DFMEA automatic updating method combining after-sales data and fault tree

Abstract

The invention discloses a DFMEA automatic updating method combining after-sales data and a fault tree. The method comprises the steps that S1, a product BOM interface is constructed based on an SQL database and JAVA; s2, according to the product after-sales service data, constructing a product fault mode knowledge base interface based on JAVA and Vue; s3, constructing a product fault tree interface according to the product BOM interface and the product fault mode knowledge base interface; and S4, realizing automatic updating of the DFMEA table of the product. Based on intelligent data integration of after-sales service data and a product fault mode knowledge base and an automatic DFMEA updating mechanism based on a fault tree, key fault links of a product can be intelligently identified, and a DFMEA table can be automatically generated and updated. According to the method, functional characteristics and structural characteristics of the product can be embodied, and potential fault mode analysis of the product in the design stage can be effectively carried out, so that key links possibly causing product faults can be intelligently identified.

Description

Technical Field

[0001] The present invention relates to the technical field of product reliability in manufacturing industry, and in particular to a DFMEA automatic updating method combining after-sales data and a fault tree. Background Art

[0002] The structure of manufacturing products has become more complex as their functions become more and more abundant, and the functions of products produced by different manufacturing companies have become more and more similar year by year, which makes users pay more attention to product performance when choosing products. Since reliability is one of the important indicators to measure the performance of manufacturing products, manufacturers have paid more attention to product reliability in recent years, and reliability technology has become an important means to improve the performance of manufacturing products. Compared with other electronic products with simple structures, the loading mode and relative movement form between components of manufacturing products are more complex. During their working process, they will be affected by mechanical stress, electrical stress and thermal stress, so the product failure mode is also more complex. Predicting and analyzing a certain potential failure mode of a product alone cannot represent the overall reliability of manufacturing products. How to predict the potential failure mode of manufacturing products through efficient and accurate methods and establish a failure mode library for them is the key to improving the reliability level of manufacturing products.

[0003] DFMEA (Design Failure Mode and Effects Analysis) is a reliability analysis method specifically for the product design stage, used to identify and evaluate potential failure modes in product design and their impact on product performance at an early stage. DFMEA is used to analyze the functions of the defined system, subsystem or related components, the relationship between its internal elements and the relationship between the elements outside the system boundary, so as to identify possible design defects and minimize the risk of potential failures. FTA (Fault Tree Analysis) is a systematic reliability analysis method used to identify and analyze the root causes that may cause product or system failures. The FTA method constructs a fault tree, decomposes the top event into a series of basic events and traces the failure path, thereby revealing the cause and effect relationship of the failure. However, in traditional DFMEA failure analysis and FTA failure attribution, it is easy to cause fuzzy and overlapping definitions and interpretations of failure modes, failure causes, and failure effects, which in turn causes incomplete analysis of the potential failure modes of the product. Summary of the invention

[0004] In order to overcome the shortcomings of the above technologies, the present invention provides a DFMEA automatic update method combining after-sales data and fault trees. Based on the intelligent data integration of after-sales service data and product failure mode knowledge base and the automatic DFMEA update mechanism based on the fault tree, the key failure links of the product can be intelligently identified and the DFMEA table can be automatically generated and updated. The method can reflect the functional characteristics and structural features of the product, and can effectively analyze the potential failure modes of the product in the design stage, thereby intelligently identifying the key links that may cause product failures.

[0005] Terminology explanation:

[0006] 1. SQL: Structured Query Language.

[0007] 2. BOM: Bill of Materials.

[0008] 3. Doccano: data annotation platform.

[0009] 4. API: Application Programming Interface, application programming interface.

[0010] 5. HTTP: Hyper Text Transfer Protocol.

[0011] 6. NLP: Natural Language Processing.

[0012] 7. RPN: Risk Priority Number, risk priority number.

[0013] 8. Axios: An HTTP client used to send asynchronous HTTP requests in JavaScript applications.

[0014] The technical solution adopted by the present invention to overcome the technical problems is:

[0015] A DFMEA automatic update method combining after-sales data and fault tree comprises the following steps:

[0016] Step S1, constructing a product BOM interface based on SQL database and JAVA: based on product design requirements and similarity, decompose the product into hierarchical structures, clarify the assembly relationship, component functions and the relationship between hierarchical structures; design the SQL product database table structure to record component information and relationships, and construct the product BOM interface;

[0017] Step S2: Based on the product after-sales service data, build a product failure mode knowledge base interface based on JAVA and Vue: sort out the product after-sales service data, classify the failure descriptions and associate them with the components in the product BOM table, create an SQL product failure mode database, and realize intelligent identification of failures through JAVA and Vue;

[0018] Step S3, constructing a product fault tree interface according to the product BOM interface and the product failure mode knowledge base interface: creating a product fault tree structure in SQL according to the product BOM and the product failure mode knowledge base, so that the product fault tree is associated with the product BOM and the product failure mode knowledge base; constructing a recursive query and update mechanism of the fault tree to achieve dynamic display; using the JAVA backend to operate the fault tree data, and developing a JAVA fault tree interface;

[0019] Step S4, realize automatic update of product DFMEA table: design DFMEA SQL database to store DFMEA data, realize automatic update through dynamic interaction with product BOM, product failure mode knowledge base and product fault tree; use JAVA backend to synchronously update DFMEA data so that DFMEA data is automatically updated when product BOM and product failure mode change; develop front end based on Vue to realize intelligent display and interaction of DFMEA.

[0020] Furthermore, step S1 specifically includes the following steps:

[0021] Step S11, designing a database of product BOM table: first, create a product BOM table containing basic product information, including at least product ID, product model and product description, identify products of different models through the product BOM table, set the product BOM table as the core table, and use it to manage parent products, sub-components and hierarchical nesting relationships;

[0022] Step S12, constructing a backend framework of the product BOM interface based on JAVA: defining an entity class corresponding to the product BOM database in the backend framework, and implementing query, level expansion, and structure update business logic according to the hierarchical structure of the product BOM; writing an API controller to implement adding new product materials, updating product BOM relationships, and deleting product BOM structures;

[0023] Step S13, using the Vue framework to build the front-end interface of the product BOM: Axios is used to realize data communication between the front-end and the back-end, and the front-end calls the back-end API interface through HTTP requests to realize the functions of adding, deleting, modifying and querying.

[0024] Further, in step S11, the product BOM table is used to display the product management interface of all product basic information, including adding, deleting, modifying and querying operations; the product BOM hierarchy is displayed in a form so that the user can expand and view the product hierarchy relationship step by step, and the user can click on the current BOM data to export an Excel file.

[0025] Furthermore, step S2 specifically includes the following steps:

[0026] Step S21, pre-define several failure mode categories in combination with the failure types appearing in the product after-sales service data, and set the pre-defined failure mode categories as labels in Doccano;

[0027] Step S22, building an association between historical fault description texts and fault mode category labels: importing historical fault description texts into Doccano one by one, and matching each fault description text with the most relevant fault mode category label through manual annotation;

[0028] Step S23, building an association between historical fault description texts and components in the product BOM: setting each component of the product in the product BOM as a label in Doccano, and matching each fault description text with the corresponding product component category label through manual labeling;

[0029] Step S24, create an SQL database structure for historical fault mode data, construct a historical fault mode SQL database table structure for storing product components and fault modes, and import the fault description text data annotated in step S22 into the historical fault mode SQL database; when the newly input fault description text data is annotated, automatically add the new fault description text data to the historical fault mode SQL database so that the fault mode knowledge base can be updated in real time;

[0030] Step S25: Build a product failure mode knowledge base interface based on JAVA and Vue, use the JAVA backend framework to process the storage and logical query of failure data, use the stored historical failure mode data, and use the NLP model to perform text analysis and probability matching on the newly input failure description text, identify potential failure modes, and return the failure mode that best matches it.

[0031] Furthermore, step S3 specifically includes the following steps:

[0032] Step S31, constructing the SQL database table structure of the product fault tree: constructing the SQL product fault tree database structure including the relationship between the fault event name, the fault event type and the hierarchical structure according to the product BOM and the historical fault mode knowledge base; associating the product fault tree SQL database structure with the product BOM and the product fault mode knowledge base database;

[0033] Step S32, constructing a recursive query and update mechanism for the product fault tree: constructing a recursive query in the product fault tree SQL database, starting from the top-level node, recursively obtaining all child nodes, and when the product BOM or product failure mode knowledge base data is updated, triggering the data update operation of the fault tree to synchronize the fault tree with the latest data;

[0034] Step S33, constructing the fault tree backend logic and frontend interface: using the JAVA backend service to process the addition, deletion, modification and query operations of the fault tree data, and developing a tree-structured fault tree interface based on the Vue framework;

[0035] Step S34, automatic update of the fault tree and database linkage: when the product BOM or product failure mode knowledge base is updated, the update of the fault tree SQL database is automatically triggered, that is, when a failure mode of a component is added, the system automatically creates a fault node for the component and updates the fault tree front-end interface.

[0036] Further, in step S33, the fault tree backend logic and frontend interface are constructed, specifically including:

[0037] Interact with the fault tree SQL database through the JAVA backend to process the addition, deletion, modification and query operations of the fault number: provide the functions of adding, deleting, modifying and querying fault tree nodes through the node management module, recursively obtain the structural information of the fault tree from the fault tree SQL database, and return it to the front end in a tree data format;

[0038] The fault tree interface is developed based on the Vue framework. The fault tree nodes and hierarchical relationships are displayed through the visual tree component so that users can see the fault hierarchy structure and realize the addition, deletion, modification and query interaction of nodes. After updating the data in the fault tree SQL database, the fault tree interface is refreshed through the front-end asynchronous request to realize real-time data update.

[0039] Furthermore, step S4 specifically includes the following steps:

[0040] Step S41, constructing the core database structure of the DFMEA table: designing the DFMEA database structure so that real-time automatic updates can be achieved through dynamic data interaction with the product BOM, product failure mode knowledge base and product failure tree;

[0041] Step S42, design a product similarity algorithm: obtain product similarity by calculating the edit distance between the product BOM and the new product BOM; set a similarity threshold, if the similarity between the new product BOM and the product BOM exceeds the similarity threshold, the new product can inherit the initial DFMEA table from the product;

[0042] Step S43, designing an intelligent update mechanism for the DFMEA table: obtaining the failure mode associated with each component from the product failure mode knowledge base. When a component in the product BOM is added or modified, the system automatically updates the failure mode of the component to the DFMEA table by querying the product failure mode knowledge base; for each component in the product BOM, the system automatically updates these failure modes to the DFMEA table by searching for the associated ID; by extracting severity and frequency data from the product failure mode knowledge base, the system automatically associates these data with the corresponding fields of the DFMEA table; according to the structure of the fault tree, when the failure mode of a component affects its parent component, the system automatically increases its severity rating, and the system determines the potential impact of the fault on the system by recursively analyzing the hierarchical relationship of the fault tree; according to the product fault tree, when a component fails, the system automatically calculates the impact path of the fault on the entire product system, and updates these results to the "fault impact" field in the DFMEA table; when the fault tree changes, the system re-evaluates the position of the fault in the entire fault propagation path through automatic calculation, and updates the impact range and severity rating of the fault;

[0043] Step S44, automatic calculation of risk priority number RPN in product DFMEA: the system automatically calculates the RPN value based on the latest severity, frequency and detection, and the system automatically fills in the standardized rules in the product failure mode knowledge base and updates the RPN in real time;

[0044] Step S45: construct the intelligent update system framework of the product DFMEA table:

[0045] First, build the backend service development: the data synchronization module synchronizes the product BOM, product failure mode knowledge base and product fault tree table in the SQL database through the JAVA backend service; whenever the product BOM or product failure mode knowledge base is updated, the system automatically triggers the update operation of the DFMEA table; and handles the automatic calculation of severity, frequency and detection, combined with the recursive calculation of the impact range of the fault tree, so that the system automatically updates the DFMEA table, obtains product information through the API interface and returns the failure mode and failure impact information;

[0046] Secondly, the intelligent display and interaction of the DFMEA table is realized through the Vue framework: the front-end interface dynamically displays the information in the DFMEA table, including failure mode, severity, probability of occurrence, impact range and RPN value, by calling the back-end API interface. When the background automatically updates the DFMEA table, the front-end interface notifies the user through the web socket protocol, prompting that the DFMEA table has been updated, and allows the user to view the new DFMEA results immediately, so that the user can modify certain fields in the DFMEA on the front-end interface. The system submits all updates to the back-end through the API and updates the records in the DFMEA database at the same time.

[0047] Furthermore, in step S42, the product similarity between the product BOM and the new product BOM is calculated, specifically, the cosine similarity between the two is calculated, and the calculation formula is as follows:

[0048] (1)

[0049] In formula (1), A represents the feature vector of product BOM, and B represents the feature vector of new product BOM. Indicates that the feature vector currently being calculated is Quantity, represents the feature vector dimension, The first Quantity, represents the first feature vector B A quantity.

[0050] Furthermore, in step S44, the calculation formula of the risk priority number RPN is as follows:

[0051] RPN=S×O×D(2)

[0052] In formula (2), S represents severity, O represents frequency, and D represents detection.

[0053] The beneficial effects of the present invention are:

[0054] 1. The present invention automatically updates the product BOM, product failure mode knowledge base and product failure tree through the system, reflects the functional characteristics and structural characteristics of the product, enables the new product DFMEA table to be automatically updated interactively, and ensures the comprehensiveness of the product DFMEA construction.

[0055] 2. The present invention uses the cosine similarity algorithm to provide an initial DFMEA table for new products by calculating the similarity value between the new product BOM and the product BOM, so as to more quickly and accurately identify potential failures in the new product design stage, intelligently identify the key links that may cause product failures, and ensure the timeliness, completeness and accuracy of the new product design process. BRIEF DESCRIPTION OF THE DRAWINGS

[0056] Figure 1 The present invention is a flowchart of a DFMEA automatic update method combining after-sales data and fault tree according to an embodiment of the present invention.

[0057] Figure 2 This is a schematic diagram of the structure of the fault tree of product C according to an embodiment of the present invention. DETAILED DESCRIPTION

[0058] In order to facilitate those skilled in the art to better understand the present invention, the present invention is further described in detail below in conjunction with the accompanying drawings and specific embodiments. The following is only exemplary and does not limit the protection scope of the present invention.

[0059] The present invention discloses a DFMEA automatic updating method combining after-sales data and a fault tree, comprising the following steps:

[0060] Step S1, constructing a product BOM interface based on SQL database and JAVA: based on product design requirements and similarity, decompose the product into hierarchical structures, clarify the assembly relationship, component functions and the relationship between hierarchical structures; design the SQL product database table structure to record component information and relationships, and construct the product BOM interface;

[0061] Step S2: Based on the product after-sales service data, build a product failure mode knowledge base interface based on JAVA and Vue: sort out the product after-sales service data, classify the failure descriptions and associate them with the components in the product BOM table, create an SQL product failure mode database, and realize intelligent identification of failures through JAVA and Vue;

[0062] Step S3, constructing a product fault tree interface according to the product BOM interface and the product failure mode knowledge base interface: creating a product fault tree structure in SQL according to the product BOM and the product failure mode knowledge base, so that the product fault tree is associated with the product BOM and the product failure mode knowledge base; constructing a recursive query and update mechanism of the fault tree to achieve dynamic display; using the JAVA backend to operate the fault tree data, and developing a JAVA fault tree interface;

[0063] Step S4, realize automatic update of product DFMEA table: design DFMEA SQL database to store DFMEA data, realize automatic update through dynamic interaction with product BOM, product failure mode knowledge base and product fault tree; use JAVA backend to synchronously update DFMEA data so that DFMEA data is automatically updated when product BOM and product failure mode change; develop front end based on Vue to realize intelligent display and interaction of DFMEA.

[0064] In order to better understand the above technical solution, the exemplary embodiments of the present invention will be described in detail below in conjunction with the accompanying drawings. These are only exemplary embodiments of the present invention. However, it should be understood that the present invention can also be implemented in various forms and is not limited to the embodiments described herein. These embodiments are intended to enable those skilled in the art to understand the present invention more clearly and thoroughly.

[0065] This embodiment is described by taking a certain injection molding machine factory as an example, and includes the following steps:

[0066] Step S1: construct product BOM interface based on SQL database and JAVA.

[0067] (1) Some fields of the product BOM structure of an injection molding machine factory are shown in Table 1 below.

[0068] Table 1 Some fields of the product BOM structure of an injection molding machine factory

[0069]

[0070] (2) Obtain the initial DFMEA of the new product by calculating the similarity between the product BOM and the new product BOM.

[0071] The component hierarchy in the BOM table of a certain injection molding machine factory's products is shown in Table 2 below.

[0072] Table 2 Component hierarchy in the BOM table of a certain injection molding machine factory

[0073]

[0074] The component hierarchy in the BOM table of a newly designed product of an injection molding machine factory is shown in Table 3 below.

[0075] Table 3 Component hierarchy in the BOM table of a new design product of an injection molding machine factory

[0076]

[0077] Construct the feature vectors of the product and the newly designed product respectively, as follows:

[0078] First, convert the product BOM and the new product BOM into feature vectors: the component names in the product BOM and the new product BOM are combined into a "feature word bag" containing all possible components. After de-reorganization, all components form the following feature word bags: ["moving template", "tail plate", "fixed template", "screw injection mechanism", "elbow connecting rod mechanism", "mold adjustment mechanism", "lubrication mechanism", "ejection cylinder", "hydraulic valve group", "hydraulic mechanism", "fuselage"].

[0079] Secondly, for the product BOM and the newly designed product BOM, a binary vector is created based on the feature word bag. The feature vector corresponds to a component in the feature word bag, and 0 indicates that the component does not exist and 1 indicates that the component exists.

[0080] Then, the components in the product are mapped to the positions in the feature word bag, and the feature vector is set to A; the components in the newly designed product are mapped to the positions in the feature word bag, and the feature vector is set to B. The feature vector A and the feature vector B are expressed as follows:

[0081] A=[1,1,1,0,1,1,0,1,1,0,1],

[0082] B=[1, 1, 0, 1, 0, 1, 1, 0, 0, 1, 1].

[0083] Among them, the components in the BOM table of the newly designed product: the moving template, the tail plate, the screw injection mechanism, the mold adjustment mechanism, the lubrication mechanism, the hydraulic mechanism, and the fuselage have corresponding positions of 1 in the feature vector B, and the remaining positions are 0.

[0084] Calculate the cosine similarity between the product BOM and the new product BOM:

[0085]

[0086] In the above formula, A represents the feature vector of product BOM, and B represents the feature vector of new product BOM. Indicates that the feature vector currently being calculated is Quantity, represents the feature vector dimension, The first Quantity, represents the first feature vector B A quantity.

[0087] Calculate the dot product of the product BOM and the new design product BOM feature vector as follows:

[0088] Compute the Euclidean norm: and are the Euclidean norms of eigenvector A and eigenvector B, respectively, as follows:

[0089]

[0090]

[0091]

[0092]

[0093]

[0094] Set a similarity threshold. If the similarity between the newly designed product BOM and the product BOM reaches the preset similarity threshold, the initial DFMEA table of the new product can be generated based on the DFMEA table of the product.

[0095] (3) Realize data update and interaction between the front-end and back-end of product BOM table based on JAVA.

[0096] Step S2: Build a product failure mode knowledge base interface based on JAVA and Vue according to product after-sales service data.

[0097] In this embodiment, the historical fault description text in the product after-sales service data is annotated using the data annotation platform Doccano, specifically: L1: substandard assembly quality, L2: appearance problems, L3: missing parts, L4: paint powder accumulation and agglomeration, L5: single part quality problems. The above predefined fault categories are set as labels in the data annotation platform Doccano.

[0098] Based on the data in the product BOM, Doccano is used to annotate the main components of the product, specifically: S1: moving template, S2: fixed template, S3: mold adjustment mechanism, S4: hydraulic mechanism, S5: fuselage. The above components are set as labels in the data annotation platform Doccano.

[0099] In this embodiment, it is assumed that some faults occur in product C. According to the product BOM structure, the annotated fault modes are mapped to matching components, as shown in Table 4.

[0100] Table 4 Product C failure mode description text-component-fault label matching table

[0101]

[0102] Build a historical failure mode SQL database table that stores product components and failure modes.

[0103] Step S3: construct a product fault tree interface according to the product BOM interface and the product failure mode knowledge base interface.

[0104] In this embodiment, first, the relationship between the fault event name, fault event type and hierarchical structure of the product fault tree is determined according to the product BOM and the product fault mode knowledge base, wherein the fault event type includes the fault mode, fault impact, etc. The structural diagram of the fault tree of product C is as follows: Figure 2 shown.

[0105] Secondly, a recursive query is constructed in the product fault tree SQL database to dynamically display fault information according to the fault tree structure.

[0106] Then, the JAVA backend service is used to process the addition, deletion, modification and query operations of the fault tree data, and the Vue framework is used to display the fault tree structure on the front end to realize interaction with users.

[0107] Step S4: Automatically update the product DFMEA table.

[0108] In this embodiment, the SQL database of DFMEA is designed to store DFMEA data, and automatic updating is achieved through dynamic interaction with the product BOM, product failure mode knowledge base and product fault tree.

[0109] The system identifies the failure modes in the failure mode knowledge base through SQL database query, searches for specific component information in the new product BOM, and obtains the failure description, severity, frequency, and detectability information of the corresponding component in the failure mode knowledge base.

[0110] The system automatically calculates the risk priority number RPN in the product DFMEA based on the latest severity, frequency and detection. The calculation formula of the risk priority number RPN is: RPN=S×O×D, where S represents severity, O represents frequency, and D represents detection. The system automatically fills the newly added component information into the DFMEA table and updates the failure mode and impact content related to the component.

[0111] Real-time synchronization and update of DFMEA is achieved through JAVA: When a new product BOM or product failure mode knowledge base is updated, the JAVA backend automatically synchronizes the updated content of the DFMEA table through a data trigger mechanism, and the Vue frontend dynamically displays the updated DFMEA content. Users can view the latest failure modes and related information of each component of the new product.

[0112] The above only describes the basic principles and preferred embodiments of the present invention. Those skilled in the art may make many changes and improvements based on the above description, and these changes and improvements should fall within the protection scope of the present invention.

Claims

1. A DFMEA automatic update method combining after-sales data and fault tree, characterized in that: The steps include: Step S1, constructing a product BOM interface based on SQL database and JAVA: based on product design requirements and similarity, decomposing the product into hierarchical structures, clarifying the assembly relationship, component functions and the relationship between hierarchical structures; Design SQL product database table structure to record component information and relationships, and build product BOM interface; Step S2: Based on the product after-sales service data, build a product failure mode knowledge base interface based on JAVA and Vue: sort out the product after-sales service data, classify the failure descriptions and associate them with the components in the product BOM table, create an SQL product failure mode database, and realize intelligent identification of failures through JAVA and Vue; Step S3, constructing a product fault tree interface according to the product BOM interface and the product failure mode knowledge base interface: creating a product fault tree structure in SQL according to the product BOM and the product failure mode knowledge base, so that the product fault tree is associated with the product BOM and the product failure mode knowledge base; constructing a recursive query and update mechanism of the fault tree to achieve dynamic display; using the JAVA backend to operate the fault tree data, and developing a JAVA fault tree interface; Step S4, realize automatic update of product DFMEA table: design DFMEA SQL database to store DFMEA data, realize automatic update through dynamic interaction with product BOM, product failure mode knowledge base and product fault tree; use JAVA backend to synchronously update DFMEA data so that DFMEA data is automatically updated when product BOM and product failure mode change; develop front end based on Vue to realize intelligent display and interaction of DFMEA.

2. The DFMEA automatic update method combining after-sales data and fault tree according to claim 1 is characterized in that: Step S1 specifically includes the following steps: Step S11, designing a database of product BOM table: first, create a product BOM table containing basic product information, including at least product ID, product model and product description, identify products of different models through the product BOM table, set the product BOM table as the core table, and use it to manage parent products, sub-components and hierarchical nesting relationships; Step S12, constructing a backend framework of the product BOM interface based on JAVA: defining an entity class corresponding to the product BOM database in the backend framework, and implementing query, level expansion, and structure update business logic according to the hierarchical structure of the product BOM; writing an API controller to implement adding new product materials, updating product BOM relationships, and deleting product BOM structures; Step S13, using the Vue framework to build the front-end interface of the product BOM: Axios is used to realize data communication between the front-end and the back-end, and the front-end calls the back-end API interface through HTTP requests to realize the functions of adding, deleting, modifying and querying.

3. The DFMEA automatic update method combining after-sales data and fault tree according to claim 2 is characterized in that: In step S11, the product BOM table is used to display the product management interface of all product basic information, including adding, deleting, modifying and querying operations; The product BOM hierarchy is displayed in a form so that users can expand and view the product hierarchy relationship level by level and click on the current BOM data to export an Excel file.

4. The DFMEA automatic update method combining after-sales data and fault tree according to claim 1 is characterized in that: Step S2 specifically includes the following steps: Step S21, pre-define several failure mode categories in combination with the failure types appearing in the product after-sales service data, and set the pre-defined failure mode categories as labels in Doccano; Step S22, building an association between historical fault description texts and fault mode category labels: importing historical fault description texts into Doccano one by one, and matching each fault description text with the most relevant fault mode category label through manual annotation; Step S23, building an association between historical fault description texts and components in the product BOM: setting each component of the product in the product BOM as a label in Doccano, and matching each fault description text with the corresponding product component category label through manual labeling; Step S24, create an SQL database structure for historical fault mode data, construct a historical fault mode SQL database table structure for storing product components and fault modes, and import the fault description text data annotated in step S22 into the historical fault mode SQL database; when the newly input fault description text data is annotated, automatically add the new fault description text data to the historical fault mode SQL database so that the fault mode knowledge base can be updated in real time; Step S25: Build a product failure mode knowledge base interface based on JAVA and Vue, use the JAVA backend framework to process the storage and logical query of failure data, use the stored historical failure mode data, and use the NLP model to perform text analysis and probability matching on the newly input failure description text, identify potential failure modes, and return the failure mode that best matches it.

5. The DFMEA automatic update method combining after-sales data and fault tree according to claim 1 is characterized in that: Step S3 specifically includes the following steps: Step S31, constructing the SQL database table structure of the product fault tree: constructing the SQL product fault tree database structure including the relationship between the fault event name, the fault event type and the hierarchical structure according to the product BOM and the historical fault mode knowledge base; associating the product fault tree SQL database structure with the product BOM and the product fault mode knowledge base database; Step S32, constructing a recursive query and update mechanism for the product fault tree: constructing a recursive query in the product fault tree SQL database, starting from the top-level node, recursively obtaining all child nodes, and when the product BOM or product failure mode knowledge base data is updated, triggering the data update operation of the fault tree to synchronize the fault tree with the latest data; Step S33, constructing the fault tree backend logic and frontend interface: using the JAVA backend service to process the addition, deletion, modification and query operations of the fault tree data, and developing a tree-structured fault tree interface based on the Vue framework; Step S34, automatic update of the fault tree and database linkage: when the product BOM or product failure mode knowledge base is updated, the update of the fault tree SQL database is automatically triggered, that is, when a failure mode of a component is added, the system automatically creates a fault node for the component and updates the fault tree front-end interface.

6. The DFMEA automatic update method combining after-sales data and fault tree according to claim 5 is characterized in that: In step S33, the fault tree backend logic and frontend interface are constructed, which specifically includes: Interact with the fault tree SQL database through the JAVA backend to process the addition, deletion, modification and query operations of the fault number: provide the functions of adding, deleting, modifying and querying fault tree nodes through the node management module, recursively obtain the structural information of the fault tree from the fault tree SQL database, and return it to the front end in a tree data format; The fault tree interface is developed based on the Vue framework. The fault tree nodes and hierarchical relationships are displayed through the visual tree component so that users can see the fault hierarchy structure and realize the addition, deletion, modification and query interaction of nodes. After updating the data in the fault tree SQL database, the fault tree interface is refreshed through the front-end asynchronous request to realize real-time data update.

7. The DFMEA automatic update method combining after-sales data and fault tree according to claim 1 is characterized in that: Step S4 specifically includes the following steps: Step S41, constructing the core database structure of the DFMEA table: designing the DFMEA database structure so that real-time automatic updates can be achieved through dynamic data interaction with the product BOM, product failure mode knowledge base and product failure tree; Step S42, design a product similarity algorithm: obtain product similarity by calculating the edit distance between the product BOM and the new product BOM; set a similarity threshold, if the similarity between the new product BOM and the product BOM exceeds the similarity threshold, the new product can inherit the initial DFMEA table from the product; Step S43, design an intelligent update mechanism for the DFMEA table: obtain the failure mode related to each component from the product failure mode knowledge base. When a component in the product BOM is added or modified, the system automatically updates the failure mode of the component to the DFMEA table by querying the product failure mode knowledge base; for each component in the product BOM, the system automatically updates these failure modes to the DFMEA table by searching for the associated ID; by extracting severity and frequency data from the product failure mode knowledge base, the system automatically associates these data with the corresponding fields of the DFMEA table; according to the structure of the fault tree, when the failure mode of a component affects its parent component, the system automatically increases its severity rating, and the system determines the potential impact of the fault on the system by recursing the hierarchical relationship of the fault tree; according to the product fault tree, when a component fails, the system automatically calculates the impact path of the fault on the entire product system, and updates these results to the "fault impact" field in the DFMEA table; when the fault tree changes, the system re-evaluates the position of the fault in the entire fault propagation path through automatic calculation, and updates the impact range and severity rating of the fault; Step S44, automatic calculation of risk priority number RPN in product DFMEA: the system automatically calculates the RPN value based on the latest severity, frequency and detection, and the system automatically fills in the standardized rules in the product failure mode knowledge base and updates the RPN in real time; Step S45: construct the intelligent update system framework of the product DFMEA table: First, build the backend service development: the data synchronization module synchronizes the product BOM, product failure mode knowledge base and product fault tree table in the SQL database through the JAVA backend service; whenever the product BOM or product failure mode knowledge base is updated, the system automatically triggers the update operation of the DFMEA table; and handles the automatic calculation of severity, frequency and detection, combined with the recursive calculation of the impact range of the fault tree, so that the system automatically updates the DFMEA table, obtains product information through the API interface and returns the failure mode and failure impact information; Secondly, the intelligent display and interaction of the DFMEA table is realized through the Vue framework: the front-end interface dynamically displays the information in the DFMEA table, including failure mode, severity, probability of occurrence, impact range and RPN value, by calling the back-end API interface. When the background automatically updates the DFMEA table, the front-end interface notifies the user through the web socket protocol, prompting that the DFMEA table has been updated, and allows the user to view the new DFMEA results immediately, so that the user can modify certain fields in the DFMEA on the front-end interface. The system submits all updates to the back-end through the API and updates the records in the DFMEA database at the same time.

8. The DFMEA automatic update method combining after-sales data and fault tree according to claim 7 is characterized in that: In step S42, the product similarity between the product BOM and the new product BOM is calculated, specifically, the cosine similarity between the two is calculated, and the calculation formula is as follows: （1） In formula (1), A represents the feature vector of product BOM, and B represents the feature vector of new product BOM. Indicates that the feature vector currently being calculated is Quantity, represents the feature vector dimension, The first Quantity, The first eigenvector of B A quantity.

9. The DFMEA automatic update method combining after-sales data and fault tree according to claim 7, characterized in that: In step S44, the calculation formula of the risk priority number RPN is as follows: RPN=S×O×D(2) In formula (2), S represents severity, O represents frequency, and D represents detection.

Images