A knowledge graph construction method based on failure mode and effect analysis information

Abstract

The present invention discloses a knowledge graph construction method based on failure mode and effect analysis information, including: constructing a functional block diagram of the agreed hierarchy of equipment structure and FMEA tables under each agreed hierarchy; according to the functional block diagram of the agreed hierarchy of equipment structure, hierarchically determining node type labels and supplementing each node attribute, and importing each component node into the knowledge graph; hierarchically connecting the component nodes in the knowledge graph according to the node type labels, and determining the backbone structure of the knowledge graph according to the subordinate relationship between each component node; taking each relevant fault information of the equipment as a fault information node and importing it into the knowledge graph; establishing the relationship between each component node and the fault information node according to the relevant fault information of each component under each agreed hierarchy in the FMEA table; deleting redundant nodes and relationships to obtain the final knowledge graph. The present invention can realize the intuitive display of the fault association between each component in the equipment, which is convenient for quickly and effectively searching for FMEA information.

Description

Technical Field

[0001] The present invention relates to the field of information visualization technology, and more specifically to a method for constructing a knowledge graph based on failure mode and impact analysis information. Background Art

[0002] During the equipment development phase, reliability, maintainability and supportability are the three most basic quality characteristics of the equipment, which require key design and analysis to ensure that the equipment maintains good performance during subsequent use. Among them, supportability analysis and design determine whether the design characteristics of the equipment and the planned support resources can meet the requirements of peacetime combat readiness and wartime use, affect the combat readiness integrity and mission continuity of the equipment, and restrict the combat effectiveness of the equipment.

[0003] Failure Mode and Effects Analysis (FMEA) is a commonly used support analysis technology. It is necessary to analyze various possible failure modes of the product during the development stage, explore the causes of the failure modes, and evaluate the impact of the failure modes on the use, function or status of the equipment. In the FMEA analysis process, great emphasis is placed on accumulating historical experience, and a corresponding failure mode and related information database is usually established. However, the relevant information database is mostly represented in the form of attribute tables, tree structure diagrams, etc., which cannot intuitively display all the failure-related information of each component and the hierarchical associations between components.

[0004] Therefore, how to provide a knowledge graph construction method that can intuitively display all fault-related information of each component and the hierarchical associations between components is a problem that technical personnel in this field urgently need to solve. Summary of the invention

[0005] In view of this, the present invention provides a knowledge graph construction method based on failure mode and effect analysis information, which can realize the intuitive display of fault correlation between components in equipment, and facilitate the rapid and effective search for FMEA information.

[0006] In order to achieve the above object, the present invention adopts the following technical solution:

[0007] A method for constructing a knowledge graph based on failure mode and effect analysis information, comprising:

[0008] According to the component structure composition relationship of the equipment and related fault information, the functional block diagram of the equipment structure agreed level and the FMEA table at each agreed level are obtained;

[0009] According to the equipment structure agreed hierarchical functional block diagram, node type labels are determined hierarchically and the attributes of each node are supplemented, and each component node is imported into the knowledge graph;

[0010] According to the node type labels, the component nodes in the knowledge graph are hierarchically connected, and according to the subordinate relationship between the component nodes, the backbone structure of the knowledge graph is determined;

[0011] Each type of relevant fault information of the equipment is used as a fault information node and imported into the knowledge graph;

[0012] According to the relevant fault information of each component at each agreed level in the FMEA table, the relationship between each component node and the fault information node is established;

[0013] Delete redundant nodes and relationships to obtain the final knowledge graph.

[0014] It can be seen from the above technical solutions that, compared with the prior art, the present invention discloses a method for constructing a knowledge graph based on failure mode and impact analysis information, which not only optimizes the convenience of querying component failure information in equipment, but also supplements the correlation between component failures. In addition, considering that new failure modes may occur during the use of equipment, the knowledge graph can dynamically adjust the addition of nodes and relationships, which is conducive to the dynamic update of equipment failure information and can effectively record historical failure experiences. At the same time, the present invention further improves the construction of the information database of equipment failure mode and impact analysis, provides an intuitive reference for the subsequent improvement design of the equipment according to the failure information, and also provides a new direction for the intelligent construction of equipment.

[0015] Optionally, the step of creating a functional block diagram and FMEA table of the equipment structure agreement hierarchy according to the component structure composition relationship of the equipment and related fault information includes:

[0016] According to the structural composition relationship of the equipment, the equipment is divided into agreed levels, and the functional analysis of the structural composition of each level is carried out to obtain the functional block diagram of the equipment structure agreed level;

[0017] Starting from the lowest agreed level of the equipment, analyze the relevant fault information of the components from bottom to top to obtain the FMEA table at each agreed level of the equipment;

[0018] The equipment structure agreed hierarchical functional block diagram and the FMEA table are integrated as input information of the knowledge graph.

[0019] Optionally, the relevant fault information includes: fault mode, cause of each fault mode and fault impact.

[0020] Optionally, the equipment structure agreed hierarchy functional block diagram includes the system composition, agreed hierarchy and component functions of the equipment.

[0021] Optionally, the step of agreeing on a hierarchical functional block diagram according to the equipment structure, hierarchically determining node type labels and supplementing attributes of each node, and importing each component node into a knowledge graph includes:

[0022] According to the equipment structure agreed hierarchy functional block diagram, determine each component node in the initial agreed hierarchy, the intermediate agreed hierarchy and the lowest agreed hierarchy of the equipment;

[0023] Using the relevant data of each component in each agreed level as the node attribute of the component, and adding the node attribute to the corresponding component node;

[0024] Each component node containing the agreed hierarchical level, component name and node attribute information is used as a node type label and imported into the knowledge graph.

[0025] Optionally, the node attributes include: failure rate, mean time between failures and mean time between repairs.

[0026] Optionally, establishing the relationship between each component node and the fault information node according to the relevant fault information of each component at each agreed level in the FMEA table includes:

[0027] According to the FMEA table under each agreed level, starting from the initial agreed level, the subordinate relationship between each component node in each agreed level and the corresponding fault information node is added layer by layer. BRIEF DESCRIPTION OF THE DRAWINGS

[0028] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on the provided drawings without paying creative work.

[0029] Figure 1 The accompanying drawing is a flow chart of a method for constructing a knowledge graph based on failure mode and impact analysis information provided by the present invention;

[0030] Figure 2 The accompanying drawing is a functional block diagram of the equipment structure provided by the present invention;

[0031] Figure 3 The accompanying drawing is a schematic diagram of the knowledge graph structure after each component node is imported provided by the present invention;

[0032] Figure 4 The accompanying drawing is a schematic diagram of the main structure of the knowledge graph provided by the present invention;

[0033] Figure 5 The accompanying drawing is a schematic diagram of the knowledge graph structure after the fault information node is imported provided by the present invention;

[0034] Figure 6The accompanying drawing is a schematic diagram of the knowledge graph structure after the relationship between each component node and the fault information node is imported provided by the present invention;

[0035] Figure 7 The accompanying drawing is a schematic diagram of the knowledge graph structure before deleting redundant relations provided by the present invention;

[0036] Figure 8 The accompanying drawing is a schematic diagram of the knowledge graph structure after deleting redundant relations provided by the present invention;

[0037] Fig. 9 The accompanying drawing is a schematic diagram of the final knowledge graph structure provided by the present invention. DETAILED DESCRIPTION

[0038] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0039] like Figure 1 As shown, an embodiment of the present invention discloses a method for constructing a knowledge graph based on failure mode and impact analysis information, comprising the following steps:

[0040] S1. According to the component structure composition relationship of the equipment and related fault information, obtain the functional block diagram of the equipment structure agreed level and the FMEA table at each agreed level;

[0041] S2. According to the agreed hierarchical functional block diagram of the equipment structure, determine the node type label hierarchically and supplement the attributes of each node, and import each component node into the knowledge graph;

[0042] S3. According to the node type labels, the component nodes in the knowledge graph are hierarchically connected, and the backbone structure of the knowledge graph is determined according to the subordinate relationship between the component nodes;

[0043] S4. Take each relevant fault information of the equipment as a fault information node and import it into the knowledge graph;

[0044] S5. According to the relevant fault information of each component at each agreed level in the FMEA table, establish the relationship between each component node and the fault information node;

[0045] S6. Delete redundant nodes and relationships to obtain the final knowledge graph.

[0046] Below, each of the above steps is described in detail.

[0047] S1. According to the component structure composition relationship of the equipment and related fault information, obtain the functional block diagram of the agreed level of the equipment structure and the FMEA table at each agreed level. Specifically including:

[0048] S11. According to the structural composition relationship of the equipment, the equipment is divided into agreed levels, and the functional analysis of the structural composition of each level is performed to obtain the functional block diagram of the equipment structure agreed level. Among them, the structural composition relationship of the equipment can be obtained based on the equipment working principle diagram and design drawings and other materials.

[0049] Taking a certain type of aircraft as an example, the equipment agreement level is divided from the whole aircraft to the bottom. According to the actual maintenance situation, the lowest agreement level is determined to be the SRU level, the initial agreement level is the equipment level, and the intermediate agreement level can be the system level (second level), subsystem level (third level), LRU level (fourth level), etc. The corresponding functional analysis is carried out on the structural composition of each level, and the structural agreement level functional block diagram of the aircraft is drawn, such as Figure 2 As shown.

[0050] S12. Analyze the relevant fault information of the components from the lowest agreed level of the equipment from bottom to top, and obtain the FMEA table at each agreed level of the equipment. Among them, the relevant fault information includes the fault mode, fault cause and fault impact. The fault mode is determined specifically from the function, task and use environment of the component, and the cause of the fault mode is analyzed comprehensively by internal factors and external factors to evaluate its fault impact.

[0051] Taking the SRU parts at the lowest agreed level as the protection object, its fault information can be analyzed as shown in Table 1:

[0052] Table 1 FMEA table of lowest agreed level parts

[0053]

[0054] Continuing to analyze the previous agreed level, combined with the lowest agreed level fault information results, we further obtain Table 2:

[0055] Table 2 FMEA table of intermediate agreed level parts

[0056]

[0057] S13. Integrate the equipment structure agreed hierarchy function block diagram and FMEA table as input information for the knowledge graph.

[0058] S2. According to the agreed hierarchical functional block diagram of the equipment structure, determine the node type label hierarchically and supplement the attributes of each node, and import each component node into the knowledge graph. Specifically include:

[0059] S21. According to the equipment structure agreement level functional block diagram, determine the initial agreement level, intermediate agreement level and lowest agreement level of the equipment. Set the initial agreement level as the equipment node, the lowest agreement level as the shop replaceable unit (SRU) node, and the intermediate agreement level as the system node, subsystem node, line replaceable unit (LRU) node, etc. according to the equipment situation.

[0060] S22, taking the relevant data of each component in each agreed level as the node attribute of the component, and adding the node attribute to the corresponding component node. The relevant data may be failure rate, mean time between failures, mean time between repairs, etc.

[0061] Taking the above-mentioned aircraft as an example, its label type is defined as the agreed level where the component or system is located, and each component structure of each agreed level is numbered to sort out the component node information table, as shown in Table 3:

[0062] Table 3 Component node information table

[0063] id labels name Failure rate 0 Equipment level equipment 0.00898 1 System Level System 1 0.057 2 System Level System 2 0.034 3 System Level System 0.011 4 LRU level LRU1 1.48 5 LRU level LRU2 0.24 6 LRU level LRU3 0.59 7 LRU level LRU4 1.17 8 SRU Level SRU1 40 9 SRU Level SRU2 2.5 10 SRU Level SRU3 1.3 11 SRU Level SRU4 0.7 12 SRU Level SRU5 0.5 13 SRU Level SRU6 10.51 14 SRU Level SRU7 0.04 15 SRU Level SRU8 1.78

[0064] S23. Take each component node containing the agreed hierarchy level, component name and node attribute information as a node type label and import it into the knowledge graph. Convert Table 3 into a csv file and import the component node information into neo4j to obtain the following: Figure 3 The knowledge graph shown.

[0065] S3. According to the node type labels, the component nodes in the knowledge graph are hierarchically connected, and the backbone structure of the knowledge graph is determined according to the subordinate relationship between the component nodes. Specifically, it includes:

[0066] S31: According to the structural hierarchical relationship in the equipment structure convention hierarchical function block diagram, connect from the equipment node from top to bottom, and add the inclusion relationship between the equipment node and the system node;

[0067] S32: By analogy, the inclusion relationship between each agreed level is added layer by layer until the nodes of the lowest agreed level are connected, and the main part of the equipment knowledge graph is completed.

[0068] Taking the above aircraft as an example, the node id represents a structure of a specific level. The entity triple formed by this information indicates that the output node contains the input node, such as the equipment node contains the system 1 node. The structure composition node relationship connection table is sorted out, as shown in Table 4:

[0069] Table 4 Structure composition node relationship connection information table

[0070] Outgoing node id Ingress id relation 0 1 contain 0 2 contain 0 3 contain 1 4 contain 1 5 contain 2 6 contain 2 7 contain 4 8 contain 4 9 contain 5 10 contain 5 11 contain 6 12 contain 6 13 contain 7 14 contain 7 15 contain

[0071] Convert the above table into a csv file and import the node relationship connection information into neo4j to get the following: Figure 4 The backbone structure of the knowledge graph shown.

[0072] S4. Take each relevant fault information of the equipment as a fault information node and import it into the knowledge graph to improve the knowledge graph. Specifically include:

[0073] S41: According to the FMEA table, summarize the existing failure modes, failure causes and failure effects, and treat each mode, cause and effect as a failure information node.

[0074] Taking the above aircraft as an example, there are three types of fault information nodes, namely, fault mode node, fault cause node, and fault impact node. Each form under each type of node represents a node and continues to be numbered independently. The summary is shown in Table 5:

[0075] Table 5 Fault information node table

[0076] id labels name 16 Failure Mode Failure Mode 1 17 Failure Mode Failure Mode 2 18 Failure Mode Failure Mode 3 19 Failure Mode Failure Mode 4 20 Failure Mode Failure Mode 5 21 Failure Mode Failure Mode 6 22 Failure Mode Failure Mode 7 23 Cause Cause 1 24 Cause Cause 2 25 Cause Cause 3 26 Cause Cause 4 27 Cause Cause 5 28 Failure Impact Fault Impact 1 29 Failure Impact Fault Impact 2 30 Failure Impact Fault Impact 3

[0077] S42: Import each fault information node into the equipment knowledge graph established in S3 to improve the knowledge related to equipment faults.

[0078] Convert the summarized fault table into a csv file and import the fault information node into neo4j to get the following: Figure 5 The knowledge graph shown.

[0079] S5. According to the relevant fault information of each component at each agreed level in the FMEA table, establish the relationship between each component node and the fault information node. Specifically include:

[0080] S51: According to the FMEA table of each agreed level, starting from the initial agreed level node, add its subordinate relationship with the fault information node;

[0081] S52: By analogy, add the subordinate relationship between each agreed level and its existing fault mode, fault cause, and fault impact node until the fault information of the lowest agreed level node is connected, and the fault information part of the equipment knowledge graph is completed.

[0082] Taking the above aircraft as an example, according to the agreed hierarchical fault information given in the FMEA analysis table of S1, the following three types of entity triples can be constructed: first, the structure component node has a certain fault mode node; second, the fault mode node is caused by the fault cause node; third, the fault of the structure component node will cause a certain fault impact. The following relationship table between the structure component node and the fault information node can be sorted out, as shown in Table 6, Table 7, and Table 8:

[0083] Table 6 Relationship between structure component nodes and failure mode nodes

[0084] Outgoing Node Ingress relation 4 22 has_failure 4 20 has_failure 5 17 has_failure 7 18 has_failure 8 16 has_failure 9 17 has_failure 9 18 has_failure 10 19 has_failure 11 18 has_failure 12 20 has_failure 12 21 has_failure 13 19 has_failure 14 21 has_failure 15 22 has_failure

[0085] Table 7 Fault mode and fault cause node relationship table

[0086]

[0087]

[0088] Table 8 Relationship between fault modes and fault-affected nodes

[0089] Outgoing Node Ingress relation 8 28 has_effect 9 29 has_effect 9 30 has_effect 4 29 has_effect 4 28 has_effect 10 29 has_effect 11 28 has_effect 5 30 has_effect 12 28 has_effect 12 28 has_effect 13 29 has_effect 14 28 has_effect 15 28 has_effect 7 29 has_effect

[0090] Convert the above table into a csv file and import the fault relationship connection information into neo4j to get the following: Figure 6 The knowledge graph shown.

[0091] S6. Delete redundant nodes and relationships to obtain the final knowledge graph. Specifically include:

[0092] S61: Check whether there are repeated component nodes or relationships, such as whether the same components under different subsystems are added repeatedly. If so, delete and integrate them;

[0093] S62: Check whether there are repeated fault information nodes or relationships. For example, if different parts have the same fault mode, check whether the fault mode node is redundant. If so, delete and integrate them. After sorting, a complete equipment knowledge graph is obtained.

[0094] Taking the above aircraft as an example, it is found that LRU1 and SRU1 are in a subordinate relationship and both point to fault impact 1. It can be determined that the triplet of LRU1 and fault impact 1 is a redundant relationship, and the relationship is deleted. The before and after comparison is shown in the figure below. Figure 7-8 shown.

[0095] After deleting redundant nodes and relationships, we get Figure 7 The final knowledge graph shown.

[0096] In this specification, each embodiment is described in a progressive manner, and each embodiment focuses on the differences from other embodiments. The same or similar parts between the embodiments can be referred to each other. For the device disclosed in the embodiment, since it corresponds to the method disclosed in the embodiment, the description is relatively simple, and the relevant parts can be referred to the method part.

[0097] The above description of the disclosed embodiments enables one skilled in the art to implement or use the present invention. Various modifications to these embodiments will be apparent to one skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention will not be limited to the embodiments shown herein, but rather to the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A method for constructing a knowledge graph based on failure mode and effect analysis information, characterized in that: include: According to the component structure composition relationship of the equipment and related fault information, the functional block diagram of the equipment structure agreed level and the FMEA table at each agreed level are obtained; According to the equipment structure agreed hierarchical functional block diagram, node type labels are determined hierarchically and the attributes of each node are supplemented, and each component node is imported into the knowledge graph; According to the node type labels, the component nodes in the knowledge graph are hierarchically connected, and according to the subordinate relationship between the component nodes, the backbone structure of the knowledge graph is determined; Each type of relevant fault information of the equipment is used as a fault information node and imported into the knowledge graph; According to the relevant fault information of each component at each agreed level in the FMEA table, the relationship between each component node and the fault information node is established; Delete redundant nodes and relationships to obtain the final knowledge graph.

2. A method for constructing a knowledge graph based on failure mode and effect analysis information according to claim 1, characterized in that: According to the component structure composition relationship of the equipment and related fault information, the equipment structure agreed hierarchical functional block diagram and FMEA table are created, including: According to the structural composition relationship of the equipment, the equipment is divided into agreed levels, and the functional analysis of the structural composition of each level is carried out to obtain the functional block diagram of the equipment structure agreed level; Starting from the lowest agreed level of the equipment, analyze the relevant fault information of the components from bottom to top to obtain the FMEA table at each agreed level of the equipment; The equipment structure agreed hierarchical functional block diagram and the FMEA table are integrated as input information of the knowledge graph.

3. A method for constructing a knowledge graph based on failure mode and effect analysis information according to claim 1 or 2, characterized in that: The relevant fault information includes: fault modes, causes of each fault mode and fault impact.

4. A method for constructing a knowledge graph based on failure mode and effect analysis information according to claim 1 or 2, characterized in that: The equipment structure agreed hierarchy functional block diagram includes the system composition, agreed hierarchy and component functions of the equipment.

5. The method for constructing a knowledge graph based on failure mode and effect analysis information according to claim 1, characterized in that: The hierarchical functional block diagram is agreed upon according to the equipment structure, node type labels are determined hierarchically and attributes of each node are supplemented, and each component node is imported into the knowledge graph, including: According to the equipment structure agreed hierarchy functional block diagram, determine each component node in the initial agreed hierarchy, the intermediate agreed hierarchy and the lowest agreed hierarchy of the equipment; Using the relevant data of each component in each agreed level as the node attribute of the component, and adding the node attribute to the corresponding component node; Each component node containing the agreed hierarchical level, component name and node attribute information is used as a node type label and imported into the knowledge graph.

6. A method for constructing a knowledge graph based on failure mode and effect analysis information according to claim 5, characterized in that: The node attributes include: failure rate, mean time between failures and mean time between repairs.

7. The method for constructing a knowledge graph based on failure mode and effect analysis information according to claim 5, characterized in that: The establishing of the relationship between each component node and the fault information node according to the relevant fault information of each component at each agreed level in the FMEA table includes: According to the FMEA table under each agreed level, starting from the initial agreed level, the subordinate relationship between each component node in each agreed level and the corresponding fault information node is added layer by layer.

Images