A failure analysis standard process generation method and system

Abstract

The present application relates to a kind of failure analysis standard process generation method and system, belong to failure analysis technical field, solve the problem that existing failure analysis standard process cannot be automatically generated.There is: construct failure analysis knowledge graph, based on failure analysis knowledge graph, the physical structure class and fault phenomenon class entity of new part and its attribute and relationship are constructed, according to entity similarity, the entity alignment relationship of new part and old part is established;In turn, the detection item entity of the fault phenomenon class of new part is taken out, according to entity alignment relationship, the detection item entity of the old part corresponding to current detection item entity is obtained, according to the physical structure class and detection process class entity associated with the detection item entity of old part, the detection process class entity associated with current detection item entity of new part and its attribute and relationship are constructed;According to the detection process class entity of new part and its attribute and relationship, failure analysis standard process is generated.Failure analysis standard process automation and standardization are realized.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of failure analysis, in particular to a failure analysis standard process generation method and system. BACKGROUND

[0002] Failure analysis generally simulates and reproduces the phenomenon of failure by analyzing and verifying the failure mode and phenomenon, finds out the cause of failure, and excavates the mechanism of failure. Failure analysis has strong practical significance in improving product quality, technical development, improvement, product repair and arbitrating failure accidents. With the increase of equipment, the storage capacity of industrial failure analysis data will increase exponentially.

[0003] Industrial failure analysis data is of various types, complex structure and strong correlation. There are structured data such as product-related requirements, bill of materials (BOM), three-dimensional models, semi-structured data such as design drawings, technical documents, various documents, and time series data of production lines, equipment and intelligent products. These data have strong correlation, but are often scattered in different business links and information systems, and traditional organizational barriers and information silos break the internal correlation between these data.

[0004] In industrial failure analysis, a standard detection process is generally used to analyze the failure analysis process of industrial assembly, but the standard detection process used is generally manually written, which is time-consuming and labor-intensive, and the level of engineers varies, and each person writes differently, so the quality of writing is not uniform, and the readability and maintainability are not high. Moreover, the loss of industry knowledge and experience due to personnel turnover makes it difficult to quickly obtain a standard detection process for failure analysis when a new mechanical part fails. SUMMARY

[0005] In view of the above analysis, the embodiments of the present application aim to provide a failure analysis standard process generation method and system to solve the problem that the existing failure analysis standard process cannot be automatically generated.

[0006] In one aspect, the embodiments of the present application provide a failure analysis standard process generation method, comprising the following steps:

[0007] Constructing a failure analysis knowledge graph, constructing physical structure class and fault phenomenon class entities and their attributes and relationships of a new part based on the failure analysis knowledge graph, and establishing an entity alignment relationship between the new part and the old part according to the entity similarity;

[0008] The detection item entity of the new part is sequentially taken out, the detection item entity corresponding to the old part of the current detection item entity is obtained according to the entity alignment relationship, and the detection process class entity associated with the detection item entity of the old part, and attributes and relationships of the detection process class entity are constructed according to the physical structure class and the detection process class entity associated with the detection item entity of the old part;

[0009] According to the detection process class entity of the new part and the attributes and relationships thereof, a failure analysis standard process is generated.

[0010] Based on the further improvement of the above method, the failure analysis knowledge graph is constructed, including:

[0011] According to the bill of materials and the netlist, the physical structure class entity and the attributes and relationships thereof are established, and the physical structure class entity includes a part entity and a line entity;

[0012] According to the test file and the test log, the fault phenomenon class entity and the attributes and relationships thereof are established, and the fault phenomenon class entity includes a detection item entity and a test command entity;

[0013] According to the historical failure analysis standard process, the detection process class entity and the attributes and relationships thereof are established, and the detection process class entity includes a detection step entity and a detection result entity;

[0014] According to the relationship between the historical failure analysis standard process and the test file, the relationship between the detection step entity and the detection item entity is established;

[0015] Based on the physical structure class entity, the operation step description attribute value in the detection step entity is subjected to word segmentation processing to obtain a step word segmentation result, and the relationship between the physical structure class entity and the detection step entity is established according to the step word segmentation result.

[0016] Based on the further improvement of the above method, the relationship between the physical structure class entity and the detection step entity is established according to the step word segmentation result, including:

[0017] The step word segmentation result is compared with the entity name of the physical structure class entity, if they are the same, the entity corresponding to the entity name is taken as a tail entity, the detection step entity corresponding to the step word segmentation result is taken as a head entity, the operation type attribute value in the head entity is taken as an operation relationship, and a triple of <head entity-operation relationship-tail entity> is established.

[0018] Based on the further improvement of the above method, the failure analysis knowledge graph is constructed, further including:

[0019] Based on the failure analysis document writing specification, the part-of-speech sequence templates of each operation type are extracted, the part-of-speech of the step segmentation result is labeled, and the labeled result is compared with the part-of-speech sequence templates. The step segmentation results consistent with the comparison are compared, the relationship between the physical structure class entity and the detection step entity or the relationship between the detection step entity and the test command entity is established according to the part-of-speech sequence template.

[0020] Based on the further improvement of the above method, the entity alignment relationship between the new part and the old part is established according to the entity similarity, including:

[0021] According to the entity category, each new part entity that has not established an alignment relationship is compared with each old part entity in turn, the similarity of the entity name, the first text spliced by the entity name and its attributes and attribute values, and the second text spliced by the entity name and its adjacent entity name, and when the similarity is the largest and greater than the threshold, the corresponding new part entity and old part entity are established as a triple of <new part entity-alignment-old part entity>.

[0022] Based on the further improvement of the above method, the word2vec model is used to obtain the word vector of the entity name; the Oag-Bert model is used to obtain the text vector of the first text and the second text.

[0023] Based on the further improvement of the above method, the detection process class entity associated with the detection item entity of the old part is used to construct the detection process class entity associated with the current detection item entity of the new part and its attributes and relationships, including:

[0024] The detection process class entity associated with the detection item entity of the old part is obtained, and each detection step entity in it is taken out in turn. If all physical structure class entities or test command entities associated with the current detection step entity exist in the alignment relationship with the current new part, the current detection step entity and its attributes are copied as the detection step entity and its attributes of the new part, the current detection item entity of the new part is associated, the relationship between the detection step entity of the new part and the physical structure class entity or test command entity of the new part is established according to the alignment relationship, and the detection process class entity associated with the current detection item entity of the new part and its attributes and relationships are obtained.

[0025] Based on the further improvement of the above method, the failure analysis standard process is generated according to the detection process class entity and its attributes and relationships of the new part, including:

[0026] The detection step entities in the detection process class entity of the new part are taken out in turn as the basic steps.

[0027] According to the operation kind attribute value of each detection step entity, a corresponding operation step description template is obtained, a physical structure class entity name or a test command attribute of the test command entity associated with the detection step entity is filled into the operation step description template, an operation step description is obtained, and the operation step description is added to the basic step;

[0028] According to the other detection step entities associated with each detection step entity, the step number attribute values of the other detection step entities are taken as next operation, and the next operation is added to the basic step;

[0029] According to the detection result entity associated with the detection step entity, a physical structure class entity name associated with the detection result entity is taken as a detection fault point, and the detection fault point is added to the basic step;

[0030] The basic step is output, and a failure analysis standard process is obtained.

[0031] In another aspect, an embodiment of the present application provides a failure analysis standard process generation system, comprising:

[0032] A knowledge entity fusion module is configured to construct physical structure class and fault phenomenon class entities of a new part and attributes and relationships thereof based on a failure analysis knowledge graph, and establish an entity alignment relationship between the new part and an old part according to entity similarity;

[0033] A knowledge entity reasoning module is configured to sequentially take out a detection item entity of a fault phenomenon class of the new part, acquire a detection item entity of the old part corresponding to the current detection item entity according to the entity alignment relationship, and construct a detection flow class entity associated with the current detection item entity of the new part and attributes and relationships thereof according to the physical structure class and detection flow class entities associated with the detection item entity of the old part.

[0034] A detection flow generation module is configured to generate a failure analysis standard process according to the detection flow class entity of the new part and attributes and relationships thereof.

[0035] Further improvements based on the above system further comprise:

[0036] A knowledge graph construction module is configured to construct a failure analysis knowledge graph according to historical data, comprising:

[0037] Physical structure class entities and attributes and relationships thereof are established according to a bill of materials and a netlist, and the physical structure class entities include part entities and circuit entities;

[0038] Fault phenomenon class entities and attributes and relationships thereof are established according to test files and test logs, and the fault phenomenon class entities include detection item entities and test command entities;

[0039] According to the historical failure analysis standard process, detection process class entities and their attributes and relationships are established, and the detection process class entities include detection step entities and detection result entities;

[0040] According to the relationship between the historical failure analysis standard process and the test file, the relationship between the detection step entity and the detection project entity is established.

[0041] Based on the physical structure class entity, the operation step description attribute in the detection step entity is segmented, the step segmentation result is obtained, and the relationship between the physical structure class entity and the detection step entity is established according to the step segmentation result.

[0042] Compared with the prior art, the present application can at least realize one of the following beneficial effects: making full use of historical data, mining the relationship of failure analysis, constructing a failure analysis knowledge graph, assisting in realizing the related knowledge query of the failure analysis standard process, and assisting in realizing the modification of the failure analysis standard process by artificial; generating a unified failure analysis standard process through the knowledge graph, solving the errors and differences caused by manual writing, and facilitating the optimization of the failure analysis standard process; through the knowledge entity fusion, knowledge entity reasoning and knowledge entity association query of new and old parts, taking the detection step as the dimension, automatically performing entity reuse and relationship migration, generating the failure analysis standard process of the new part, improving the work efficiency and accuracy, and realizing the automation and standardization of the failure analysis standard process.

[0043] In the present application, the above-mentioned technical solutions can be combined with each other to realize more preferred combination solutions. Other features and advantages of the present application will be described in the subsequent specification, and some advantages will become apparent from the specification, or will be understood by implementing the present application. The purpose and other advantages of the present application can be realized and obtained from the contents specifically pointed out in the specification and the drawings. BRIEF DESCRIPTION OF DRAWINGS

[0044] The accompanying drawings are included to provide a further understanding of the application and are incorporated herein and constitute a part of the application. The same reference numerals in the several figures indicate same or similar components.

[0045] Figure 1 A failure analysis standard process generation method flowchart in embodiment 1 of the present application is shown. DETAILED DESCRIPTION

[0046] The preferred embodiments of the present application will be specifically described below with reference to the accompanying drawings, wherein the drawings constitute a part of the present application and are used to illustrate the principles of the embodiments of the present application, and are not used to limit the scope of the present application.

[0047] Embodiment 1

[0048] One specific embodiment of the present application discloses a failure analysis standard process generation method, as shown in the following steps: Figure 1

[0049] S10: Constructing a failure analysis knowledge graph;

[0050] S11: Based on the failure analysis knowledge graph, constructing physical structure class and fault phenomenon class entities of a new part and their attributes and relationships, and establishing entity alignment relationship between the new part and the old part according to entity similarity;

[0051] S12: Taking out the detection item entity of the fault phenomenon class of the new part in turn, obtaining the detection item entity of the old part corresponding to the current detection item entity according to the entity alignment relationship, and constructing the detection process class entity associated with the current detection item entity of the new part and its attributes and relationships according to the physical structure class and detection process class entities associated with the detection item entity of the old part;

[0052] S13: Generating a failure analysis standard process according to the detection process class entity of the new part and its attributes and relationships.

[0053] Compared with the prior art, the failure analysis standard process generation method provided by the embodiment fully utilizes and mines historical data of failure analysis, constructs a failure analysis knowledge graph, and automatically generates a failure analysis standard process of a new part through knowledge entity fusion, knowledge entity reasoning and knowledge entity correlation query of the new and old parts.

[0054] Specifically, the failure analysis knowledge graph is constructed through step S10, and the specific process is divided into steps S101-S106, as shown below:

[0055] S101: According to the bill of materials and the netlist, the physical structure class entity and its attributes and relationships are established, and the physical structure class entity includes part entity and line entity.

[0056] It should be noted that the bill of materials is also called product structure table and material list, which disassembles the raw materials, spare parts and assemblies of a product, and records each single material according to the material code, name, specification, unit quantity, loss, etc. in the order of the manufacturing process, arranges it into a list, and is used to obtain assembly information in the field of failure analysis; the netlist is used to describe the connection relationship between parts in the assembly.

[0057] In this embodiment, the parts and lines and their related attributes are obtained by analyzing the bill of materials, and are instantiated as part entities and line entities in the knowledge graph; the association relationship between the parts and the lines, and the order between the parts on the same line are obtained by analyzing the netlist, the triple relationship between the part entities and the line entities is established, and the order attribute value is added in the part entity, so that the maintenance engineer can detect the parts according to the order and reduce repeated operations.​

[0058] Exemplarily, the Component, Description, VendorName and VendorPartNo in the bill of materials (BOM) are respectively mapped to the attribute values of the part name, detailed description of the part, supplier name and supplier code in the part entity. If there is a case where an assembly includes multiple parts, the assembly code to which the part belongs is added to the attributes of the part entity.

[0059] It should be noted that each entity has a unique entity identifier, which is generated according to preset rules, for example, by splicing multiple attribute values, so as to facilitate subsequent direct acquisition of related attribute values according to the entity identifier.

[0060] S102: According to the test file and the test log, the fault phenomenon class entity and its attributes and relationships are established, and the fault phenomenon class entity includes the detection item entity and the test command entity.

[0061] It should be noted that the test file is a collection of detection items for each fault phenomenon of the assembly, and is used to obtain the detection items; the test log is a log file recorded by the test machine when testing the assembly according to the test file, and is used to obtain the test commands, test standard return values and other related information required during testing.

[0062] When the test file and the test log are parsed, the respective to-be-parsed contents are intercepted from the original files according to the file format and content identifier; based on the TF-IDF algorithm, the to-be-parsed content of the test file content is extracted for keywords, and a keyword is obtained for each detection item. The keyword is used as the detection item name, and the specific test command and standard return value when the test is passed corresponding to each detection item are extracted from the to-be-parsed content of the test log, and are respectively instantiated as the detection item entity and the test command entity; and the association relationship between the detection item entity and the test command entity is established according to the keyword.

[0063] S103: According to the historical failure analysis standard process, the detection process class entity and its attributes and relationships are established, and the detection process class entity includes the detection step entity and the detection result entity.

[0064] It should be noted that the historical failure analysis standard process is generally a manually written detection process, which is used to obtain each detection step and its determination result under each detection process, and the determination result indicates the faulty part.

[0065] The embodiment extracts operation step name, operation type, operation step description, operation step return value type, and next step operation name of normal and abnormal detection by analyzing the historical failure analysis standard process, generates operation step number, number and name of the detection process to which the operation step belongs according to preset rules, and instantiates the detection step entity. Since the next step operation of normal and abnormal detection is also instantiated as a detection step entity, the next step operation of normal and / or abnormal detection between detection step entities is established by using the next step operation. The judgment result and related part name are extracted, instantiated as a detection result entity, and the corresponding physical structure class entity is obtained according to the related part name, and the association between the detection result entity and the detection step entity and the physical structure class entity is established.

[0066] Specifically, the operation type of the detection step includes visual inspection, measurement, disassembly, and sending command.

[0067] S104: According to the relationship between the historical failure analysis standard process and the test file, the relationship between the detection project entity and the detection step entity is established.

[0068] It should be noted that the detection file and the historical failure analysis standard process used for each fault phenomenon are known based on domain business knowledge. One test file extracts multiple detection projects, multiple detection projects correspond to one detection process, and one detection process has multiple detection steps. The embodiment realizes the acquisition of the detection step according to the detection project name by establishing the association between the detection project name and the detection process name, that is, the relationship between the detection project entity and the detection step entity is established.

[0069] S105: Based on the physical structure class entity, the operation step description attribute value in the detection step entity is processed to obtain a step segmentation result, and the relationship between the physical structure class entity and the detection step entity is established according to the step segmentation result.

[0070] It should be noted that the operation step description attribute value in the detection step entity is a specific operation step manually written, and usually one operation step involves a detection object and a detection method. The detection object is the physical structure class entity that has been established in the knowledge graph. In order to improve the segmentation accuracy, a corpus database is constructed based on the corpus data in the industrial analysis field, and the established physical structure class entity name is used as an auxiliary library, and the segmentation result is obtained by jieba segmentation.

[0071] The segmentation result of the operation step description is used to extract the relationship between entities, including:

[0072] The step segmentation result is compared with the entity name of the physical structure class entity. If they are the same, the entity corresponding to the entity name is taken as the tail entity, the detection step entity corresponding to the step segmentation result is taken as the head entity, the operation category attribute value in the head entity is taken as the operation relationship, and a triple of <head entity-operation relationship-tail entity> is established.

[0073] For example, the operation step description of the detection step entity G004 is: "measure whether the power size of C2000 is between (10dBm-30dBm), please input the measurement result". After jieba segmentation, "measure whether the power size of A is between (B-C), please input the measurement result" is obtained, where A=C2000, B=10dBm, and C=30dBm. A, B, and C are compared with the entity name of the physical structure class entity. A corresponds to the part entity with the entity name C2000. Then a triple of <G004-measure-C2000> is established, that is, the detection step entity G004 is associated with the part entity C2000.

[0074] Considering some special cases, such as non-standard operation step description, and operation category is sending command, the operation step description involves test command. For these special cases, the operation step description described above which cannot obtain entity relationship is further analyzed by step S105.

[0075] S106: Based on the failure analysis file writing specification, the part-of-speech sequence template of each operation category is extracted. After part-of-speech tagging of the step segmentation result, the part-of-speech sequence template is compared with the part-of-speech sequence template. For the consistent step segmentation result, the relationship between the physical structure class entity and the detection step entity, or the relationship between the detection step entity and the test command entity is established according to the part-of-speech sequence template.

[0076] For example, the part-of-speech sequence template is: v n x n. The operation step description of the detection step entity K002 is: "check pre-charge resistor, pre-charge relay". After part-of-speech tagging, it is obtained: check / v pre-charge resistor / n, / x pre-charge relay / n, which is consistent with the part-of-speech sequence template. If the entities corresponding to the pre-charge resistor and the pre-charge relay can be queried in the knowledge graph, then the relationship of check is taken as the triple, and the relationships between the detection step entity and the battery pack, the pre-charge resistor, and the pre-charge relay entity are established respectively.

[0077] Through the construction of the knowledge graph, the utilization rate of historical data is improved, and it is convenient for operators to obtain the failure analysis detection process according to the fault phenomenon. When a new part fails, the detection steps and detection results of the old part are used, the knowledge entities are fused and reasoned, and the failure analysis standard process of the new part is automatically generated.

[0078] In step S11, based on the concept system of the knowledge entities in the failure analysis knowledge graph, the physical structure class and the fault phenomenon class entities of the new part and their attributes and relationships are constructed according to the above steps S101-S102. According to the entity similarity, the entity alignment relationship between the new part and the old part is established, including:

[0079] According to the entity category, for each new part entity that has not established an alignment relationship, compare the entity name, the first text of the entity name and its attributes and attribute values, and the second text of the entity name and its adjacent entity name with each old part entity in turn, and establish a <new part entity-alignment-old part entity> triple when the similarity of the entity name, the first text of the entity name and its attributes and attribute values, and the second text of the entity name and its adjacent entity name is the largest and greater than a threshold value.

[0080] Specifically, for each type of entity, the alignment relationship is established in turn according to the following three methods:

[0081] ① Use the word2vec model to obtain the word vector of the entity name of each new part and old part, calculate the similarity between the word vectors of the entity name of each new part and each old part, and establish a <new part entity-alignment-old part entity> triple when the word vectors corresponding to the new part entity and the old part entity have the maximum similarity and are greater than a threshold value;

[0082] ② For the remaining new parts that have not established an alignment relationship and each old part, splice the entity name and its attributes and attribute values into a first text, obtain the first text vector using the Oag-Bert model, calculate the similarity between the first text vectors of each remaining new part and each old part, and establish a <new part entity-alignment-old part entity> triple when the first text vectors corresponding to the new part entity and the old part entity have the maximum similarity and are greater than a threshold value;

[0083] ③ For the remaining new parts that have not established an alignment relationship and each old part, splice the entity name and its adjacent entity name into a second text, obtain the second text vector using the Oag-Bert model, calculate the similarity between the second text vectors of each remaining new part and each old part, and establish a <new part entity-alignment-old part entity> triple when the second text vectors corresponding to the new part entity and the old part entity have the maximum similarity and are greater than a threshold value.

[0084] Preferably, when splicing the second text, the adjacent entity is not considered as a detection process class entity.

[0085] In step S11, the component entity, circuit entity, inspection item entity, and test command entity of the new component are aligned with the old component. However, there are cases where some entities of the new component cannot be aligned with the old component. In this embodiment, an interactive interface is provided for the misaligned entities, which receives the alignment relationship between the new and old entities configured by the operator.

[0086] Next, in step S12, starting with each inspection item entity of the new part, and by aligning with the inspection item entities of the old part, the associated inspection process class entities are identified sequentially, and the inspection process class entities of the new part, along with their attributes and relationships, are inferred, including:

[0087] Obtain the inspection process entity associated with the inspection item entity of the old part, and extract each inspection step entity in turn. If all physical structure entities or test command entities associated with the current inspection step entity have an alignment relationship with the current new part, then copy the current inspection step entity and its attributes to the inspection step entity and its attributes of the new part, associate the current inspection item entity of the new part, and establish the relationship between the inspection step entity of the new part and the physical structure entity or test command entity of the new part according to the alignment relationship, so as to obtain the inspection process entity associated with the current inspection item entity of the new part, its attributes and relationships.

[0088] It should be noted that a single inspection project is associated with multiple inspection steps, and each inspection step involves one or more inspection objects. Depending on the type of operation of the inspection step, the inspection objects may include parts, circuits, or test commands. Each inspection step under the aligned old part inspection project is retrieved. If its associated inspection objects are all aligned with the new part, then the inspection step can be reused for the new part and associated with the new part's inspection project. Based on the new part information and preset rules, the inspection process number to which the inspection step belongs is modified, and it is instantiated as a new part's inspection step entity. According to the new part's inspection objects aligned with the old inspection objects, the relationship between the new part's inspection step entity and the new part's physical structure entity or test command entity is established.

[0089] For example, the entity name of the new part's inspection item "No Signal" has a similarity of 0.92 with the entity name of the old part's inspection item "Poor Signal," which is greater than the preset threshold of 0.8, thus establishing an alignment relationship. In the inspection steps associated with "Poor Signal": the first inspection step's operation type is visual inspection, the inspection step number is A001, the number of the inspection process it belongs to is APP2D502, the name of the inspection process it belongs to is "signal," and the associated inspection object is part R3300. This part is aligned with part R3301 in the new part. Therefore, the attribute information of the first inspection step is copied to the inspection step of the new part's inspection item "No Signal." Based on the inspection area and the name of the new part, the number of the inspection process it belongs to is modified to APP3D201, and it is associated with part R3301.

[0090] Preferably, if the detection step entity of the old part is associated with the detection result entity, the detection result entity of the new part is also reused, and the detection result entity of the new part is established and associated with the detection step entity of the new part.

[0091] In the prior art, a large number of detection items are generated by a testing machine during detection of a part, and the number of the detection items is in the level of hundreds of thousands. If the detection items are manually compared between the new and old parts, a huge labor and time cost will be consumed. Based on the entity alignment, the entity reuse and relationship migration are automatically performed by using the alignment relationship between the entities and taking the detection step as the dimension, so that the work efficiency and accuracy are greatly improved.

[0092] In step S13, in order to realize the standardization of the detection process, the operation step description manually written in the historical file is no longer used, and a plurality of operation step description templates are subdivided according to the operation type, and the entities associated with the detection step entity are automatically filled.

[0093] Specifically, according to the detection process class entity of the new part and the attributes and relationships thereof, a failure analysis standard process is generated, including:

[0094] The detection step entities in the detection process class entity of the new part are sequentially taken out as basic steps;

[0095] According to the operation type attribute value of each detection step entity, the corresponding operation step description template is obtained, the physical structure class entity name or the test command attribute of the test command entity associated with the detection step entity is filled into the operation step description template, the operation step description is obtained, and the operation step description is added to the basic step;

[0096] According to the next operation relationship of each detection step entity under the condition of normal and / or abnormal detection, the corresponding detection step entity is obtained, the step number attribute value is taken as the corresponding next operation, and the next operation is added to the basic step;

[0097] According to the detection result entity associated with the detection step entity, the physical structure class entity name associated with the detection result entity is taken as the detection fault point, and the detection fault point is added to the basic step;

[0098] The basic step is output to obtain the failure analysis standard process.

[0099] Exemplarily, the failure analysis standard process shown in Table 1 is generated. The step number and operation type in Table 1 are directly derived from the attribute value of the detection step entity, the operation step description is filled into the corresponding operation step description template of the inspection and measurement according to the associated part entity, the next step is the step number of the detection step entity corresponding to the next operation of the detection normal (0) and / or detection abnormal (1), and the fault point is the fault part corresponding to the detection result (0) and / or the detection result (1) when the detection is normal according to the associated detection result.

[0100] Table 1: Example of failure analysis standard process

[0101]

[0102]

[0103] Compared with the prior art, the failure analysis standard process generation method of the embodiment fully utilizes historical data, mines the relationship of failure analysis, constructs a failure analysis knowledge graph, assists in realizing related knowledge query of the failure analysis standard process, and assists in realizing manual modification of the failure analysis standard process. A unified failure analysis standard process is generated through the knowledge graph, errors and differences caused by manual writing are solved, and the failure analysis standard process is facilitated to be optimized. Through knowledge entity fusion, knowledge entity reasoning and knowledge entity association query of new and old parts, entity reuse and relationship migration are automatically performed with the detection step as the dimension, the failure analysis standard process of the new part is generated, the work efficiency and accuracy are improved, and the automation and standardization of the failure analysis standard process are realized.

[0104] Embodiment 2

[0105] Another embodiment of the application discloses a failure analysis standard process generation system, so as to realize the failure analysis standard process generation method in embodiment 1. The specific implementation mode of each module is referred to the corresponding description in embodiment 1. The system comprises:

[0106] The knowledge entity fusion module is configured to construct the physical structure class and the fault phenomenon class entity and the attributes and relationships thereof of the new part based on the failure analysis knowledge graph, and establish the entity alignment relationship between the new part and the old part according to the entity similarity.

[0107] The knowledge entity reasoning module is configured to sequentially take out the detection item entity of the fault phenomenon class of the new part, acquire the detection item entity of the old part corresponding to the current detection item entity according to the entity alignment relationship, and construct the detection flow class entity associated with the current detection item entity of the new part according to the physical structure class and the detection flow class entity associated with the detection item entity of the old part.

[0108] The detection flow generation module is configured to generate the failure analysis standard flow according to the detection flow class entity of the new part and the attributes and relationships thereof.

[0109] It should be noted that in the system of the present embodiment, the failure analysis knowledge graph is implemented by a knowledge graph construction module, which is configured to construct the failure analysis knowledge graph according to historical data, including:

[0110] According to the bill of materials and the netlist, the physical structure class entity and the attributes and relationships thereof are established, and the physical structure class entity includes a part entity and a circuit entity.

[0111] According to the test file and the test log, the fault phenomenon class entity and the attributes and relationships thereof are established, and the fault phenomenon class entity includes a detection item entity and a test command entity.

[0112] According to the historical failure analysis standard flow, the detection flow class entity and the attributes and relationships thereof are established, and the detection flow class entity includes a detection step entity and a detection result entity.

[0113] According to the relationship between the historical failure analysis standard flow and the test file, the relationship between the detection item entity and the detection step entity is established.

[0114] Based on the physical structure class entity, the operation step description attribute in the detection step entity is subjected to word segmentation processing to obtain a step word segmentation result, and the relationship between the physical structure class entity and the detection step entity is established according to the step word segmentation result.

[0115] It should be noted that the detection steps in the failure analysis standard flow are usually many, and after being exported, all the detection step information can be seen as shown in Table 1, and the steps are executed by comparison during use. Preferably, by using the system of the present embodiment, the corresponding detection flow is queried according to the fault phenomenon of the defective assembly, the first operation is automatically displayed, and after the maintenance engineer executes the operation steps, the actual detection value is input, the system automatically judges and displays the corresponding fault point, or continues to display the next operation, which is convenient for the use of the maintenance engineer.

[0116] Exemplarily, in Table 1, step A001 is associated with part C9400, the associated detection result is that when the detection is abnormal, the fault point is C9400, and the step A002 establishes the next operation relationship of the detection normal, then the maintenance engineer displays the operation category and operation step description of step A001 when performing failure analysis, when the part C9400 is detected to have a problem, the actual detection value 1 is input, the fault point C9400 is obtained, and one failure analysis is completed; when the part C9400 is detected to have no problem, the actual detection value 0 is input, step A002 is executed, and the operation category and operation step description of step A002 are displayed again, when the part R3300 is detected to have no problem, the actual detection value 0 is input, step A003 is executed, and the fault point is obtained until one failure analysis is completed.

[0117] Since the failure analysis standard process generation system of the embodiment and the foregoing generation method are related to each other and can be mutually referred to, repeated description is not given here. Since the system embodiment and the method embodiment have the same principle, the system embodiment also has the corresponding technical effects of the method embodiment.

[0118] Those skilled in the art can understand that all or part of the processes of the above-mentioned embodiments can be completed by a computer program instructing related hardware, and the program can be stored in a computer readable storage medium. The computer readable storage medium is a disk, an optical disk, a read-only memory, a random access memory, or the like.

[0119] The above is only a preferred specific embodiment of the present application, but the protection scope of the present application is not limited to this. Any person skilled in the art can easily think of changes or replacements within the technical range disclosed by the present application, which should be covered within the protection scope of the present application.

Claims

1. A failure analysis standard procedure generation method characterized by, The method comprises the following steps: constructing a failure analysis knowledge graph, constructing physical structure class and fault phenomenon class entities and their attributes and relationships of the new part based on the failure analysis knowledge graph, and establishing an entity alignment relationship between the new part and the old part according to entity similarity; sequentially taking out the detection item entity of the fault phenomenon class of the new part, obtaining the detection item entity of the old part corresponding to the current detection item entity according to the entity alignment relationship, and constructing the detection process class entity and its attributes and relationships associated with the current detection item entity of the new part according to the detection item entity of the old part. According to the detection process class entity and its attributes and relationships of the new part, a failure analysis standard process is generated. The method comprises the following steps: According to the bill of materials and the netlist, the physical structure class entity and its attributes and relationships are established, and the physical structure class entity includes part entity and line entity. According to the test file and the test log, the fault phenomenon class entity and its attributes and relationships are established, and the fault phenomenon class entity includes detection item entity and test command entity. According to the historical failure analysis standard process, the detection process class entity and its attributes and relationships are established, and the detection process class entity includes detection step entity and detection result entity. According to the relationship between the historical failure analysis standard process and the test file, the relationship between the detection item entity and the detection step entity is established. Based on the physical structure class entity, the operation step description attribute value in the detection step entity is segmented to obtain a step segmentation result, and the relationship between the physical structure class entity and the detection step entity is established according to the step segmentation result.

2. The failure analysis standard flow generation method according to claim 1, characterized by, The method comprises the following steps: The step segmentation result is compared with the entity name of the physical structure class entity, if they are the same, the entity corresponding to the entity name is taken as the tail entity, the detection step entity corresponding to the step segmentation result is taken as the head entity, the operation type attribute value in the head entity is taken as the operation relationship, and the triple of <head entity-operation relationship-tail entity> is established.

3. The failure analysis standard flow generation method according to claim 1, characterized by, The method further comprises the following steps: Based on the failure analysis file writing specification, the part-of-speech sequence template of each operation type is extracted, the part-of-speech sequence template is compared with the part-of-speech sequence template after the step segmentation result is part-of-speech tagged, the relationship between the physical structure class entity and the detection step entity or the relationship between the detection step entity and the test command entity is established for the consistent step segmentation result according to the part-of-speech sequence template.

4. The failure analysis standard flow generation method according to claim 1, characterized by, The method comprises the following steps: According to the entity category, each new part entity that has not established an alignment relationship is compared with each old part entity in turn, the similarity of the entity name, the first text obtained by splicing the entity name and its attributes and attribute values, and the second text obtained by splicing the entity name and its adjacent entity name is compared, and when the similarity is the largest and greater than a threshold value, the corresponding new part entity and old part entity are established as the triple of <new part entity-alignment-old part entity>.

5. The failure analysis standard flow generation method according to claim 3, wherein A word vector of the entity name is obtained using a word2vec model, and a text vector of the first text and the second text is obtained using an Oag-Bert model.

6. The failure analysis standard flow generation method according to claim 1, wherein The detection flow class entity associated with the detection item entity of the old part is obtained, and each detection step entity is sequentially taken out; if all physical structure class entities or test command entities associated with the current detection step entity have an alignment relationship with the current new part, the current detection step entity and its attributes are copied as the detection step entity and its attributes of the new part, the detection step entity of the new part is associated with the current detection item entity of the new part, and the relationship between the detection step entity of the new part and the physical structure class entity or the test command entity of the new part is established according to the alignment relationship, so as to obtain the detection flow class entity and its attributes and relationships associated with the current detection item entity of the new part. The failure analysis standard flow is generated according to the detection flow class entity and its attributes and relationships of the new part, including:

7. The failure analysis standard flow generation method according to claim 1, wherein Each detection step entity in the detection flow class entity of the new part is sequentially taken out as a basic step; According to the operation type attribute value of each detection step entity, a corresponding operation step description template is obtained, the physical structure class entity name or the test command attribute of the test command entity associated with the detection step entity is filled into the operation step description template, an operation step description is obtained, and the operation step description is added to the basic step; According to the next operation relationship of each detection step entity under the condition of detection normality and / or abnormality, a corresponding detection step entity is obtained, and the step number attribute value is taken as a corresponding next operation, which is added to the basic step; According to the detection result entity associated with the detection step entity, the physical structure class entity name associated with the detection result entity is taken as a detection fault point, which is added to the basic step; The basic step is output to obtain the failure analysis standard flow. The knowledge entity fusion module is configured to construct the physical structure class and the fault phenomenon class entity and their attributes and relationships of the new part based on the failure analysis knowledge graph, and establish the entity alignment relationship between the new part and the old part according to the entity similarity.

8. A failure analysis standard procedure generation system characterized by comprising: The knowledge entity reasoning module is configured to sequentially take out the detection item entity of the fault phenomenon class of the new part, obtain the detection item entity of the old part corresponding to the current detection item entity according to the entity alignment relationship, and construct the detection flow class entity and its attributes and relationships associated with the current detection item entity of the new part according to the physical structure class and the detection flow class entity associated with the detection item entity of the old part. The detection flow generation module is configured to generate the failure analysis standard flow according to the detection flow class entity and its attributes and relationships of the new part. The failure analysis standard flow generation system also includes: The knowledge graph construction module is configured to construct the failure analysis knowledge graph according to historical data, including: ​ ​ According to the bill of materials and the netlist, a physical structure class entity and attributes and relations thereof are established, the physical structure class entity including a part entity and a circuit entity; according to the test file and the test log, a fault phenomenon class entity and attributes and relations thereof are established, the fault phenomenon class entity including a detection item entity and a test command entity; According to a historical failure analysis standard process, a detection process class entity and attributes and relations thereof are established, the detection process class entity including a detection step entity and a detection result entity; According to a relationship between the historical failure analysis standard process and the test file, a relationship between the detection item entity and the detection step entity is established; Based on the physical structure class entity, a tokenization processing is performed on an operation step description attribute in the detection step entity, a step tokenization result is obtained, and a relationship between the physical structure class entity and the detection step entity is established according to the step tokenization result.

Images