JP6674831B2 - Causal relationship extracting apparatus, causal relationship extracting method, and causal relationship extracting program - Google Patents

Description

  The present invention relates to a causal relationship extracting device, a causal relationship extracting method, and a causal relationship extracting program.

  As a method for preventing a failure of the device, a failure tree analysis (FTA, Failure Tree Analysis) and a failure mode effect analysis (FMEA, Failure Mode and Effects Analysis) are known. The FTA systematically searches for the source of a defect by picking up a defect event of a product, sequentially identifying the cause thereof, and developing the defect in a hierarchical manner. The product of the FTA is a "fault tree" having a product failure event at its top event and a cause at a lower hierarchy.

  The FMEA predicts a failure mode that can occur in a part when designing the part, evaluates the influence of the failure mode on a subsystem in a higher layer of the part, and evaluates the influence of the failure mode on a system in a higher layer. Evaluate and take precautionary measures for failure modes that have a large impact. The FMEA deliverable is an “FMEA table” that includes part names, failure modes, effects on subsystems, effects on systems, measures, and the like.

  As described above, the FTA is a method of tracing a causal relationship leading to a top event from a result side to a cause side in a top-down manner. It can be said that it is a technique that follows an expression. However, both when creating a fault tree and when creating an FMEA table, it is premised that samples of causal relationships of faults that have occurred in the past are stored in a type that can be easily referred to. Only with such a sample, a fault tree can be efficiently created in a top-down manner, and an FMEA table can be efficiently created in a bottom-up manner.

  As an example of a technique for extracting a causal relationship from past examples of failure information, there is a failure occurrence cause extraction device in Patent Document 1. The device extracts a causal relationship from the defect information (FIG. 6 of Patent Document 1) by using a clue expression (FIG. 7 of Patent Document 1) as a search key. Specifically, using the clue expression "A when B", the defect information "If your building is reinforced concrete, radio waves may not be received", the cause "The building you live in is reinforced concrete", Then, the result "no radio wave" is extracted.

WO 2014/020833

  The failure occurrence cause extraction device of Patent Document 1 is easily affected by fluctuations in expression, and may not be able to extract a causal relationship depending on the expression of defect information. For example, suppose that the defect information is "radio waves may not be received. The cause is that your house is reinforced concrete." This defect information has exactly the same meaning as the above-mentioned "If the house you live in is reinforced concrete, radio waves may not be received". However, even if the clue expression “A when B is used” is used, the cause and the result cannot be extracted because “the radio wave may not be received. The cause is that the building in which you live is reinforced concrete.”

  In addition, there are many cases where a plurality of intermediate events are chained and triggered by a certain cause, resulting in a final failure. In other words, in the above example, some intermediate event 1 is caused by “the resident building is reinforced concrete”, another intermediate event 2 is caused by the intermediate event 1, and finally “the radio wave is received by the intermediate event 2”. No ".

However, the failure occurrence cause extraction device of Patent Literature 1 targets only the causal relationship between the triggering cause and the final failure. Therefore, even if an intermediate event between the cause of the trigger and the final defect is described in the defect information, it is difficult to extract the intermediate event. In combination with the above two points, the failure occurrence cause extraction device of Patent Literature 1 accumulates a wide and large amount of causal relationships, which makes it difficult to efficiently search for the causal relationships without omission. Was.
Thus, the present invention can extract a causal relationship chain including an intermediate event between a triggering cause and a final defect from the defect information, and can extract a causal relationship that is not easily affected by fluctuations in the expression of the defect information. It is intended to provide a device.

The causal relationship extraction device according to the present invention includes a storage unit that stores defect information and a hierarchical structure of components, which are examples of defects, a component / phenomenon extraction unit that extracts words indicating components and words indicating phenomena from the defect information. , The word indicating the part and the word indicating the phenomenon are combined, and for each combination of the word indicating the part and the word indicating the phenomenon, the degree of association between the word indicating the part and the word indicating the phenomenon is calculated. And a co-occurrence relation evaluation unit that determines a plurality of highly relevant combinations as causal elements based on the component, and specifies a hierarchy in the component hierarchical structure for the parts included in the causal element, and ranks the causal elements higher in the hierarchy. by ranking the order in, creating a chain of elements causal consisting only of elements of causality includes a causal relationship extraction unit that stores the chain of elements of causality in the storage unit, the And wherein the door.
Other means will be described in the embodiments for carrying out the invention.

  According to the present invention, it is possible to extract a chain of causal relations including an intermediate event between a triggering cause and a final defect from the defect information, and to extract a causal relation that is hardly affected by fluctuation of the expression of the defect information. It becomes possible to provide a device.

It is a figure explaining the composition of a causal relation extraction device. (A) is a figure explaining a causal relationship. (B) is a figure explaining a more concrete causal relationship. (A) is a figure explaining fault information. (B) is a figure explaining a parts list. (C) is a diagram illustrating a phenomenon list. FIG. 3 is a diagram illustrating a parts tree. It is a figure showing an example of causal relation information. It is a flowchart explaining a processing procedure. It is an example of an evaluation result table. It is an example of a failure tree creation screen. It is an example of a top event registration screen. It is an example of a failure tree creation screen. It is an example of the search result of a causal relationship. It is an example of an intermediate event registration screen. It is an example of a failure tree creation screen. It is an example of an FMEA table creation screen. It is an example of an FMEA table creation screen. It is an example of a calculation result of values such as IEPH and IEPA. It is an example of the result of having calculated four types of co-occurrence relation evaluation indexes.

  Hereinafter, four embodiments for carrying out the present invention will be described in detail with reference to the drawings and the like. Four embodiments consist of the first to fourth embodiments. The details will be described later, but the difference between them is the difference in information processing when the causal relationship extraction device extracts a chain of causal components from the defect information. The first embodiment includes a process of extracting a chain of causal elements and a process of creating a fault tree and an FMEA table using the extracted chain of causal elements.

<First embodiment>
(Equipment configuration)
The configuration of the causal relationship extraction device 2 will be described with reference to FIG. The causal relationship extracting device system 1 includes a causal relationship extracting device 2 and a terminal device 3. These can be connected via the network 4. The causal relationship extracting device 2 is a general computer, and includes a central control device 11, an input device 12, an output device 13, a main storage device 14, and an auxiliary storage device 15. These are interconnected by a bus. The auxiliary storage device 15 stores defect information 31, a component list 32, a phenomenon list 33, a component tree 34, a causal relationship information 35, a failure tree 36, and an FMEA table 37 (details will be described later). The auxiliary storage device 15 is an external storage device independent of the causal relationship extraction device 2, and a configuration in which both can be connected via the network 4 is also possible.

  The part / phenomenon extraction unit 21, co-occurrence relationship evaluation unit 22, causal relationship extraction unit 23, fault tree creation unit 24, and FMEA table creation unit 25 in the main storage device 14 are programs. Hereinafter, when the subject is described as “XX section”, the central control device 11 reads out each program from the auxiliary storage device 15, loads it into the main storage device 14, and then functions of each program (details described later). Shall be realized.

(Causal relationship)
The causal relationship will be described with reference to FIG. The causal relationship 51 is composed of three or more causal elements 52a, 52b, 52c,. The elements 52a and the like are connected by directed segments 53a, 53b,. Of the set (two) elements connected by one directed segment, the element on the starting side of the directed segment is the cause, and the element on the end point is the result. For example, element 52a is the result from element 52b, and element 52b is the cause from element 52a. Similarly, element 52b is the result from element 52c, and element 52c is the cause from element 52b. Note that the starting point side (cause side) of the directed line segment may be referred to as “upstream” and the end point side (result side) may be referred to as “downstream”.

  Element 52a has one part 54a and one phenomenon 55a. Element 52b has one part 54b and one phenomenon 55b. Element 52c has one part 54c and one phenomenon 55c. The auxiliary storage device 15 stores a plurality of causal relationships 51, and each of the causal relationships 51 has a causal relationship ID 56 that uniquely identifies itself. For example, the causal relationship 51 in FIG. 2A indicates the following as a whole.

The phenomenon B occurs in the part B due to the phenomenon C occurring in the part C.
The phenomenon A occurs in the component A due to the phenomenon B occurring in the component B.
In the above description, each "part" 54a or the like of the element 52a or the like means a group of physical units in which the phenomenon occurs. That is, the “parts” here are parts in a broad sense, and are a concept including all products, systems constituting the products, subsystems constituting the systems, parts constituting the subsystems (in a narrow sense), and the like.

A more specific causal relationship will be described with reference to FIG. The causal relationship 57 indicates the following as a whole.
-A short circuit occurs in the power supply circuit due to the generation of whiskers in electronic components.
-An output failure occurs in the power supply due to a short circuit in the power supply circuit.
・ Because of the output failure of the power supply, the product cannot be started.

(Defect information)
The defect information 31 will be described with reference to FIG. The defect information 31 is a report in which a defect of a device that has occurred in the past is described as a sentence for each item. Specifically, the defect information 31 includes items such as a defect name 31a, a progress 31b until the occurrence of the defect, a phenomenon 31c (for the defect), a cause 31d (for the defect), and a countermeasure 31e (for the defect). Each item has a sentence (character string). A plurality of defect information 31 about a plurality of defects are stored in the auxiliary storage device 15.

(Parts list)
The component list 32 will be described with reference to FIG. The component list 32 stores (broadly defined) components specified in advance by the user of the causal relationship extraction device 2.

(Phenomenon list)
The phenomenon list 33 will be described with reference to FIG. The phenomenon list 33 stores the phenomenon specified by the user in advance.

(Parts tree)
The component tree 34 will be described with reference to FIG. The parts tree 34 is a tree that indicates a hierarchical structure of parts in a broad sense, specified in advance by the user. FIG. 4 illustrates the following.
-"XX Product" exists at the highest level.
-The OO product has at least a "power supply", a "control unit", and a "housing" in the immediately lower hierarchy.
The “power supply” has at least a “power supply circuit” in the immediately lower hierarchy.
The "power supply circuit" has at least "electronic components" in the immediately lower hierarchy.
Preferably, the number of layers in the component tree 34 is three or more. This is because by doing so, a causal relationship in which three or more elements are linked can be created (details described later). The auxiliary storage device 15 stores a plurality of component trees 34. It is desirable that there be one parts tree 34 for each product.

(Causal information)
The causal relationship information 35 will be described with reference to FIG. In the causal relationship information 35, in association with the causal relationship ID stored in the causal relationship ID column 101, the element ID column 102 stores the element ID, the component column 103 stores the component, the phenomenon column 104 stores the phenomenon, and the downstream. The column 105 stores the element ID of the downstream element, and the upstream column 106 stores the element ID of the upstream element.

The causal relationship ID in the causal relationship ID column 101 is an identifier for uniquely specifying the causal relationship, as described above.
The element ID in the element ID column 102 is an identifier for uniquely specifying an element.
The parts in the parts column 103 are the names of parts in a broad sense.
The phenomena in the phenomena column 104 are phenomena (mainly defects) occurring in parts in a broad sense.
The element ID of the downstream element in the downstream column 105 is the element ID of an element that is a direct result of the element. "-" Is stored in the row for the most downstream element (final result).
The element ID of the upstream element in the upstream column 106 is the element ID of the element that directly causes the element. "-" Is stored in the row for the most upstream element (the cause of the trigger).

  As is clear from the above, the causal relationship information 35 represents the causal relationship (FIGS. 2A and 2B) in a table format. Note that, for simplicity of description, the causal relationship (FIGS. 2A and 2B) and the causal relationship information 35 are illustrated without branching and merging. However, the causal relationship may have branching and merging, and it goes without saying that the causal relationship information 35 can express branching and merging. Further, the number of elements having one causal relationship is not limited, but is usually three or more. The causal relationship information 35 stores a plurality of causal relationships.

(Processing procedure)
The processing procedure of the present embodiment will be described with reference to FIG. During the description of FIG. 6, FIGS. 7 to 15 will be referred to as appropriate.
In step S201, the component / phenomenon extraction unit 21 extracts components and phenomena from the defect information 31 (FIG. 3A). Specifically, first, the component / phenomenon extraction unit 21 extracts components (in a broad sense) described in the component list 32 (FIG. 3B) from all items of the defect information 31. In this case, a total of five items of “XX product”, “power supply”, “power supply circuit”, “electronic component” and “base” are extracted from the defect information 31. Note that the part / phenomenon extraction unit 21 morphologically analyzes the document of the defect information 31 and divides the document into words as pre-processing of the processing.

  Second, the part / phenomenon extraction unit 21 extracts the phenomena described in the phenomena list 33 (FIG. 3C) from all items of the defect information 31. In this case, a total of four items, “startup impossible”, “output defect”, “whisker”, and “short”, are extracted from the defect information 31.

  In step S202, the co-occurrence relation evaluation unit 22 calculates the degree of association between the part and the phenomenon. Specifically, first, the co-occurrence relation evaluation unit 22 creates a combination of the component extracted in “first” in step S201 and the phenomenon extracted in “second” in step S201. In this case, 5 × 4 = 20 combinations such as “(XX product, cannot be activated)” are created.

  Second, the co-occurrence relation evaluation unit 22 obtains, as the degree of association, a value of the result of evaluating the co-occurrence relation between the part and the phenomenon for each combination of the part and the phenomenon. Generally, a co-occurrence relation is a relation in which two words appear simultaneously in a certain document. Here, the co-occurrence relationship is a relationship in which, for example, “XX product” and “startup impossible” simultaneously appear in the defect information 31. Although there are a plurality of methods for evaluating such co-occurrence relations, here, the co-occurrence relation evaluation unit 22 uses the number of characters between the component and the phenomenon as the degree of association.

  For example, among the items of the defect information 31 (FIG. 3A), the defect name 31a has a description of "XX product cannot be activated". Here, the number of characters between the component “XX product” and the phenomenon “startup impossible” is “0”. Among the items of the defect information 31 (FIG. 3A), the phenomenon 31c includes a description that "the XX product cannot be started." Here, the number of characters between the component “XX product” and the phenomenon “startup impossible” is “1”. In this way, the co-occurrence relation evaluation unit 22 acquires one or a plurality of characters, and sets the smallest number as the relevance. The smaller the value of the degree of association, the higher the association between the component and the phenomenon. The co-occurrence relation evaluation unit 22 repeats the “second” process of step S202 for all combinations of parts and phenomena.

In step S203, the co-occurrence relation evaluation unit 22 determines a causal relation element. Specifically, first, the co-occurrence relation evaluation unit 22 associates the combination created in “first” in step S202 with the evaluation result table storing the degree of association obtained in “second” in step S202. Create
FIG. 7 is an example of the evaluation result table. FIG. 7 shows the following.
-There are five combinations of the phenomenon having the phenomenon "cannot be started" and the phenomenon.
There are five combinations of parts having the phenomenon "output failure" and the phenomenon, and the same applies to other phenomena.
-The relevance is calculated for each of the 20 combinations in total.

  Second, the co-occurrence relation evaluation unit 22 determines, as a causal relation element, a combination having the smallest relevance value from a group of combinations having the same phenomenon. In the example of FIG. 7, there are five combinations in the group of combinations having the phenomenon "cannot be activated", and there are also five values of the degree of association. The smallest value among them is “0”. Therefore, “XX product cannot be activated” is acquired as the causal element. Similarly, “power supply output failure”, “electronic component whisker”, and “power supply circuit short” are determined from other groups. It should be noted that the order of “startup disabled, output failure,...” Described in the phenomenon column of FIG. 7 and the order of XX product, power supply,. (The order is determined in the next step according to the hierarchy of the parts tree 34).

In step S204, the causal relationship extracting unit 23 orders the elements of the causal relationship. Specifically, first, the causal relationship extracting unit 23 reads out the component tree 34 (FIG. 4) from the auxiliary storage device 15.
Second, the causal relationship extracting unit 23 extracts the components “XX product”, “power supply”, “electronic component”, and “power supply circuit” included in the four elements of the causal relationship determined in “second” in step S203. The hierarchy of the parts tree 34 to which each belongs is specified. In the example of FIG. 4, the first hierarchy (highest hierarchy) for “xx product”, the second hierarchy for “power supply”, the fourth hierarchy for “electronic components”, and the third hierarchy for “power supply circuit” Is specified.

  Third, the causal relationship extracting unit 23 orders the elements of the causal relationship based on the hierarchy specified in the “second” of step S204. That is, the combinations are ordered in the order in which the hierarchy is higher. In the above example, in order from the highest order, the order is “XX product cannot be started”, “power supply output failure”, “power supply circuit short”, and “electronic component whisker”. Expressing this order using a directed line segment results in “XX product cannot be started ← power supply output failure ← power supply circuit short ← electronic component whisker”. In this way, the causal relationship extracting unit 23 orders the plurality of elements as a chain of causal components from components including lower-level components to components including higher-level components.

  Fourth, the causal relationship extracting unit 23 creates the causal relationship information 35 (FIG. 5) based on the causal relationship elements ordered in the “third” of step S204, and stores it in the auxiliary storage device 15. As described above, in the present embodiment, the clue expression is not used, and the co-occurrence relationship of words and the component tree (the hierarchical relationship of components) are used. Therefore, according to the present embodiment, a chain of causal elements can be extracted without being affected by the fluctuation of the expression.

  In FIG. 6, the processing procedure for creating the fault tree and the FMEA table continues after step S204. Prior to the description of these steps, first, the problem when the user creates the fault tree and the FMEA table will be described, and then the continuation of FIG. 6 will be described.

(FTA)
In the FTA, a user determines a “failure tree” by sequentially determining intermediate events (what components generate what phenomena) in a top-down manner from a top event toward a cause side. To complete. The top event is often an event relating to the component at the highest level. This is because a defect is often first recognized by a user as a defect as a product. The top event often occurs at the end (downstream) on the result side of the causal relationship.

The user recognizes that a component at a certain higher level has some kind of defect, and that a defect at a component at a higher level may be caused by a defect at a component at a lower level. ing. However, the user does not know specifically what kind of lower layer and what kind of component has caused any trouble.
Determining an intermediate event without any clues is difficult, especially for unskilled users. Therefore, the causal relationship extracting device 2 provides the user with a chain of the elements of the causal relationship as clues.

(FMEA)
In FMEA, a user determines a failure mode assumed for a component starting from a component in a lower hierarchy. Then, in a bottom-up manner from the combination of the part and the failure mode toward the result side, the effect of the failure mode of the part on the middle-level parts, the influence on the higher-level parts, and countermeasures. Then, the FMEA table is completed. The failure mode is a classification based on a type of a failure state such as disconnection, short circuit, breakage, wear, and deterioration of characteristics.

  It is assumed that the user designs a coil, for example. For an experienced user, it may be easy to come up with a failure mode such as “disconnection”, “short”, or “abnormal resistance” as a failure mode of the coil. However, it is not easy for an unskilled user to come up with a specific failure mode. Even if a specific failure mode is conceived, it is possible to come up with what kind of phenomenon (for example, output failure) and what kind of higher-level component (for example, power supply) it affects. It is difficult. Therefore, the causal relationship extracting device 2 provides the user with a chain of the elements of the causal relationship as clues.

Referring back to FIG. The subsequent steps S205 to S211 are processing in which the failure tree creation unit 24 supports a user who wants to create the failure tree 36.
In step S205, the fault tree creation unit 24 displays a fault tree creation screen 61 (FIG. 8). Specifically, first, upon receiving an arbitrary instruction from the user, the fault tree creation unit 24 displays a fault tree creation screen 61 on the output device 13. The failure tree creation screen 61 has a title area 62, a failure tree area 63, and a menu area 64. The menu area 64 has a top event registration button 64a, a causal relationship search button 64b, an intermediate event registration button 64c, a failure tree registration button 64d, and a cancel button 64e.

  Second, the failure tree creation unit 24 accepts that the user presses the top event registration button 64a after inputting a title in the title area 62 via the input device 12 such as a keyboard. Here, the user has entered “FTA related to“ XX product cannot be activated ”” as the title, and the failure tree area 63 remains blank.

In step S206, the fault tree creating unit 24 displays a top event registration screen 71 (FIG. 9). Specifically, first, the fault tree creating unit 24 displays the top event registration screen 71 on the output device 13. The top event registration screen 71 has a parts column 72a, a phenomenon column 72b, a registration button 73a, and a cancel button 73b.
Second, the failure tree creating unit 24 accepts that the user inputs a component of the top event in the component column 72a, inputs a phenomenon of the top event in the phenomenon column 72b, and then presses the registration button 73a. Here, the user has input “XX product” in the component column 72a and “startup impossible” in the phenomenon column 72b.

  In step S207, the fault tree creation unit 24 displays the top event on the fault tree creation screen 61 (FIG. 10). At this time, the failure tree creation unit 24 displays the element “XX product activation impossible” 63a received in “second” in step S206 in the failure tree area 63 of the failure tree creation screen 61 in FIG. Note that FIG. 10 and FIG. 13 described below are the results of the transition of the failure tree creation screen 61 of FIG.

In step S208, the fault tree creating unit 24 extracts a causal relationship. Specifically, first, the fault tree creation unit 24 accepts that the user selects the element 63a on the fault tree creation screen 61 (FIG. 10) and then presses the causality search button 64b.
Second, the failure tree creating unit 24 searches the causal relationship information 35 (FIG. 5) using the element “XX product cannot be activated” 63a as a search key, and has a component “XX product” and a phenomenon “cannot be activated”. Search for causal relationships that include elements. Then, the failure tree creating unit 24 temporarily stores all the causal relationships corresponding to the search result in the main storage device 14.

Third, the failure tree creating unit 24 displays the search result 75 (FIG. 11) on the output device 13. The search result 75 is a list of the causal relationships corresponding to the “second” in step S208, and describes the components and phenomena of each of the lowest elements in the causal relationships as headings.
Fourth, the failure tree creating unit 24 accepts that the user selects any causal relationship (any row) from the search results 75. Now, it is assumed that the user has selected “XX product cannot be activated” on the first line.

  Fifth, the fault tree creating unit 24 displays the causal relationship accepted in the “fourth” of step S208 on the output device 13 (reference numeral 76). The user knows that the direct factor that causes “XX product cannot be activated” is “power supply output failure”. The causal relationship 76 in FIG. 11 has “XX product cannot be activated” at the most downstream. However, a causal relationship 76 having “XX product cannot be activated” at a position other than the most downstream may be a search result.

In step S209, the fault tree creating unit 24 displays the intermediate event registration screen 81 (FIG. 12). Specifically, first, the fault tree creating unit 24 accepts that the user selects the top event 63a on the fault tree creating screen 61 (FIG. 10) and then presses the intermediate event registration button 64c.
Second, the fault tree creating unit 24 displays an intermediate event registration screen 81 on the output device 13. The intermediate event registration screen 81 has a component column 82a, a phenomenon column 82b, a registration button 83a, and a cancel button 83b.

  Third, the failure tree creating unit 24 inputs the part of the intermediate event immediately upstream of the top event into the parts column 82a, inputs the phenomenon of the intermediate event into the phenomenon column 82b, and then presses the registration button 83a. Accept to do. Here, the user has input "power" in the parts column 82a and "output failure" in the phenomenon column 82b.

  In step S210, the fault tree creation unit 24 displays an intermediate event on the fault tree creation screen 61 (FIG. 13). At this time, the failure tree creating unit 24 displays the element “power supply output failure” 63b received in “third” in step S209, immediately upstream of the element 63a in the failure tree area 63.

  After that, the failure tree creating unit 24 repeats the processing of steps S208 to S210. However, in the repetitive processing, the user selects the element displayed in the uppermost stream (lowest hierarchy) in the fault tree area 63 of the fault tree creation screen 61 (FIG. 13), and then presses the causality search button 64b. It shall be pressed. Then, the fault tree creating unit 24 continuously displays the causal elements 63c, 63d,... From the downstream side to the upstream side (from left to right) of the fault tree area 63. Become.

In step S211, the fault tree creating unit 24 registers the fault tree 36. When the user determines that a sufficient number of elements 63a, 63b,... Are displayed, the user presses the failure tree registration button 64d. Then, the fault tree creating unit 24 stores the continuous elements displayed in the fault tree area 63 in the auxiliary storage device 15 as the fault tree 36.
The fault tree 36 (reference numerals 63a to 63d) in FIG. 13 looks like the causal relationship 76 in FIG. 11 at first glance. However, these are different. The user can view a plurality of causal relationships 76 by selecting a plurality of rows of the search result 75 in FIG. The user sequentially selects and links elements considered to be useful from among the plurality of causal relationships 76. The result of such an operation is the fault tree 36.

The following steps S212 to S214 are processing in which the FMEA table creation unit 25 supports a user who wants to create the FMEA table 37.
In step S212, the FMEA table creation unit 25 displays the FMEA table creation screen 91 (FIG. 14). Specifically, the FMEA table creation unit 25 displays an FMEA table creation screen 91 on the output device 13 when receiving an arbitrary instruction from the user. The FMEA table creation screen 91 has a title area 92, an FMEA table area 93, and a menu area 94. The menu area 94 has a causal relationship search button 94a, an FMEA table registration button 94b, and a cancel button 94c. Note that the FMEA table creation unit 25 accepts a user inputting a title (for example, “FMEA for“ XX product ”) in the title area 92.

The table displayed in the FMEA table area 93 is the FMEA table.
The parts in the parts column 93a are parts that are designed by the user.
The function in the function column 93b is a function of a component.
The failure modes in the failure mode column 93c are, for example, classifications based on types of failure states such as disconnection, short circuit, breakage, wear, and deterioration of characteristics.
The probable cause in the presumed cause column 93d is that which is presumed to be the cause of the occurrence of the failure mode in the component.

The effect of the failure in the failure effect column 93e is the effect of the occurrence of a failure mode on a component in a higher-level component. There are many hierarchies in the upper hierarchy. In this case, sub-columns for the middle “unit” 93f and the highest “product” 93g are provided, and the effect is stored in each sub-column.
The detection method in the detection method column 93h is a specific method for detecting occurrence of a failure mode.
The failure grade in the failure grade column 93i is an index or the like indicating the severity of the failure mode.
The countermeasure in the countermeasure column 93j is a specific countermeasure for avoiding the occurrence of the failure mode.

In step S213, the FMEA table creation unit 25 extracts a causal relationship. Specifically, first, the FMEA table creation unit 25 inputs a component into the component column 93a of the FMEA table creation screen 91 (FIG. 14), selects the input component, and then executes a causality search button 94a. Accepts pressing. Here, it is assumed that the “power supply circuit” is input.
Second, the FMEA table creator 25 searches the causal relationship information 35 (FIG. 5) using the component "power supply circuit" as a search key, and searches for a causal relationship including an element having the component "power supply circuit". Then, the FMEA table creator 25 temporarily stores all the causal relationships corresponding to the search result in the main storage device 14.

Third, the FMEA table creator 25 displays the search result 75 (FIG. 11) on the output device 13. The search result 75 is a list of the causal relations corresponding to the “second” in step S213, and describes the lowermost parts and phenomena of each of the causal relations as headings.
Fourth, the FMEA table creator 25 accepts that the user selects an arbitrary causal relationship (arbitrary row) from the search results 75. Now, it is assumed that the user has selected “XX product cannot be activated” on the first line.

Fifth, the FMEA table creator 25 displays the causal relationship accepted in the "fourth" of step S213 on the output device 13 (reference numeral 76). The user sees the causal relationship 76 and recognizes, for example, the following.
-There may be a "short" as a failure mode of the component "power supply circuit".
-"Electronic component whisker" may be the cause of "power supply circuit short".
-The phenomenon "output failure" may occur in the component "power supply" of the immediately higher hierarchy due to the "power supply circuit short".
-The phenomenon "startup impossible" may occur in the component "xx product" at the highest level due to "power output failure".

  In step S214, the FMEA table creation unit 25 registers the FMEA table 37. Specifically, first, the FMEA table creation unit 25 accepts that the user inputs a failure mode or the like to the FMEA table area 93 of the FMEA table creation screen 91 (FIG. 15). FIG. 15 shows the result of the transition of the FMEA table creation screen 91 of FIG. The user indicates “short” in the failure mode column 93c, “whisker in electronic component” in the probable cause column 93d, “power supply output failure” in the unit sub-column 93f, and “XX product cannot be started” in the product sub-column 93g. Is entered. These are input by the user with reference to the causal relationship 76 (FIG. 11).

  Second, the FMEA table creator 25 accepts a user input to the FMEA table area 93 other than a failure mode or the like. The user inputs data into the function column 93b, the detection method column 93h, the failure class column 93i, and the countermeasure column 93j as appropriate. Since the user has already input the failure mode or the like in the “first” of step S214, it is easy to input the function after considering the function or the like.

Third, the FMEA table creation unit 25 accepts that the user presses the FMEA table registration button 94b. Then, the FMEA table creation unit 25 stores the table displayed in the FMEA table area 93 in the auxiliary storage device 15 as the FMEA table 37.
Thereafter, the processing procedure ends.

  The processing of steps S205 to S211 may be omitted. Also, the processing of steps S212 to S214 may be omitted. Further, the processing of steps S205 to S211 may be performed after the processing of steps S212 to S214. Note that the cancel button 64e and the like are pressed to cancel the immediately preceding input.

(Auto-fill)
In the above description, the user manually inputs characters into the component column 72a and the phenomenon column 72b of the top event registration screen 71 (FIG. 9) and the component column 82a and the phenomenon column 82b of the intermediate event registration screen 81 (FIG. 12). Examples have been described. However, the failure tree creating unit 24 may automatically input a component and a phenomenon into each column based on the causality 76 selected after the search. Further, the failure tree creation unit 24 searches the causal relationship 76 selected after searching (without waiting for the user to press the registration button 73a by displaying the top event registration screen 71) and displays the causal relationship 76 in the failure tree creation screen 61 on the failure tree creation screen 61. It may be displayed directly in the tree area 63.

  In the above, an example has been described in which the user manually inputs characters into the failure mode column 93c and the like of the FMEA table creation screen 91 (FIG. 15). However, the FMEA table creation unit 25 may automatically input a failure mode or the like in each column based on the causality 76 selected after the search. That is, the FMEA table creator 25 automatically recognizes that the user has recognized in the “fifth” of step S213, and automatically performs the input operation performed by the user in the “first” of step S214.

<Second embodiment>
In the first embodiment, in step S201, the component / phenomenon extraction unit 21 uses the component list 32 and the phenomenon list 33 when extracting components and phenomena from the defect information 31. In the second embodiment, the component / phenomenon extraction unit 21 extracts a component-like word and a phenomenon-like word without using the component list 32 and the phenomenon list 33.

(Parts-like, phenomenon-like)
The part / phenomenon extraction unit 21 calculates two values “IEPH” and “IEPA” for each word constituting the defect information 31. The value of IEPH indicates the possibility that the word belongs to the following expression tendency 1 (possibility of a phenomenon). The value of IEPA indicates the possibility that the word belongs to the following expression tendency 2 (possibility of a part).

(Expression tendency 1) An expression tendency in which the word is determined to be a phenomenon when a word such as “occurs” or “occurs” immediately follows the word. For example, such a word often follows immediately after a phenomenon such as "residual stress" and "high temperature degradation".
(Expression tendency 2) An expression tendency in which a word such as "failure" or "purchase" immediately follows the word to determine that the word is a part. For example, such words often immediately follow components such as "impeller" and "condenser".
A calculation formula for calculating “IEPH” and “IEPA” will be described below. Although IEPH and the like are abbreviations, they do not include generalized words, so detailed spelling is omitted. That is, "IEPH" and the like may be understood as simple symbols.

(IEPH)
・ IEPH = PHL × TF × IDF
・ PHL = PHDF / TDF
・ IDF = log (K / DF)
PHDF is a case where, out of all the defect information 31 stored in the auxiliary storage device 15, the word appears immediately after that with the word “occurred”, “occurred” or “occurred”. This is the number of defect information that has at least once.
The TDF is the number of pieces of defect information having at least one occurrence of the word out of all pieces of defect information 31 stored in the auxiliary storage device 15.
TF is the number of appearances of the word in the defect information.
K is the total number of all the fault information 31 stored in the auxiliary storage device 15.
DF is the number of pieces of defect information including the word out of all the defect information 31 stored in the auxiliary storage device 15.
Note that “log (K / DF)” is a natural logarithm (or a common logarithm) of “K / DF”.

(IEPA)
・ IEA = PAL × TF × IDF
・ PAL = PALDF / TDF
・ IDF = log (K / DF)
The paldf, out of all the defect information 31 stored in the auxiliary storage device 15, immediately after the word, the word “failure”, “purchase”, “acceptance”, “develop” or “develop” "Is the number of pieces of defect information having at least one occurrence of the term"".
-TF, TDF, K and DF are the same as those described for "IEPH".

  FIG. 16 shows the result of the component / phenomenon extraction unit 21 calculating values such as “IEPH” and “IEPA” for each of the words constituting the defect information 31. Referring to the IEPH column 95a of FIG. 16, it can be seen that the IEPH values for "startup disabled", "output failure", "whisker", and "short" are significantly larger than the others. Therefore, the part / phenomenon extraction unit 21 extracts these as phenomena. Referring to the IEPA column 95b of FIG. 16, it is found that the IEPA values for “xx product”, “power supply”, “power supply circuit”, “electronic component” and “base” are significantly larger than the others. Recognize. Therefore, the part / phenomenon extraction unit 21 extracts these as parts.

(IEPH and IEPA thresholds)
The part / phenomenon extraction unit 21 may extract a word having a value of IEPH greater than a certain threshold value as a phenomenon. Similarly, the component / phenomenon extraction unit 21 may extract, as a component, a word whose IEPA value is larger than another threshold value. Examples of how to determine these thresholds are as follows.

The part / phenomenon extraction unit 21 sets the average value of the IEPH (IEPA) values of all the words constituting the defect information 31 as a threshold value of the IEPH (IEPA).
The user classifies all words constituting one or a plurality of defect information 31 as a sample into “phenomenon” or “parts”. The part / phenomenon extraction unit 21 calculates the value of IEPH and the value of IEPA for all the words constituting the defect information 31. The part / phenomenon extraction unit 21 sets the smallest one of the IEPH values calculated for the word that is the phenomenon as the IEPH threshold. The component / phenomenon extraction unit 21 sets the smallest one of the IEPA values calculated for the component word as the IEPA threshold value.

<Third embodiment>
In the first embodiment, in step S202, the co-occurrence relation evaluation unit 22 uses the number of characters between the component and the phenomenon as the degree of relevance when calculating the degree of association between the component and the phenomenon. In the third embodiment, the co-occurrence relation evaluation unit 22 may calculate any of the following co-occurrence relation evaluation indexes # 1 to # 5. These co-occurrence relation evaluation indexes are indexes used in natural language processing.

# 1: co-occurrence frequency = frequency in which the central word and co-occurrence word appear simultaneously in the span in the defect information to be extracted # 2: appearance frequency ratio = co-occurrence frequency / co-occurrence word frequency # 3: T value = (Co-occurrence frequency- (center word frequency × co-occurrence word frequency) / total number of corpus words) / √co-occurrence frequency # 4: MI value = log ₂ ((co-occurrence frequency × total number of corpus words) / (center word frequency × Co-occurrence frequency))
# 5: Overall evaluation index = a × co-occurrence frequency + b × appearance frequency ratio + c × T value + d × MI value

  Here, the “central word” is a word to be evaluated, and in the present embodiment, corresponds to a phenomenon. The “co-occurrence word” is a word that appears together with the central word, and corresponds to a component in the present embodiment. The “span” is the number of characters before and after a predetermined number around the central word. “Co-occurrence word frequency” is the total number of appearances of co-occurrence words in the target defect information. The “center word frequency” is the total number of appearances of the center word in the target defect information. “Corpus total number of words” is the total number of words in the target defect information. a, b, c and d are weights.

# 1 is the simplest index. If the frequency at which parts and phenomena appear simultaneously in the span is high, the co-occurrence relation evaluation unit 22 determines that the co-occurrence relation is strong. However, in # 1, if the co-occurrence word frequency is high, the co-occurrence frequency naturally increases, and it becomes impossible to evaluate the co-occurrence relationship equally between a word having a high co-occurrence word frequency and a word having a low co-occurrence frequency.
Therefore, in # 2, by dividing the co-occurrence frequency by the co-occurrence word frequency, the co-occurrence relationship can be evaluated regardless of the co-occurrence word appearance frequency.

# 3 is an index that takes into account not only the frequency of co-occurring words but also the total number of words in the corpus, and is an index that is diverted from statistics and measures the statistical superiority of the co-occurrence relationship between two words.
# 4 is also an index that takes into account not only the co-occurrence word frequency but also the total number of corpus words, and is an index that is born from information theory and indicates the frequency at which a certain word and a partner word appear in close proximity.
# 5 is an index integrating # 1 to # 4.

  FIG. 17 shows the result of the co-occurrence relation evaluation unit 22 calculating the co-occurrence relation evaluation indexes # 1 to # 4 for a combination of a total of 20 parts and phenomena. The co-occurrence relation evaluation unit 22 extracts a part having the largest co-occurrence relation evaluation index for each of the phenomena. When the co-occurrence relation evaluation unit 22 uses, for example, “co-occurrence frequency”, the part “XX product” is extracted for the phenomenon “cannot be started”, and the phenomenon “output failure”, “whisker”, and “short” are detected. In the figure, the same component "power supply circuit" is redundantly extracted.

  The component to be extracted differs depending on which co-occurrence relation evaluation index is used. Therefore, the user acquires a combination (answer) of a part and a phenomenon that have actually occurred from all words constituting one or a plurality of defect information 31 as a sample. The co-occurrence relation evaluation unit 22 repeats the process of extracting parts for all phenomena included in the defect information 31 for the co-occurrence relation evaluation indexes # 1 to # 4. The co-occurrence relation evaluation unit 22 compares the extracted part with the answer part and obtains a number that matches both. A co-occurrence relation evaluation index that maximizes the matching number is determined as a co-occurrence relation evaluation index to be adopted.

<Fourth embodiment>
In the first embodiment, in step S202, the co-occurrence relation evaluation unit 22 uses the co-occurrence relation in one piece of target defect information 31 when calculating the degree of association between a part and a phenomenon. . On the other hand, in the fourth embodiment, the co-occurrence relation evaluation unit 22 uses the co-occurrence relation among many pieces of defect information 31 stored in the auxiliary storage device 15.

  There is a general tendency between components and the phenomena that can occur in the components. For example, phenomena that can occur in the component “capacitor” include “short circuit”, “unstable”, “noise”, and the like. Phenomena that may occur in the component “piping” include “clogging”, “leakage”, “thinning”, and the like. Therefore, it can be said that the component corresponding to the phenomenon "short circuit" is more appropriate for the "condenser" than for the "piping". The co-occurrence relation evaluation unit 22 obtains such a general tendency from among many pieces of defect information 31 stored in the auxiliary storage device 15 and uses it when extracting a combination of a part and a phenomenon.

Now, as another example of the co-occurrence relation evaluation index for a combination of a certain part and a certain phenomenon, a “trend evaluation index” of # 6 is defined.
# 6: Trend evaluation index = f (parts and phenomena) / f (parts or phenomena) x f (parts near N phenomena) / f (parts and phenomena)

  Here, f (parts and phenomena) is the number of relevant defect information 31 in the case where all the defect information 31 stored in the auxiliary storage device 15 is searched under the search condition “parts and phenomena”. Similarly, f (parts or phenomena) is the number of relevant defect information 31 in the case where all the defect information 31 stored in the auxiliary storage device 15 are searched under the search condition “parts or phenomena”. f (part near N phenomenon) is applicable when all the defect information 31 stored in the auxiliary storage device 15 is searched under the condition that the part and the phenomenon appear within a distance of "N" characters or less. This is the number of defect information 31. In the “parts” section, “XX product” and the like are entered, and in the “phenomenon” section, “startup impossible” and the like are entered. Naturally, “and” means that a part and a phenomenon appear at the same time, and “or” means that at least one of the part and the phenomenon appears. The user can set “N” to any value.

  The larger the value of “f (parts and phenomena) / f (parts or phenomena)” on the right side of # 6, the larger the number (ratio) of the parts and phenomena appearing simultaneously in one defect information 31 become. As the value of “f (part near N phenomenon) / f (part and phenomenon)” on the right side of # 6 is larger, the number (ratio) of defect information 31 in which the distance between the part and the phenomenon is shorter is larger. . The co-occurrence relation evaluation unit 22 calculates the value of the left side of # 6 for each component for each phenomenon, and extracts the phenomenon in which the value of the left side of # 6 is maximized.

  The component / phenomenon extraction unit 21, the co-occurrence relationship evaluation unit 22, the causal relationship extraction unit 23, the failure tree creation unit 24, and the FMEA table creation unit 25 display the screen displayed on the output device 13 on the output device of the terminal device 3. It is possible to In addition, these units 21 to 25 can also receive information received via the input device 12 via the input device of the terminal device 3. Therefore, if a plurality of terminal devices 3 are prepared, many users can access the causal relationship extracting device 2 at the same time.

(Effects of the embodiment)
The causal relationship extraction device 2 of the present embodiment has the following effects.
(1) The causal relationship extraction device can make a user easily refer to an example of a chain of causal relationship elements.
(2) The causal relationship extracting device allows a user to easily create a fault tree and an FMEA table by referring to an example of a chain of causal components.
(3) The causal relationship extraction device can accurately extract components and phenomena based on the tendency to express words.
(4) The causal relationship extraction device can quickly extract components and phenomena in a specific field by using the component list and the phenomenon list.
(5) By calculating the value of the co-occurrence relation evaluation index, the causal relation extraction device can accurately evaluate the co-occurrence relation between components and phenomena.
(6) The causal relationship extraction device can quickly evaluate the co-occurrence relationship between the component and the phenomenon by calculating the number of characters between the word indicating the component and the word indicating the phenomenon.
(7) The causal relationship extraction device can objectively evaluate the co-occurrence relationship between components and phenomena by calculating the frequency of occurrence of words indicating components and phenomena in a plurality of pieces of defect information.

  Note that the present invention is not limited to the above-described embodiment, and includes various modifications. For example, the above-described embodiments have been described in detail for easy understanding of the present invention, and are not necessarily limited to those having all the configurations described above. Further, a part of the configuration of one embodiment can be replaced with the configuration of another embodiment, and the configuration of one embodiment can be added to the configuration of another embodiment. Also, for a part of the configuration of each embodiment, it is possible to add, delete, or replace another configuration.

In addition, each of the above-described configurations, functions, processing units, processing means, and the like may be partially or entirely realized by hardware, for example, by designing an integrated circuit. Further, each of the above-described configurations, functions, and the like may be realized by software by a processor interpreting and executing a program that realizes each function. Information such as a program, a table, and a file for realizing each function can be stored in a memory, a hard disk, a recording device such as an SSD (Solid State Drive), or a recording medium such as an IC card, an SD card, or a DVD.
Further, the control lines and the information lines are shown to be necessary for the explanation, and not all the control lines and the information lines are necessarily shown on the product. In practice, almost all components may be considered to be interconnected.

2 Causal relationship extraction device 3 Terminal device 11 Central control device 12 Input device 13 Output device 14 Main storage device (storage unit)
15 Auxiliary storage device (storage unit)
21 Component / Phenomenon Extraction Unit 22 Co-occurrence Relationship Evaluation Unit 23 Causal Relationship Extraction Unit 24 Failure Tree Creation Unit 25 FMEA Table Creation Unit 31 Failure Information 32 Parts List 33 Phenomenon List 34 Parts Tree 35 Causal Relationship Information 36 Failure Tree 37 FMEA Table

Claims (9)

A storage unit for storing defect information and a hierarchical structure of parts, which are examples of defects;
A part / phenomenon extraction unit that extracts a word indicating a part and a word indicating a phenomenon from the defect information;
A word indicating the part is combined with a word indicating the phenomenon, and for each combination of the word indicating the part and the word indicating the phenomenon, the degree of association between the word indicating the part and the word indicating the phenomenon is determined. Calculated, a co-occurrence relation evaluation unit that determines a plurality of combinations having high relevance based on the degree of relevance as elements of a causal relation ,
By specifying a hierarchy in the hierarchical structure of the parts for the parts included in the causal element, and ranking the causal elements in the order of the higher rank, a causal relation composed only of the causal elements Creating a chain of elements, a causal relationship extracting unit that stores the chain of causal elements in the storage unit,
A causal relationship extraction device comprising: A failure tree creating unit that displays a chain of the stored causal elements on a screen when the user performs a failure tree analysis;
An FMEA table creation unit that displays a chain of causal elements stored in the storage unit on the screen when the user performs the failure mode effect analysis;
The causality extracting apparatus according to claim 1, further comprising: The part / phenomenon extraction unit includes:
Extracting a word indicating a part and a word indicating a phenomenon from the defect information based on the expression tendency of the word included in the defect information;
The causal relationship extraction device according to claim 1, wherein The storage unit,
A list of words indicating parts and a list of words indicating phenomena are stored.
The part / phenomenon extraction unit includes:
Based on the list of words indicating the parts, extract words indicating parts from the defect information,
Extracting a word indicating a phenomenon from the defect information based on a list of words indicating the phenomenon,
The causal relationship extraction device according to claim 1, wherein The co-occurrence relation evaluation unit,
Assessing co-occurrence by calculating the value of the co-occurrence metric used in natural language processing;
The causal relationship extraction device according to claim 1, wherein The co-occurrence relation evaluation unit,
Evaluating the co-occurrence relationship by calculating the number of characters between the word indicating the extracted part and the word indicating the extracted phenomenon,
The causal relationship extraction device according to claim 1, wherein The storage unit,
Storing a plurality of the defect information,
The co-occurrence relation evaluation unit,
Evaluating the co-occurrence relationship by calculating a frequency of occurrence of a word indicating the extracted part and a word indicating the extracted phenomenon in the plurality of pieces of the defect information;
The causal relationship extraction device according to claim 5, characterized in that: The storage unit of the causal relationship extraction device includes:
It stores the defect information and the hierarchical structure of parts, which are examples of defects.
The component / phenomenon extraction unit of the causality extraction device,
Extracting a word indicating a part and a word indicating a phenomenon from the defect information,
The co-occurrence relation evaluation unit of the causal relation extraction device,
A word indicating the part is combined with a word indicating the phenomenon, and for each combination of the word indicating the part and the word indicating the phenomenon, the degree of association between the word indicating the part and the word indicating the phenomenon is determined. Calculate and determine a plurality of combinations with high relevance as causal elements based on the degree of relevance ,
The causal relationship extraction unit of the causal relationship extraction device,
By specifying a hierarchy in the hierarchical structure of the parts for the components included in the causal element, and ranking the causal elements in the order of the hierarchy, the causal relation composed of only the causal elements Create a chain of elements of, and store the chain of causal elements in the storage unit,
A causal relationship extracting method of a causal relationship extracting apparatus, characterized in that: For the storage unit of the causal relationship extraction device,
The defect information and the hierarchical structure of parts, which are examples of defects, are stored.
For the part / phenomenon extraction unit of the causality extraction device,
Performing a process of extracting a word indicating a part and a word indicating a phenomenon from the defect information,
For the co-occurrence relationship evaluation unit of the causal relationship extraction device,
A word indicating the part is combined with a word indicating the phenomenon, and for each combination of the word indicating the part and the word indicating the phenomenon, the degree of association between the word indicating the part and the word indicating the phenomenon is determined. Calculating, and performing a process of determining a plurality of combinations having high relevance based on the degree of relevance as elements of a causal relationship ,
For the causal relationship extraction unit of the causal relationship extraction device,
By specifying a hierarchy in the hierarchical structure of the parts for the components included in the causal element, and ranking the causal elements in the order of the hierarchy, the causal relation composed of only the causal elements Creating a chain of elements, and executing a process of storing the chain of elements of the causal relationship in the storage unit;
A causal relationship extraction program for causing a causal relationship extracting device to function.

Images