "yyy unit." However, the functions may be realized by one or more computer programs executed by a processor, by one or more hardware circuits (e.g., FPGAs or ASICs), or by a combination thereof. When a function is realized by a program executed by a processor, the specified processing is performed using a storage device and / or an interface device, etc., as appropriate, and therefore the function may be considered to be at least a part of the processor. Processing described using a function as the subject may be processing performed by a processor or a device having the processor. A program may be installed from a program source. The program source may be, for example, a computer from which the program is distributed or a computer-readable recording medium (e.g., a non-transitory recording medium). The description of each function is merely an example; multiple functions may be combined into one function, or one function may be divided into multiple functions.

[0019] In the following description, a process may be described using a "program" as the subject, but the process described using a program as the subject may also be a process performed by a processor or a device having that processor. Two or more programs may be realized as one program, or one program may be realized as two or more programs.

[0020] In the following description, information that provides an output for an input may be described using expressions such as "xxx table," but this information may be a table of any structure, or may be a neural network that generates an output for an input, or a learning model such as a genetic algorithm or random forest. Therefore, the "xxx table" may be referred to as "xxx information." In the following description, the structure of each table is an example, and one table may be divided into two or more tables, or all or part of two or more tables may be one table.

[0021] Furthermore, in the following description, a "fault tree generation system" may be a system configured with one or more physical computers, or may be a system (e.g., a cloud computing system) implemented on a group of physical computing resources (e.g., a cloud infrastructure). When a fault tree generation system "displays" display information, it may mean that the display information is displayed on a display device possessed by the computer, or that the computer transmits the display information to a display computer (in the latter case, the display information is displayed by the display computer).

[0022] Hereinafter, the present embodiment will be described in detail with reference to the drawings.

[0023] In the following description, the same or similar components will be designated by common reference numerals, and redundant description may be omitted.

[0024] Furthermore, when there are multiple elements having the same or similar functions, the multiple elements may be described by using the same reference numeral with different subscripts to distinguish between them. On the other hand, when there is no need to distinguish between the multiple elements, the subscripts may be omitted.

[0025] <Configuration example of fault tree generation system 100> First, a configuration example of a fault tree generation system 100 according to this embodiment will be described with reference to Fig. 1. Fig. 1 is a diagram schematically illustrating an example of the configuration of an entire system including the fault tree generation system 100.

[0026] (Example of overall system configuration) The fault tree generation system 100 of this embodiment is a computer system that can generate a fault tree in which candidate causes of failures, etc. are extracted without omission or duplication during analysis, while narrowing down the extracted tree to an appropriate number, and is realized by a computer device or a server device having the components described below.

[0027] As shown in FIG. 1 , this fault tree generation system 100 is connected to information terminals 20a, 20b, 20c, . . . 20n (hereinafter, collectively referred to as “information terminals 20”) such as laptop PCs, tablets, and smartphones owned by users of the fault tree generation system 100, such as developers and maintenance personnel of facilities, via an appropriate communication network 50 such as the Internet or a dedicated line, so that data can be transmitted between them. The fault tree generation system 100 and the communication network 50 are connected via a wired connection using well-known communication equipment (not shown), but may also be connected wirelessly. The information terminals 20 and the communication network 50 are connected wirelessly, but may also be connected via a wired connection. Each user of the fault tree generation system 100 who owns an information terminal 20 is assigned a unique ID called a user ID in advance.

[0028] The information terminal 20 supplies input data to the fault tree generation system 100. The information terminal 20 also receives output data from the fault tree generation system 100. That is, each user of the fault tree generation system 100 can obtain various information, including the fault tree generated by the fault tree generation system 100, via the information terminal 20.

[0029] In this embodiment, the fault tree generation system 100 has been described as being composed of one device as shown in Fig. 1. However, for example, the fault tree generation system 100 may be composed of multiple devices.

[0030] 1, the fault tree generation system 100 is described as being connected to the information terminal 20 via the communication network 50. However, for example, the fault tree generation system may include the information terminal 20. Also, for example, the fault tree generation system may be configured to include some or all of the functions performed by the information terminal 20.

[0031] Furthermore, other devices such as a server device that stores various data may be further connected to the fault tree generation system 100 via the communication network 50.

[0032] (Example of Hardware Configuration of Fault Tree Generation System 100) Next, an example of the hardware configuration of the fault tree generation system 100 according to this embodiment will be described.

[0033] As described above, the fault tree generation system 100 of this embodiment is realized by a single general-purpose computer. In the following description, the fault tree generation system 100 is assumed to be realized by a single general-purpose computer including one or more processors 110, one or more storage devices (120, 130), one or more interface devices, and wired or wireless communication lines connecting them.

[0034] That is, the fault tree generation system 100 includes a storage device consisting of a memory 120 and a persistent storage device 130, an interface device including an input device 140 and an output device 150, a communication device 160, and a processor 110 connected thereto.

[0035] The persistent storage device 130 is an auxiliary storage device made up of a non-volatile storage element such as a flash memory. Specific examples of the persistent storage device 130 include a solid state drive (SSD) and a hard disk drive (HDD). The persistent storage device 130 stores at least a fault tree generation program (not shown). The fault tree generation program is a computer program for implementing the functions required for the fault tree generation system 100. This fault tree generation program holds, for example, various information related to variables.

[0036] The fault tree generation program is executed by the processor 110, whereby various processes for generating a fault tree are performed (described in detail below).

[0037] The fault tree generation program may be installed from a program source. The program source may be, for example, a program distribution computer or a computer-readable recording medium. The fault tree generation program may also be configured by device drivers, an operating system, various application programs located above them, and libraries that provide common functions to these programs. Furthermore, two or more programs may be implemented as one fault tree generation program, and one fault tree generation program may be implemented as two or more programs.

[0038] Furthermore, at least data relating to mathematical expressions, data relating to defect information, data relating to variables, etc. are stored in the persistent storage device 130. Details of these data and the databases that store each data will be described later.

[0039] The memory 120 is a primary storage device mainly consisting of volatile storage elements such as RAM (Random Access Memory). The memory 120 temporarily stores data representing various information read from the persistent storage device 130 and various data acquired from the information terminal 20.

[0040] The processor 110 is a processor device such as a CPU (Central Processing Unit) and various co-processors. The processor 110 loads a fault tree generation program into the memory 120 and executes it, thereby performing overall control of the fault tree generation system 100 itself and also managing a control unit that performs various processes such as calculation processing and judgment processing.

[0041] The interface device includes a communication interface device that connects to the communication network 50 and communicates with the information terminal 20, and an I / O interface device that includes an input device 140 and an output device 150.

[0042] (Example of functional blocks of the fault tree generation system 100) Next, an example of blocks of various functions included in the fault tree generation system 100 according to this embodiment will be described. Note that each block described below does not represent a hardware-based configuration, but represents a functional-based block.

[0043] The fault tree generation system 100 is configured to include the following functional blocks: a control unit, a storage unit, a user interface unit, and a communication unit (none of which are shown).

[0044] The control unit executes various data processing based on the user's operation input detected by the user interface unit, the data acquired by the communication unit, and the fault tree generation program and data stored in the storage unit. The control unit also functions as an interface for the user interface unit, the communication unit, and the storage unit.

[0045] The control unit has at least the following functional blocks: a top event input unit 210, a mathematical expression-based causal model generation unit 220, a mathematical expression-based fault tree generation unit 230, a malfunction information-based causal model generation unit 240, a malfunction information-based causal model combination unit 250, a mathematical expression decoding unit 260, and a mathematical expression processing unit 270.

[0046] The top event input unit 210 executes a process of receiving a top event input operation performed by the user of the fault tree generation system 100 via the communication unit or the input unit.

[0047] The mathematical-expression-based causal model generation unit 220 executes processing to generate a causal relationship for a mathematical expression stored in the mathematical expression database 310 based on the relationship between the variables on the left side and the variables on the right side of the mathematical expression. This causal relationship generated by the mathematical-expression-based causal model generation unit 220 is referred to as a mathematical-expression-based causal model. Furthermore, processing by the mathematical-expression-based causal model generation unit 220 to generate a mathematical-expression-based causal model is referred to as a mathematical-expression-based causal model generation processing. Details of the mathematical-expression-based causal model generation processing will be described later with reference to FIGS. 9 to 15.

[0048] The mathematical-expression-based fault tree generator 230 executes a process of generating a mathematical-expression-based fault tree by acquiring mathematical-expression-based causal models for the top events received by the top event input unit 210 from a mathematical-expression-based causal model database 320 (described later) and combining all the acquired mathematical-expression-based causal models. This process executed by the mathematical-expression-based fault tree generator 230 is referred to as a mathematical-expression-based fault tree generation process. Details of the mathematical-expression-based fault tree generation process will be described later with reference to FIG. 16 .

[0049] The defect-information-based causal model generation unit 240 extracts, by natural language processing, the causal relationships of defects contained in defect information stored in the defect information database 330 (described later), and executes processing to generate a defect-information-based causal model.

[0050] The malfunction-information-based causal model combining unit 250 combines the mathematical-expression-based fault tree generated by the mathematical-expression-based fault tree generating unit 230 with the malfunction-based causal models stored in the malfunction-information-based causal model database 340 (described later) and outputs the combined results.

[0051] The mathematical expression interpretation unit 260 executes at least a process for generating variable conditions. This process executed by the mathematical expression interpretation unit 260 is referred to as a variable condition generation process. Details of the variable condition generation process will be described later with reference to FIGS. 19 and 20.

[0052] The mathematical expression processor 270 executes a process of appropriately reducing the branches of the generated fault tree by using the variable conditions generated in the variable condition generation process. Details of this process will be described later with reference to FIG. 19 and FIGS. 21 to 24.

[0053] The control unit is configured using the processor 110, and can realize these functional blocks by reading the above-mentioned fault tree generation program from the persistent storage device 130 onto the memory 120 and executing it. Note that the control unit may be configured using a logic circuit such as an FPGA (Field Programmable Gate Array) instead of the processor 110. The control unit may also be configured by combining the processor 110 and a logic circuit.

[0054] The storage unit is configured using a storage device, for example, consisting of memory 120 and persistent storage device 130, and stores a fault tree generation program that supplies various processing instructions to the control unit, and data representing various information used in the processing executed by the control unit.

[0055] The storage unit has at least the following functional blocks: a mathematical expression database 310, a mathematical expression-based causal model database 320, a malfunction information database 330, a malfunction information-based causal model database 340, a variable database 350, a variable knowledge database 360, an expression format mathematical expression database 370, and a variable relation database 380.

[0056] The mathematical formula database 310 is a database for storing and managing mathematical formulas related to physical phenomena.

[0057] An example of the configuration of the mathematical formula database 310 is shown in FIG. 2. The mathematical formula database 310 has a record for each mathematical formula ID, which is an identifier for uniquely identifying each mathematical formula. The record represents the mathematical formula ID, the name of the product, part, etc. (hereinafter referred to as "part, etc.") to which the mathematical formula represented by the mathematical formula ID is applied, the variables on the left side of the mathematical formula, and the variables on the right side of the mathematical formula. In other words, the mathematical formula database 310 records, for each record, the mathematical formula ID, the application target of the mathematical formula represented by the mathematical formula ID, the variables on the left side, and the variables on the right side.

[0058] For example, in a part in which a shaft portion is press-fitted into a cylindrical portion for fastening, the friction force F generated in the shaft portion can be calculated using the following (Equation 1).

[0059]

number

[0060] 2, the above (Equation 1) is assigned the equation ID "1" in the equation database 310. In addition, the record in the equation database 310 corresponding to the above (Equation 1) records that the applicable object of the equation is "a fastening part between a shaft portion and a cylindrical portion (a fastening part between a shaft and a cylinder)," that the variable on the left side of the equation is "a friction coefficient of the shaft portion of the part (shaft: friction force)," and that the variables on the right side of the equation are "a friction coefficient between the cylindrical portion and the shaft portion of the part (between the cylinder and the shaft: friction coefficient), an internal pressure acting between the cylindrical portion and the shaft portion of the part (between the cylinder and the shaft: internal pressure), an area of ​​a joint between the cylindrical portion and the shaft portion of the part (joint: area), and an inner diameter of the cylindrical portion of the part (cylinder: inner diameter)."

[0061] As mentioned above, in the records of the mathematical formula database 310, the variables on the left and right sides are recorded as a set with the part to which the mathematical formula is applied in a format such as "part:variable". For example, "friction coefficient between the cylindrical part and the shaft part" is recorded in the format such as "between the cylinder and the shaft: friction coefficient". When there are multiple variables, they are stored separated by ",".

[0062] Moreover, the internal pressure P between the cylindrical portion and the shaft portion is calculated by the following (Equation 2).

[0063]

number

[0064] 2, the above (Equation 2) is assigned a mathematical formula ID of "2" in the mathematical formula database 310. Furthermore, the record in the mathematical formula database 310 corresponding to the above (Equation 2) records that the target of application of the mathematical formula is "a fastening part between a shaft portion and a cylindrical portion (a fastening part between a shaft and a cylinder)," that the variable on the left side of the mathematical formula is "an internal pressure acting between the cylindrical portion and the shaft portion of the part (internal pressure between the cylinder and the shaft)," and that the variables on the right side of the mathematical formula are "an inner diameter of the cylindrical portion of the part (cylinder: inner diameter), an outer diameter of the cylindrical portion (cylinder: outer diameter), an elastic modulus of the shaft portion of the part (shaft: elastic modulus), and an interference of the shaft portion (shaft: interference)."

[0065] Furthermore, the area A of the joint between the cylindrical portion and the shaft portion of the part is calculated by the following (Equation 3).

[0066]

number

[0067] 2, the above (Equation 3) is assigned the equation ID "3" in the equation database 310. Furthermore, the record in the equation database 310 corresponding to the above (Equation 3) records that the applicable object of the equation is "a fastening part between a shaft portion and a cylindrical portion (a fastening part between a shaft and a cylinder)," that the variable on the left side of the equation is "the area of ​​the joint between the cylindrical portion and the shaft portion of the part (joint: area)," and that the variables on the right side of the equation are "the inner diameter of the cylindrical portion of the part (cylinder: inner diameter), and the insertion length of the shaft portion of the part (shaft: insertion length)."

[0068] The mathematical-expression-based causal model database 320 is a database for storing and managing the mathematical-expression-based causal models generated by the mathematical-expression-based causal model generating unit 220.

[0069] The defect information database 330 is a database for managing defect information.

[0070] The defect information-based causal model database 340 is a database for storing and managing the defect information-based causal models generated by the defect information-based causal model generating unit 240.

[0071] The variable database 350 is a database for storing and managing variables extracted from technical documents.

[0072] An example of the configuration of the variable database 350 is shown in FIG. 3. The variable database 350 has a record for each variable symbol, which is a symbol representing a variable. The record represents this variable symbol and the phenomenon that is the target of the variable represented by the variable symbol (hereinafter referred to as the "target phenomenon"). In other words, the variable database 350 defines the target phenomenon of the variable represented by the variable symbol by recording the variable symbol and the target phenomenon in association with each other. In the example shown in FIG. 3, the variable database 350 records that the target phenomenon of the variable represented by the variable symbol "T" is "torque."

[0073] The variable knowledge database 360 ​​is a database for storing and managing variable knowledge, which is various types of knowledge related to variables.

[0074] An example of the configuration of the variable knowledge database 360 ​​is shown in FIG. 4. The variable knowledge database 360 ​​has a record for each target product, facility, etc. (hereinafter collectively referred to as "product, etc."). The record represents the name of the target product, etc., the name of a part of the product, etc., the name of a portion of the part, a numerical value representing the size of the portion, and the unit of the numerical value. The numerical value is recorded as a numerical range by recording a minimum and a maximum value. In other words, the variable knowledge includes at least one of a physical quantity related to a physical phenomenon and a design value related to the design of the product, etc. The variable knowledge database 360 ​​centrally manages variable knowledge by linking and recording the target product, etc., the part of the product, etc., the portion of the part, the numerical value representing the size of the portion, and the unit of the numerical value. In the example shown in FIG. 4, the variable knowledge database 360 ​​records that the "diameter" of the "shaft" of a "wind turbine" is "10 cm to 100 cm."

[0075] The expression format mathematical expression database 370 is a database for managing the expression format of each mathematical expression.

[0076] An example of the configuration of the expression-form mathematical formula database 370 is shown in FIG. 5. The expression-form mathematical formula database 370 has a record for each mathematical formula ID. The record represents the mathematical formula ID and the expression format of the mathematical formula represented by the mathematical formula ID (hereinafter referred to as an "expression-form mathematical formula"). In other words, the expression-form mathematical formula database 370 centrally manages the expression formats of the mathematical formulas by linking and recording the mathematical formula ID and the expression format of the mathematical formula represented by the mathematical formula ID. According to the example shown in FIG. 5, the expression-form mathematical formula database 370 records that the above (Formula 1) represented by the mathematical formula ID "1" is expressed in the format "T = \mu PA \frac{d_1}{2}".

[0077] The variable relation database 380 is a database for storing and managing variable relations that represent the relationship between the variables on the left side of each mathematical expression related to physical phenomena and the variables on the right side of the mathematical expression.

[0078] An example of the configuration of the variable relation database 380 is shown in Figure 6. The variable relation database 380 has a record for each mathematical formula ID. The record represents the mathematical formula ID, a variable symbol (hereinafter referred to as the "left-hand side variable symbol") representing the variable on the left side of the mathematical formula represented by the mathematical formula ID, the state of the variable on the left side of the mathematical formula (hereinafter referred to as the "left-hand side variable state"), the range of values ​​that the variable on the right side of the mathematical formula can take when the variable on the left side is increased (hereinafter referred to as the "right-hand side variable range when the variable on the left side is increased"), the state of the variable on the right side of the mathematical formula in the case of the left-hand side variable state (hereinafter referred to as the "right-hand side variable state"), and whether or not the left-hand side variable state in the mathematical formula is the target of a selection operation by the user of the fault tree generation system 100. If the left-hand side variable state in the mathematical formula is the target of a selection operation by the user of the fault tree generation system 100, "1" is recorded in the record, and if it is not the target of a selection operation, "0" is recorded. 6, the variable relation database 380 records, in association with a formula ID, the left-hand side variable symbol, the left-hand side variable state, the right-hand side variable range when the left-hand side variable is increased, and the right-hand side variable state of the formula represented by the formula ID, and whether or not the left-hand side variable state has been selected in the formula, thereby managing the relationship between the left-hand side variable and the right-hand side variable in the formula in a unified manner. According to the example shown in Fig. 6, for the (Formula 1) represented by formula ID "1", when the left-hand side variable represented by the symbol "\mu" is in the "increasing" state, the right-hand side variable is also in the "increasing" state, the range of values ​​that the right-hand side variable in the formula can take when the left-hand side variable is increased is "(-∞ to +∞)", and the left-hand side variable state in the formula is the target of a selection operation by the user of the fault tree generation system 100.

[0079] The storage unit may also have a variable thesaurus database as a functional block.

[0080] An example of the configuration of this variable synonym dictionary database is shown in Figure 7. The variable synonym dictionary database has a record for each target phenomenon. The record represents a general name that represents this target phenomenon, synonyms for that name, and applications of that name and synonyms. In other words, the variable synonym dictionary database centrally manages the name of the target phenomenon and its synonyms by linking and recording the name of the target phenomenon, synonyms for that name, and applications of that name and synonyms. According to the example shown in Figure 7, the target phenomenon represented by the name "torque" has a synonym called "frictional force," and it is recorded that these terms are applied to "fastening."

[0081] The control unit can execute the various processes described above by reading and writing this information in the storage unit.

[0082] The user interface unit accepts input operations from the user and is responsible for user interface-related processing such as image display and audio output. The user interface unit has the functional blocks of an input unit and an output unit. The input unit detects various operations from the user. The input unit is configured using an input device 140 such as a keyboard, pointing device, or touch panel. The output unit displays the created production and logistics plan on a display device, and also displays various screens on the display device and outputs audio. The output unit is configured using an output device 150 such as an LCD display or touch screen.

[0083] The communication unit is responsible for communication processing with other devices, such as the information terminals 20 owned by each user of the fault tree generation system 100, via the Internet (an example of the communication network 50). The communication unit is configured using a communication device 160, such as a NIC (Network Interface Card) or an HBA (Host Bus Adapter).

[0084] That is, each component of the fault tree generation system 100 is realized by hardware including a processor 110, storage devices such as memory 120 and persistent storage device 130, wired or wireless communication lines and interface devices that connect them, and software stored in the storage devices (120, 130) that supplies processing instructions to the arithmetic units.

[0085] In this embodiment, the fault tree generation system 100 has been described as being implemented as a single computer. However, these functions may be implemented by multiple interconnected computers or server devices. Furthermore, the fault tree generation system 100 may include a general-purpose computer such as a laptop PC with a web browser installed thereon, or may include a web server or various portable devices.

[0086] Furthermore, the description of each function is an example, and multiple functions may be combined into one function, or one function may be divided into multiple functions.

[0087] <Causal model> As described above, the mathematical-expression-based causal model generation unit 220 generates a mathematical-expression-based causal model from the relationship between the variables on the left side and the variables on the right side of the mathematical expressions stored in the mathematical expression database 310.

[0088] Furthermore, as described above, the defect information-based causal model generation unit 240 extracts the causal relationships of defects contained in the defect information stored in the defect information database 330 using natural language processing, and generates a defect information-based causal model.

[0089] Here, the causal model is a model of a causal relationship.

[0090] Figure 8 shows an example of the configuration of a causal model.

[0091] Each element 801 that makes up the causal chain is made up of a part 802 and a phenomenon 803. These elements are expressed in the order of causality as a chain. In the example shown in Figure 8, the element on the right causes the element on the left.

[0092] The relationships connecting elements include an AND condition 804 and an OR condition 805. The AND condition 804 indicates that the element on the right is triggered when all of the connected elements occur. On the other hand, the OR condition 805 indicates that the element on the right is triggered when any one of the connected elements occurs.

[0093] The causal model illustrated in FIG. 8 indicates that the occurrence of phenomenon D809 in part D808 or the occurrence of phenomenon E811 in part E810 causes phenomenon B807 in part B806.

[0094] The causal model illustrated in FIG. 8 also indicates that phenomenon B807 occurs in part B806 and phenomenon C813 occurs in part C812, resulting in phenomenon A803 occurring in part A802.

[0095] A causal model generated from a mathematical formula in this way is called a mathematical formula-based causal model. On the other hand, a causal model generated from the causal relationships of defects described in defect information is called a defect information-based causal model.

[0096] <Formula-based causal model generation processing> Next, the mathematical expression-based causal model generation process executed by the fault tree generation system 100 according to this embodiment will be described with reference to FIGS.

[0097] FIG. 9 is a flowchart 900 showing an example of the flow of the mathematical-expression-based causal model generation process performed in this embodiment.

[0098] In step S901, the control unit of the fault tree generation system 100 executes a process in which the mathematical-formula-based causal model generation unit 220 determines the incompatibility of the variable on the left side as a malfunction event, and further generates elements of the large variable and the small variable on the left side as its causes. For example, in the following (Equation 4), as shown in Fig. 10, the character string "incompatibility" is added to the variable "component A: variable A" on the left side, resulting in "component A: variable A incompatibility" 1001.

[0099] Part A: Variable A = f (Part B1: Variable B1, Part B2: Variable B2, , Part Bn: Variable Bn) (Equation 4)

[0100] In the above (Equation 4), each variable is expressed in the format of "component: variable." For example, the left side of the above (Equation 1) is expressed in the format of "shaft: frictional force." Variables have a range within which a product operates normally, and if that range is exceeded, a defect occurs in the product. When a variable exceeds the normal range, it may become larger or smaller than the normal range. For this reason, as a cause of "component A: variable A unsuitable" 1001, the control unit of the fault tree generation system 100 adds the string "large" or "small" to "component A: variable A" to generate elements "component A: variable A large" 1002 and "component A: variable A small" 1003. When the control unit of the fault tree generation system 100 completes the processing in step S901, the control unit proceeds to step S902.

[0101] In step S902, the control unit of the fault tree generation system 100 executes processing in which the mathematical-formula-based causal model generation unit 220 generates a model with the inadequacy of the variable on the right side as an element, considering the causes of the large variable and small variable on the left side generated in step S901 as a case where the variable on the right side becomes inadequate. For example, in the case of the above (Formula 4), as illustrated in Fig. 11, "Component A: Large variable A" 1002 and "Component A: Small variable A" 1003 generated in step S901 are respectively added with "Component B1: Inadequate variable B1" 1104, "Component B2: Inadequate variable B2" 1105, ..., and "Component Bn: Inadequate variable Bn" 1106. 11, the illustration of "Component B1: Variable B1 Unsuitable" 1104, "Component B2: Variable B2 Unsuitable" 1105, ..., "Component Bn: Variable Bn Unsuitable" 1106, which are added to "Component A: Variable A Small" 1003, is omitted. When the control unit of the fault tree generation system 100 completes the processing in step S902, it proceeds to step S903.

[0102] In step S903, the control unit of the fault tree generation system 100 causes the mathematical-formula-based causal model generation unit 220 to execute a process of generating elements of a large variable and a small variable on the right-hand side as factors of the variable incompetence on the right-hand side generated in step S902. As in step S901, this is based on the idea that there is a range in which a product operates normally, and that if that range is exceeded, a malfunction will occur in the product, and that when the normal range is exceeded, the variable may be larger or smaller than the normal range. For example, in the case of the above (Equation 4), as illustrated in FIG. 12, "Component B1: Variable B1 Large" 1201 and "Component B1: Variable B1 Small" 1202 are added to "Component B1: Variable B1 Incompetence" 1104, "Component B2: Variable B2 Incompetence" 1105, ..., "Component Bn: Variable Bn Incompetence" 1106 generated in step S902. 12, the same factors as those for "Component A: Variable A Large" are also added to "Component A: Variable A Small." This enables the fault tree generation system 100 to generate a mathematical-expression-based causal model. Upon completing the processing in step S903, the control unit of the fault tree generation system 100 ends the mathematical-expression-based causal model generation processing shown in the flowchart 900 of FIG. 9.

[0103] In this way, the mathematical-expression-based causal model generation unit 220 generates a mathematical-expression-based causal model by the above steps S901 to S903 from the relationship between the variables on the left side and the variables on the right side of the mathematical expression stored in the mathematical expression database 310, as described above. That is, if there are variables on the left side and the right side, the mathematical-expression-based causal model generation unit 220 can generate a mathematical-expression-based causal model by the above steps S901 to S903. Therefore, the control unit of the fault tree generation system 100 causes the mathematical-expression-based causal model generation unit 220 to acquire variables from the "variables on the left side" column and the "variables on the right side" column of the table in the mathematical expression database 310 illustrated in FIG. 2, and generates a mathematical-expression-based causal model. For example, when generating a causal model for the above (Equation 1), the variables on the left side are "axis: friction force" and the variables on the right side are "between cylinder and axis: friction coefficient, between cylinder and axis: internal pressure, joint: area, cylinder: inner diameter," and a formula-based causal model is generated.

[0104] An example of a mathematical-expression-based causal model generated by the mathematical-expression-based causal model generation unit 220 based on the above (Equation 1) is shown in Figure 13. Similarly, an example of a mathematical-expression-based causal model generated by the mathematical-expression-based causal model generation unit 220 based on the above (Equation 2) is shown in Figure 14, and an example of a mathematical-expression-based causal model generated by the mathematical-expression-based causal model generation unit 220 based on the above (Equation 3) is shown in Figure 15. The mathematical-expression-based causal model generated by the mathematical-expression-based causal model generation unit 220 is stored in a mathematical-expression-based causal model database 320.

[0105] <Formula-based fault tree generation process> Next, the mathematical expression-based fault tree generation process executed by the fault tree generation system 100 according to this embodiment will be described.

[0106] First, the control unit of the fault tree generation system 100 executes a process in which the mathematical-expression-based fault tree generation unit 230 extracts a mathematical-expression-based causal model related to the top event received by the top event input unit 210 from the mathematical-expression-based causal model database 320. For example, when an input operation specifying "axis: inadequate friction force" as the top event is received from the user of the fault tree generation system 100, the mathematical-expression-based causal model database 320 is searched and a mathematical-expression-based causal model including "axis: inadequate friction force" is extracted from the mathematical-expression-based causal model database 320. As a result, the mathematical-expression-based causal model generated based on the above-mentioned (Equation 1) and illustrated in FIG. 6 is extracted as a search result.

[0107] Next, the control unit of the fault tree generation system 100 executes a process of searching the mathematical-expression-based causal model database 320 for a mathematical-expression-based causal model including each element of the mathematical-expression-based causal model extracted in step S1601. For example, when a search is performed using "between cylinder and shaft: large internal pressure", which is one of the elements of the mathematical-expression-based causal model generated based on the above (Equation 1), the mathematical-expression-based causal model illustrated in FIG. 7, which is generated based on the above (Equation 2), is extracted as a search result. Similarly, when a search is performed using "joint: large area", which is one of the elements of the mathematical-expression-based causal model generated based on the above (Equation 1), the mathematical-expression-based causal model illustrated in FIG. 8, which is generated based on the above (Equation 3), is extracted as a search result.

[0108] Furthermore, the control unit of the fault tree generation system 100 uses the mathematical-expression-based fault tree generation unit 230 to combine these matched mathematical-expression-based causal models with the mathematical-expression-based causal model generated based on the above (Equation 1). Furthermore, individual elements of the combined mathematical-expression-based causal models are searched, and if any are found, they are combined. This process is repeated and generated, and this is called a mathematical-expression-based fault tree. An example of a mathematical-expression-based fault tree generated by the mathematical-expression-based fault tree generation process is shown in FIG. 16. Furthermore, FIG. 17 shows the correspondence between the above (Equation 1) and the mathematical-expression-based fault tree generated based on the above (Equation 1), as shown in FIG. 16. In this way, the fault tree generation system 100 can generate fault trees for all top events without omissions or duplications.

[0109] The fault tree generation system 100 may calculate and display a score for each element of the generated fault tree. In such a case, the fault tree generation system 100 may gray out or hide elements with low scores. There are two ways to calculate the score:

[0110] The first method is to use a formula. As mentioned above, when a formula-based causal model is generated based on the above formula (1), an example is shown in Figure 6. Two factors are ultimately generated as factors contributing to "shaft: large friction coefficient": "cylinder-to-shaft: large friction coefficient" and "cylinder-to-shaft: small friction coefficient." In the above formula (1), when the variable on the right side, "cylinder-to-shaft: friction coefficient," increases, the variable on the left side, "shaft: friction coefficient," increases. Conversely, when the variable on the right side, "cylinder-to-shaft: friction coefficient," decreases, the variable on the left side, "shaft: friction coefficient," decreases. In other words, "cylinder-to-shaft: large friction coefficient" is a more appropriate factor contributing to "shaft: large friction coefficient," and "cylinder-to-shaft: large friction coefficient" has a higher score. In this way, the score is calculated based on the effect on the left side when the variable on the right side is changed.

[0111] The second method is to use past performance, that is, past defect information. Specifically, this method calculates the score based on the number of defect information items in which a defect event in the fault tree and an event that is a candidate for its cause are both described. For example, the defect information database 330 is searched to find the number of defect information items in which both "shaft: large friction coefficient" and "between cylinder and shaft: large friction coefficient" are described, and the number of defect information items in which both "shaft: large friction coefficient" and "between cylinder and shaft: small friction coefficient" are described. If there are many defect information items in which both "shaft: large friction coefficient" and "between cylinder and shaft: large friction coefficient" are described, this means that "between cylinder and shaft: large friction coefficient" is more likely and appropriate as a cause of "shaft: large friction coefficient."

[0112] <Outline of the Formula Decoding Unit and Formula Processing Unit> The fault tree generation system 100 according to this embodiment can appropriately reduce the branches of the generated fault tree as illustrated in Fig. 18 by the processes executed by the formula interpretation unit 260 and the formula processing unit 270, respectively. As a result, the fault tree generation system 100 can generate a fault tree in which cause candidates of faults, etc. are extracted without omission or duplication, and which enables the user of the fault tree generation system 100 to efficiently perform cause investigation work when a malfunction occurs and review work at the design stage. Fig. 18 shows how the branches of the fault tree illustrated in Fig. 17 are appropriately reduced as a result of the processes executed by the formula interpretation unit 260 and the formula processing unit 270. Next, an overview of the formula interpretation unit 260 and the formula processing unit 270 will be described below with reference to Figs. 19 to 24.

[0113] FIG. 19 is a diagram schematically illustrating an example of the mathematical formula decoding unit 260 and the mathematical formula processing unit 270. As shown in FIG.

[0114] 19, the fault tree generation system 100 appropriately reduces the branches of the generated fault tree by executing, in this order, the variable condition generation process performed by the mathematical formula interpretation unit 260 (described in detail later in relation to FIG. 20) and the process performed by the mathematical formula processing unit 270 (described in detail later in relation to FIG. 21). At this time, the mathematical formula interpretation unit 260 performs the variable condition generation process using, as input data, the target device, component, phenomenon, and mathematical formula ID input by the user of the fault tree generation system 100 and the variables selected by the user, as shown in FIG. 19. Note that, in the fault tree generation system 100, the user can input a mathematical formula by specifying a mathematical formula ID and reading the corresponding mathematical formula from the mathematical formula database 310, as shown in FIG. 19. In addition, the user can directly input the mathematical formula into a predetermined input field. Furthermore, in the fault tree generation system 100, when a mathematical formula is directly input, the notation of the mathematical formula may be a representation-style formula, as illustrated in FIG. 19, a simulation-style formula, or even a formula expressed in another notation.

[0115] (Variable condition generation process of the mathematical formula decoding unit 260) FIG. 20 is a flowchart 2000 showing an example of the flow of the variable condition generation process executed by the mathematical expression decoding unit 260.

[0116] In step S2001, the control unit of the fault tree generation system 100 executes a process in which the formula decoding unit 260 compares the variable data stored in the variable database 350 with the characters of the formula input by the user, starting from the beginning, and creates a variable list for the formula. After completing the process in step S2001, the control unit of the fault tree generation system 100 proceeds to step S2002.

[0117] In step S2002, the control unit of the fault tree generation system 100 executes a process of creating a numerical value list by causing the formula interpretation unit 260 to search the variable knowledge database 360 ​​for the variables included in the variable list generated in step S2001, and obtaining the numerical values ​​and units of the records corresponding to the conditions related to the target devices, etc. and components input by the user of the fault tree generation system 100. When the process in step S2002 is completed, the control unit of the fault tree generation system 100 proceeds to step S2003.

[0118] In step S2003, the control unit of the fault tree generation system 100 executes a process of calculating a multiple list that aligns the powers of the same type of unit (length, weight, etc.) for the numerical value list created in step S2002 using the mathematical formula interpretation unit 260. After completing the process in step S2003, the control unit of the fault tree generation system 100 proceeds to step S2004.

[0119] In step S2004, the control unit of the fault tree generation system 100 executes a process in which the numerical values ​​in the numerical value list created in step S2002 are multiplied by the values ​​in the multiple list calculated in step S2003 using the mathematical formula interpretation unit 260. After completing the process in step S2004, the control unit of the fault tree generation system 100 proceeds to step S2005.

[0120] In step S2005, the control unit of the fault tree generation system 100 causes the mathematical formula interpretation unit 260 to execute processing to output the minimum, median, and maximum values ​​of the range of numerical values ​​as variable conditions for all variables in the mathematical formula. This generates variable conditions for the mathematical formula. Upon completing the processing in step S2005, the control unit of the fault tree generation system 100 ends the variable condition generation processing shown in the flowchart 2000 of FIG. 20.

[0121] (Mathematical processing unit 270) FIG. 21 is a flowchart 2100 showing an example of the flow of processing executed by the mathematical-formula processing unit 270.

[0122] In step S2101, the control unit of the fault tree generation system 100 executes processing to select one variable candidate by the mathematical processing unit 270 in response to a selection operation by the user received via the input unit or the selection unit. The fault tree generation system 100 displays the setting user interface (2200, 2300) of the mathematical processing unit 270 illustrated in Figures 22 and 23 on the display device of the information terminal 20 held by the user or on the output device 150 of the fault tree generation system 100, and receives the selection operation and / or input operation by the user. When the control unit of the fault tree generation system 100 completes the processing in step S2101, it proceeds to step S2102.

[0123] In step S2102, the control unit of the fault tree generation system 100 executes processing to create a combination list of variable conditions for other variables using the mathematical processing unit 270. When the processing in step S2102 is completed, the control unit of the fault tree generation system 100 proceeds to step S2103.

[0124] In step S2103, the control unit of the fault tree generation system 100 executes a process of selecting one record from the combination list created in step S2102 using the mathematical processing unit 270. The control unit of the fault tree generation system 100 repeatedly executes the processes of steps S2103 to S2107 for each record in the combination list. After completing the process in step S2103, the control unit of the fault tree generation system 100 proceeds to step S2104.

[0125] In step S2104, the control unit of the fault tree generation system 100 uses the mathematical processing unit 270 to calculate an equation in which the variable condition of the variable on the left side of the equation is applied to the equation. For example, when the variable on the left side of the equation is ≧0, the calculation is performed assuming that the right side of the equation is ≧0. Then, for the result obtained by the calculation process, a process is performed to obtain extreme value conditions and new variable conditions. When the process in step S2104 is completed, the control unit of the fault tree generation system 100 proceeds to step S2105.

[0126] In step S2105, the control unit of the fault tree generation system 100 executes processing to generate a variable condition by combining the variable condition and the variable condition obtained immediately before, using the mathematical processing unit 270. When the processing in step S2105 is completed, the control unit of the fault tree generation system 100 proceeds to step S2106.

[0127] In step S2106, the control unit of the fault tree generation system 100 executes the process of differentiating the formula and performing a simulation using the formula processing unit 270. Specifically, the formula is differentiated, the conditions under which the differentiated function becomes 0 are calculated, the presence or absence of an extreme value is checked, and a simulation is performed under the variable conditions. When the control unit of the fault tree generation system 100 completes the process in step S2106, it proceeds to step S2107.

[0128] In step S2107, the control unit of the fault tree generation system 100 executes a process in which the mathematical processing unit 270 determines whether the result of the simulation executed in step S2106 is positive or negative, and records the result in an increase / decrease relationship list.

[0129] The control unit of the fault tree generation system 100 repeatedly executes the processes of steps S2103 to S2107 for each record in the combination list. After executing the processes of steps S2103 to S2107 for all records in the combination list and completing the process of step S2107, the control unit returns to step S2101 and repeatedly executes the processes of steps S2101 to S2107 for all variable candidates. That is, the control unit of the fault tree generation system 100 repeats the processes of steps S2101 to S2107 until they have been executed for all variables on the right-hand side of the equation. After executing the processes of steps S2101 to S2107 for all variable candidates and completing the process of step S2107, the control unit of the fault tree generation system 100 ends the process shown in the flowchart 2100 of FIG. 21.

[0130] As a result, the fault tree generation system 100 can appropriately reduce the branches of the generated fault tree, as shown in Fig. 24. After generating a fault tree in this way in which all candidate causes of faults, etc. are extracted without omission or duplication, the fault tree generation system 100 appropriately reduces the branches of the fault tree and presents it to the user. At this time, the fault tree may be presented to the user from the beginning in a state in which the branches have been appropriately reduced, or the fault trees both before and after the branches have been appropriately reduced may be presented to the user in a form that allows them to compare them.

[0131] <How to build the variable knowledge database 360> Next, a method for constructing the variable knowledge database 360 ​​described above in relation to FIG. 4 will be described with reference to FIG.

[0132] FIG. 25 is a diagram showing an example of a method for constructing the variable knowledge database 360. As shown in FIG.

[0133] As shown in FIG. 25, the control unit of the fault tree generation system 100 extracts various data stored in the variable knowledge database 360, such as the name of the target product, the names of the parts of the product, the names of the portions of the parts, numerical values ​​representing the sizes of the portions, and the units of the numerical values, from various technical documents such as design documents, CAD data, catalogs, past analysis reports, standards such as JIS, catalogs of the target product, and papers, using well-known natural language processing techniques.

[0134] The above-described embodiment of the present invention can be summarized as follows.

[0135] (1) The fault tree generation system 100 is a system for generating a fault tree, and includes at least a processor 110 and storage devices (120, 130). The storage devices (120, 130) include at least a mathematical formula database 310 that stores mathematical formulas related to physical phenomena, a variable database 350 that stores variables extracted from technical documents, a variable knowledge database 360 ​​that stores variable knowledge, which is knowledge related to variables, and a variable relationship database 380 that stores variable relationships that represent the relationships between variables on the left side of mathematical formulas related to the physical phenomena and variables on the right side of the mathematical formulas. The processor 110 develops the causal relationships of the malfunction events based on the mathematical expressions stored in the mathematical expression database 310, and generates a mathematical expression-based causal model that is a model representing the causal relationships; based on the generated mathematical expression-based causal model, generates and outputs a fault tree that combines causal relationships related to any malfunction events from the developed causal relationships of the malfunction events; based on the variable database 350 and the variable knowledge database 360, generates a variable list of the mathematical expression related to the physical phenomenon, the variables, and variable conditions that are possible numerical values ​​for the variables; extracts variable relationships between variables on the left side and variables on the right side of the mathematical expression from the mathematical expression related to the physical phenomenon and the generated variable list and variable conditions, and stores them in the variable relationship database 380; and when generating the fault tree, constructs causal relationships related to the malfunction events based on the variable relationships stored in the variable relationship database 380, thereby reducing the number of branches in the fault tree. In this way, the fault tree generation system 100 can generate and output a fault tree in which candidate causes of failures, etc. are extracted without omission or duplication, while narrowing down the extracted fault tree to an appropriate number. As a result, the user of the fault tree generation system 100 can perform fault tree analysis using a fault tree in which candidate causes of failures, etc. are extracted without omission or duplication, while the number of branches is appropriately reduced, thereby enabling efficient cause investigation when a malfunction occurs and review work at the design stage.

[0136] (2) In the output fault tree, both the resulting higher-level events and the causal lower-level events are expressed as physical quantities, and the relationship between the physical quantities of the higher-level events and the lower-level events is expressed by mathematical formulas.

[0137] (3) The processor 110 combines the generated fault tree with the causal relationships extracted from the malfunction information.

[0138] (4) The variable knowledge includes at least one of a physical quantity related to a physical phenomenon and a design value related to the design of a product or the like.

[0139] (5) When generating a formula-based causal model, processor 110 determines the factors that cause the variable on the left side of the formula to become inappropriate as being when the variable on the left side is large and when the variable on the left side is small, and develops the causal relationship of the defect according to the variable relationship between the variable on the left side of the formula and the variable on the right side of the formula stored in variable relationship database 380.

[0140] (6) When generating a formula-based causal model, processor 110 determines the factors that cause the variable on the right side of the formula to become inappropriate as being when the variable on the right side is large and when the variable on the right side is small, and develops the causal relationship of the malfunction according to the variable relationship between the variable on the left side of the formula and the variable on the right side of the formula stored in variable relationship database 380.

[0141] (7) When constructing a causal relationship related to the malfunction event, the processor 110 accepts a selection operation or an input operation for specifying a variable candidate.

[0142] (8) The processor 110 accepts an input operation for specifying a variable condition when constructing a causal relationship related to the malfunction event.

[0143] The present invention is not limited to the above-described embodiment, and can be implemented using any components without departing from the spirit of the present invention.

[0144] The above-described embodiments, examples, and modifications are merely examples, and the present invention is not limited to these details as long as the features of the invention are not impaired. Furthermore, although various embodiments, examples, and modifications have been described above, the present invention is not limited to these details. Other aspects that can be considered within the scope of the technical idea of ​​the present invention are also included within the scope of the present invention.

[0145] In the above figures, the control lines and information lines shown are those that are considered necessary for explanation, and do not necessarily show all the control lines and information lines that are actually implemented. For example, it can be considered that almost all components are actually connected to each other.

[0146] The above-described layout of the various functional units, processing units, and databases of the fault tree generation system 100 is merely an example. The layout of the various functional units, processing units, and databases can be changed to an optimal layout in terms of the performance, processing efficiency, communication efficiency, etc. of the hardware and software included in the fault tree generation system 100. [Explanation of symbols]

[0147] 100: Fault tree generation system

Claims

1. 1. A fault tree generation system for generating a fault tree, comprising: comprising at least a processor and a storage device; The storage device a mathematical formula database storing mathematical formulas relating to physical phenomena; a variable database for storing variables extracted from technical documents; a variable knowledge database that stores variable knowledge, which is knowledge about variables; a variable relation database that stores variable relations that represent the relationship between variables on the left side of a mathematical expression related to the physical phenomenon and variables on the right side of the mathematical expression; and The processor: An inadequacy of a variable on the left side of a mathematical formula stored in the mathematical formula database is regarded as a malfunction event, and elements of large and small variables of the variable on the left side are generated as factors of the malfunction event, and the factors of large and small variables of the variable on the left side generated are considered to be factors when a variable on the right side of a mathematical formula stored in the mathematical formula database becomes inappropriate, and the inadequacy of the variable on the right side is generated as an element, and elements of large and small variables of the variable on the right side are generated as factors of the inadequacy of the variable on the right side generated, thereby generating a mathematical formula-based causal model which is a model representing the causal relationship of the malfunction event from the relationship between the variable on the left side and the variable on the right side of the mathematical formula stored in the mathematical formula database, generating and outputting a fault tree that combines causal relationships related to any malfunction event from among the causal relationships of the malfunction events based on the generated mathematical formula-based causal model; generating a list of variables for a mathematical formula relating to the physical phenomenon, the variables, and variable conditions that are the values ​​that the variables can take, based on the variable database and the variable knowledge database; extracting variable relationships between variables on the left side and variables on the right side of the mathematical formula from the mathematical formula related to the physical phenomenon and the generated variable list and variable conditions, and storing the extracted variable relationships in the variable relationship database; Using the generated variable conditions, differentiation and simulation are performed for the variables on the left side and the right side of the equation relating to the physical phenomenon, and based on the results of the simulation, an increase / decrease relationship of each variable with respect to the malfunction event is determined, and based on the determined increase / decrease relationship, causal relationships included in the fault tree that do not affect the malfunction event are excluded, thereby reducing the number of branches of the fault tree. Fault tree generation system.

2. The storage device further includes a defect information database for storing defect information, the processor combines, with the generated fault tree, causal relationships extracted from failure information representing information about the failure event; The fault tree generation system of claim 1 .

3. 2. The fault tree generation system according to claim 1, wherein the variable knowledge includes at least one of a physical quantity related to a physical phenomenon and a design value related to the design of a product or facility.

4. When generating the mathematical-formula-based causal model, the processor determines that the cause of the inappropriateness of a variable on the left side of the mathematical formula is a case where the variable on the left side is large and a case where the variable on the left side is small. The fault tree generation system of claim 1 .

5. When generating the mathematical-formula-based causal model, the processor determines that the cause of the inappropriateness of the variable on the right side of the mathematical formula is a case where the variable on the right side is large and a case where the variable on the right side is small. The fault tree generation system of claim 1 .

6. 2. The fault tree generation system according to claim 1, wherein the processor accepts a selection operation or an input operation for specifying a variable candidate when constructing a causal relationship related to the malfunction event.

7. 2. The fault tree generation system according to claim 1, wherein the processor accepts an input operation for specifying a variable condition when constructing a causal relationship related to the malfunction event.

8. 1. A fault tree generation method for generating a fault tree, comprising: a computer having at least a processor and a storage device, The system has at least a mathematical formula database that stores mathematical formulas related to physical phenomena, a variable database that stores variables extracted from technical documents, a variable knowledge database that stores variable knowledge that is knowledge related to variables, and a variable relation database that stores variable relations that represent the relationship between variables on the left side of mathematical formulas related to the physical phenomena and variables on the right side of the mathematical formulas, An inadequacy of a variable on the left side of a mathematical formula stored in the mathematical formula database is regarded as a malfunction event, and elements of large and small variables of the variable on the left side are generated as factors of the malfunction event, and the factors of large and small variables of the variable on the left side generated are considered to be factors when a variable on the right side of a mathematical formula stored in the mathematical formula database becomes inappropriate, and the inadequacy of the variable on the right side is generated as an element, and elements of large and small variables of the variable on the right side are generated as factors of the inadequacy of the variable on the right side generated, thereby generating a mathematical formula-based causal model which is a model representing the causal relationship of the malfunction event from the relationship between the variable on the left side and the variable on the right side of the mathematical formula stored in the mathematical formula database, generating and outputting a fault tree that combines causal relationships related to any malfunction event from among the causal relationships of the malfunction events based on the generated mathematical formula-based causal model; generating a list of variables for a mathematical formula relating to the physical phenomenon, the variables, and variable conditions that are the values ​​that the variables can take, based on the variable database and the variable knowledge database; extracting variable relationships between variables on the left side and variables on the right side of the mathematical formula from the mathematical formula related to the physical phenomenon and the generated variable list and variable conditions, and storing the extracted variable relationships in the variable relationship database; Using the generated variable conditions, differentiation and simulation are performed for the variables on the left side and the right side of the equation relating to the physical phenomenon, and based on the results of the simulation, an increase / decrease relationship of each variable with respect to the malfunction event is determined, and based on the determined increase / decrease relationship, causal relationships included in the fault tree that do not affect the malfunction event are excluded, thereby reducing the number of branches of the fault tree. Fault tree generation methods.

Images