A ship maintainability design recommendation method and system

Abstract

The present application discloses a ship maintainability design recommendation method and system, which relates to the field of ship maintainability design technology. The method includes: obtaining ship maintainability design knowledge and ship DFMEA knowledge, processing them, and constructing a ship maintainability design knowledge base; constructing a ship maintainability design ontology based on the ship maintainability design knowledge base; the ship maintainability design ontology is a structured representation of the knowledge base; obtaining text data input by designers, using natural language processing technology to perform entity recognition on the text data, and obtaining entities and their corresponding classes; based on the ship maintainability design ontology and the entities and their corresponding classes, obtaining the ship's maintainability design strategy. The present application realizes the effective management of ship maintainability design knowledge and ship DFMEA knowledge by constructing a ship maintainability design ontology, and can provide objective ship maintainability design strategies based on the data input by designers, thereby improving maintenance efficiency.

Description

Technical Field

[0001] The present application relates to the technical field of ship maintainability design, and in particular to a ship maintainability design recommendation method and system. Background Art

[0002] In recent years, with the rapid development of the shipbuilding industry, the structure of ships has become increasingly complex and their functions have become increasingly complete. Correspondingly, the requirements for ship maintainability have also increased. The so-called ship maintainability refers to the ability of ships to be repaired and maintained conveniently and quickly when they encounter failures or damage. Good ship maintainability can shorten maintenance time, reduce maintenance costs, and improve the availability and reliability of ships. To achieve ship maintainability design, it is necessary to comprehensively apply professional knowledge in multiple fields such as ship design, mechanical design, electrical engineering, materials science, ergonomics, reliability engineering, and maintenance engineering.

[0003] Although the continuous progress of ship design and construction technology has accumulated a wealth of knowledge related to maintainability design, this knowledge is often scattered in different fields, and there is a large amount of knowledge that cannot be effectively expressed and transferred, such as skills and experience that are difficult to express. At present, there is a lack of a systematic sharing mechanism, which makes it difficult to effectively manage and apply this knowledge, and therefore it cannot be effectively integrated into the research and development process of new models. In addition, the traditional maintainability design method mainly relies on manual experience judgment, there is a lot of repetitive work, and there may be problems of strong subjectivity and inefficiency. It is difficult to fully learn from and utilize past design experience, and it is difficult to provide quantitative improvement suggestions. Summary of the invention

[0004] The purpose of this application is to provide a ship maintainability design recommendation method and system, which can realize the effective management of knowledge and automatically provide corresponding ship maintainability design strategies, thereby improving the efficiency and reliability of ship maintenance.

[0005] To achieve the above objectives, this application provides the following solutions:

[0006] In a first aspect, the present application provides a ship maintainability design recommendation method, the ship maintainability design recommendation method comprising:

[0007] Acquire ship knowledge data, process the ship knowledge data, and construct a ship maintainability design knowledge base; the ship knowledge data includes ship maintainability design knowledge and ship DFMEA knowledge;

[0008] According to the ship maintainability design knowledge base, a ship maintainability design ontology is constructed; the ship maintainability design ontology is a structured representation of the ship maintainability design knowledge base;

[0009] Obtain text data input by the designer, perform entity recognition on the text data using natural language processing technology, and obtain entities and their corresponding classes;

[0010] Based on the ship maintainability design ontology and entities and their corresponding classes, a ship maintainability design strategy is obtained.

[0011] In a second aspect, the present application provides a ship maintainability design recommendation system, the ship maintainability design recommendation system comprising: a knowledge base module, a knowledge representation module and a maintainability design recommendation module;

[0012] The knowledge base module is used to obtain ship knowledge data, process the ship knowledge data, and construct a ship maintainability design knowledge base; the ship knowledge data includes ship maintainability design knowledge and ship DFMEA knowledge;

[0013] The knowledge representation module is used to construct a ship maintainability design ontology according to the ship maintainability design knowledge base; the ship maintainability design ontology is a structured representation of the ship maintainability design knowledge base;

[0014] The maintainability design recommendation module is used to obtain text data input by designers, use natural language processing technology to perform entity recognition on the text data, obtain entities and their corresponding classes, and obtain the ship's maintainability design strategy based on the ship maintainability design ontology and the entities and their corresponding classes.

[0015] According to the specific embodiments provided in this application, this application discloses the following technical effects:

[0016] The present application provides a ship maintainability design recommendation method and system, the method comprising: obtaining ship knowledge data, processing the ship knowledge data, constructing a ship maintainability design knowledge base, the ship knowledge data including ship maintainability design knowledge and ship DFMEA knowledge; constructing a ship maintainability design ontology based on the ship maintainability design knowledge base, the ship maintainability design ontology being a structured representation of the ship maintainability design knowledge base; obtaining text data input by designers, using natural language processing technology to perform entity recognition on the text data, obtaining entities and their corresponding classes; obtaining the ship maintainability design strategy based on the ship maintainability design ontology and the entities and their corresponding classes. The present application realizes the effective management and structured representation of ship maintainability design knowledge and DFMEA knowledge by constructing a ship maintainability design ontology, thereby integrating rich design knowledge into the new model development process and improving the availability and effectiveness of knowledge. Moreover, the present application utilizes natural language processing technology to perform entity recognition on the text data input by the designer, which can automatically extract relevant entities and their classes, and combined with the ship maintainability design ontology, it can automatically generate maintainability design strategies for specific ships. This process can reduce human intervention and subjective influences, thereby improving the efficiency and reliability of maintainability design. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0018] Figure 1 This is an application environment diagram of a ship maintainability design recommendation method in one embodiment of the present application;

[0019] Figure 2 A schematic diagram of a process flow of a ship maintainability design recommendation method provided in one embodiment of the present application;

[0020] Figure 3 for Figure 2 A detailed flow chart of the steps for building a ship maintainability design knowledge base;

[0021] Figure 4 for Figure 2 A detailed flow chart of the steps for constructing the ship maintainability design ontology;

[0022] Figure 5 for Figure 2 A detailed flow chart of entity recognition and ship maintainability design strategy query steps;

[0023] Figure 6 A visual diagram of the relevant knowledge of the cooling system failure mode and the corresponding maintenance design strategy in one embodiment of the present application;

[0024] Figure 7 A schematic diagram of the functional modules of a ship maintainability design recommendation system provided in one embodiment of the present application. DETAILED DESCRIPTION

[0025] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.

[0026] The ship maintainability design recommendation method provided in the embodiment of the present application can be applied to Figure 1 In the application environment shown. Among them, the terminal 102 communicates with the server 104 through the network. The data storage system can store the data that the server 104 needs to process. The data storage system can be set up separately, integrated on the server 104, or placed on the cloud or other servers. The terminal 102 can send the acquired ship knowledge data and the text data entered by the designer to the server 104. After receiving the ship knowledge data and the text data entered by the designer, the server 104 processes the ship knowledge data, constructs a ship maintainability design knowledge base, and constructs a ship maintainability design ontology based on the ship maintainability design knowledge base; uses natural language processing technology to perform entity recognition on the text data, obtains entities and their corresponding classes, and obtains the ship's maintainability design strategy based on the ship maintainability design ontology and the entities and their corresponding classes. The server 104 can feedback the obtained ship's maintainability design strategy to the terminal 102. In addition, in some embodiments, the ship maintainability design recommendation method can also be implemented independently by the server 104 or the terminal 102. For example, the terminal 102 can directly perform corresponding processing on the ship knowledge data and the text data entered by the designer, or the server 104 can obtain the ship knowledge data and the text data entered by the designer from the data storage system and process them.

[0027] Among them, the terminal 102 can be but is not limited to various desktop computers, laptops, smart phones, tablet computers, Internet of Things devices and portable wearable devices, and the server 104 can be implemented by an independent server or a server cluster composed of multiple servers, and can also be a cloud server.

[0028] In an exemplary embodiment, Figure 2As shown, a ship maintainability design recommendation method is provided. The method is executed by a computer device, and can be executed by a computer device such as a terminal or a server alone, or by a terminal and a server together. In the embodiment of the present application, the method is applied to Figure 1 The server 104 in the example is used as an example to illustrate the method, which includes the following steps 201 to 204. Among them:

[0029] Step 201, acquiring ship knowledge data, and processing the ship knowledge data to construct a ship maintainability design knowledge base; the ship knowledge data includes ship maintainability design knowledge and ship DFMEA (Design Failure Mode and Effects Analysis) knowledge.

[0030] Furthermore, ship knowledge data is obtained and processed to construct a ship maintainability design knowledge base, such as Figure 3 As shown, specifically including:

[0031] Step 2011, obtaining ship knowledge data; ship knowledge data includes ship maintainability design knowledge, maintainability design cases and ship DFMEA knowledge accumulated historically; the sources of ship knowledge data are also very wide, including design documents, maintenance records, fault reports, design cases and technical manuals, etc. These data are diverse and complex, but can be generally classified into three data formats, including structured data (such as database tables), semi-structured data (such as XML, JSON), and unstructured data (such as text files, PDF documents).

[0032] Step 2012, performing data cleaning and format conversion operations on the ship knowledge data to obtain intermediate format data (such as JSON, CSV) for further processing; the data cleaning includes removing redundant information and noise to ensure data quality.

[0033] Step 2013, performing data mapping and conversion operations on the intermediate format data, including entity extraction, attribute extraction and relationship extraction, to obtain RDF (Resource Description Framework) triple data; specifically: first, based on the mapping rules, identify entities (such as "faulty component", "maintenance task" and "maintenance personnel", etc.) from the intermediate format; then, extract their attributes (such as "name", "type" and "description", etc.), and identify the relationship between entities (such as "faulty component requires maintenance task", "maintenance task is completed by maintenance personnel", etc.); finally, convert the intermediate format data into the form of RDF triples (subject, predicate, object), such as (ship, name, "A"), (faulty component, maintenance task, task ID123), (task ID123, executor, maintenance personnel), etc.

[0034] Step 2014, the RDF triple data is integrated, including disambiguation and coreference resolution, to obtain the ship maintainability design knowledge base to ensure the accuracy and consistency of knowledge; wherein, disambiguation refers to identifying and resolving the different references of entities with the same name. For example, "pump" can refer to multiple types of pumps, such as water pumps, oil pumps, hydraulic pumps, etc. In a maintenance manual or troubleshooting guide, if only "pump" is mentioned without providing more information, it may cause ambiguity. Disambiguation can be achieved through context-based word sense disambiguation, which extracts context features and compares them to select the interpretation that best fits the context. For example, it can be determined whether "pump" refers to a water pump or an oil pump through the description of the surrounding text. Coreference resolution refers to identifying and resolving the phenomenon that different named entities refer to the same thing. Ships are usually composed of multiple systems and subsystems, which may appear under different names. For example, "propulsion system" and "engine" may refer to the same system. Co-reference resolution can be achieved using a rule-driven co-reference resolution method, which is based on linguistic rules such as pronoun agreement, name matching, and semantic similarity to determine whether they refer to the same entity. By combining the context and pronoun agreement rules, it can be determined whether "propulsion system" and "engine" refer to the same object.

[0035] By converting the complex and diverse ship knowledge data into a standardized RDF format, RDF is a universal data model suitable for representing any form of knowledge, which can facilitate the unification and management of knowledge.

[0036] Step 202, based on the ship maintainability design knowledge base, construct a ship maintainability design ontology, that is, an ontology composed of ship maintainability design knowledge and ship DFMEA knowledge; specifically, the ship maintainability design ontology is a structured representation of the ship maintainability design knowledge base, and the ship maintainability design ontology is constructed and used as the core of the knowledge base, thereby ensuring the standardization, searchability and scalability of knowledge.

[0037] In the ontology construction process, the following terms are usually included:

[0038] Concept: A concept is a core abstract entity in the knowledge domain and is a generalized representation of real things. For example, "ship parts", "fault types" and "maintenance strategies" are all concepts. Concepts are abstractions of actual objects, events, states, etc., and are generally further refined into classes during the modeling process.

[0039] Class: A class is a realization of a concept and a classification collection of similar things. For example, the concept "ship parts" can become a class, which contains specific parts such as "main engine", "auxiliary engine", etc. Classes can be seen as more formal expressions of concepts and are usually used to classify and organize knowledge in ontology construction. Each class usually contains individuals with certain common characteristics.

[0040] Data Property: Data property is a property of a class, which describes the characteristics of a class or instance. For example, the class "main engine" may have properties such as "power" and "model", which represent different dimensions or characteristics of the class. Attributes are used in the ontology to describe the characteristics of a class or an instance of a class, and often have data types (such as numeric values, text) and constraints.

[0041] Object Property: Object properties are the associations between classes. For example, there can be a relationship between the class "Ship Components" and the class "Failure Mode" such as "Possible Failures". Object properties are usually the key part of building an ontology, helping to indicate how to establish associations between different classes.

[0042] Instance: An instance is a specific manifestation of a class, an actual object that corresponds to an entity. Each class can have multiple instances. For example, a specific engine model in the "Main Engine" class is an instance. Instantiation can be seen as a description of a specific thing in reality.

[0043] When constructing the ship maintainability design ontology, according to different knowledge levels and needs, it can be analyzed from three perspectives, namely, the structure domain, the design domain, and the fault domain; among them: the structure domain is used to represent the hierarchical relationship and association relationship of the ship system structure, including concepts such as system, subsystem, component, structural composition and functional requirements; the design domain is used to describe the parameters and requirements of the ship maintainability design, and express the overall design information and maintainability design information in the ship design stage, including concepts such as design attributes, design constraints, maintainability requirements, maintainability design parameters and maintenance design strategies; the fault domain is used to represent potential failure modes and their causes and effects, and describes the failure information of ship parts, product information and structural relationships related to the failed parts, as well as the analysis results of the failure, including concepts such as failed products, structural composition, failure modes, failure causes, failure effects, maintenance methods, severity and occurrence probability.

[0044] Furthermore, based on the ship maintainability design knowledge base, a ship maintainability design ontology is constructed, such as Figure 4 As shown, specifically including:

[0045] Step 2021, determine the domain and scope of the ship maintainability design ontology; cooperate with domain experts to clarify the scope and details of the knowledge that needs to be represented, including all concepts, attributes and relationships related to ship maintainability design knowledge and ship DFMEA knowledge.

[0046] Step 2022, based on the domain and scope of the ship maintainability design ontology, extract core concepts related to building the ship maintainability design ontology (such as "ship components", "fault types", "maintenance strategies", etc.) from the ship maintainability design knowledge base, and confirm the extracted concepts with domain experts to ensure that they meet actual needs and usage scenarios.

[0047] Step 2023, according to the attributes and functions of the core concepts, define the class and attribute structure; generally define one or more top-level classes in the ontology, representing the most abstract concepts in the domain, and the top-level classes provide a framework and reference for subsequent class definitions. For example, "ship system", "design information", "fault analysis" and the like can be set as top-level classes; further refine under the top-level classes to construct subclasses, following the "top-down" approach, starting from the top-level classes, and refine layer by layer to concretize into subclasses. According to the attributes and functions corresponding to the core concepts, the core concepts in the knowledge domain are defined as corresponding classes, each class represents an entity type in the domain, such as "ship system", "subsystem", "fault mode", etc., define data attributes for each class, clarify the data type and constraints of the data attributes, and establish hierarchical relationships and associations through object attributes to form a hierarchical structure (such as "ship system" has "propulsion system", "power system", etc.).

[0048] Step 2024, using an ontology editing tool to create classes, data attributes and object attributes, and using an ontology language to describe and store the ship maintainability design knowledge base according to classes, data attributes and object attributes, to obtain a ship maintainability design ontology, thereby ensuring the standardization and extensibility of the ontology.

[0049] Furthermore, the ontology editing tool is Protégé software; and the ontology language is OWL language.

[0050] In this application, based on the structure domain, design domain, and fault domain, the ship maintainability design ontology describes the basic concepts of ship maintainability design knowledge and ship DFMEA knowledge through different classes.

[0051] The main categories include: Ship, Component, Maintenance Design Strategy, Maintenance Activity, Design Parameters, Reliability Parameters, Maintainability Parameters, General Design, Product Entity, Failure Mode, Failure Cause, Failure Effect, Detection Method, Risk Priority Number (RPN), Product Hierarchy and Product Criticality.

[0052] The definitions of some object attributes (relationships) in the ship maintainability design ontology are shown in Table 1:

[0053] Table 1 Definition of some object attributes (relationships)

[0054]

[0055] The definitions of some data attributes in the ship maintainability design ontology are shown in Table 2:

[0056] Table 2 Definition of some data attributes

[0057]

[0058] This application uses the ontology editing tool Protégé to systematically represent and manage these data attributes and relationships. The ship maintainability design ontology can provide comprehensive data support for designers. On this basis, the entire knowledge base is analyzed, and the data in RDF triple format is stored in the ontology according to the corresponding class, data attribute, and object attribute, that is, an instance is constructed to realize the representation of knowledge.

[0059] Step 203, obtaining the text data input by the designer, and performing entity recognition (Named Entity Recognition, NER) on the text data using natural language processing (NLP) technology to obtain entities and their corresponding classes, such as ship model, design parameters and failure mode.

[0060] Furthermore, the text data input by the designer is obtained, and the text data is subjected to entity recognition using natural language processing technology to obtain entities and their corresponding classes (entities are specific manifestations of classes, which are actual objects corresponding to instances), such as Figure 5 As shown, specifically including:

[0061] Step 2031, obtaining text data input by designers, such as potential failure modes of new ship models, and converting them into a unified format to ensure data consistency and processability.

[0062] Step 2032, input the text data into the trained NER model, the text data will be automatically processed by the tokenizer and converted into the format required by the NER model (word unit ID, attention mask, paragraph tag), and the entity and its corresponding class will be obtained, such as equipment, component, failure mode, etc. For example, when the designer inputs a piece of text data, the trained NER model will output the recognized entity and its corresponding class. For the input text "the ship cooling system may have a coolant leak or pump failure", the model may output: [("ship", "equipment"), ("cooling system", "component"), ("coolant leak", "failure mode"), ("pump failure", "failure mode")].

[0063] The determination process of the trained NER model is as follows:

[0064] Construct a sample data set; each sample in the sample data set includes historical sample data and corresponding label data; the historical sample data is data obtained after removing noise, segmenting and marking parts of speech on historical text data, and the label data includes entity labels and their corresponding class labels. The specific steps are:

[0065] a. Remove noise from historical text data, such as extra spaces, special characters, and punctuation marks, and segment the text into independent words or phrases. Jieba is a commonly used tool for Chinese word segmentation, and NLTK is used for English word segmentation. For example, for the historical text data "There are some potential failure modes in the ship's cooling system, such as coolant leakage and pump failure", the word segmentation result is ["ship", "of", "cooling system", "in", "exist", "some", "potential", "of", "failure mode", ", ", "such as", "coolant leakage", "and", "pump failure"];

[0066] b. Label each word with its part of speech, such as noun, verb, adjective, etc. This can help the model better understand the role and relationship of words. For the above word segmentation results, it can be labeled as [("ship", noun), ("of", particle), ("cooling system", noun), ("in", particle), ("exist", verb), ("some", numeral), ("potential", adjective), ("of", particle), ("failure mode", noun), ("", punctuation), ("such as", conjunction), ("coolant leakage", noun), ("and", conjunction), ("pump failure", noun)].

[0067] Each sample input of the sample data set is converted into the input format of the BERT model (Bidirectional Encoder Representations from Transformers), namely, token IDs, attention masks and segment IDs, and input into the BERT model. Then, the parameters of the BERT model are fine-tuned to obtain a fine-tuned model, and the model performance is evaluated through a validation set, and the parameters of the fine-tuned model are adjusted again to improve the accuracy. Finally, the fine-tuned model is used as the trained NER model (Named Entity Recognition). Among them, the BERT model is a general pre-trained language model that can provide powerful text representation capabilities for various natural language processing tasks (including NER). The NER model is a task model specifically used to identify named entities (such as names, places, etc.) in text. The NER task can be implemented by fine-tuning based on BERT.

[0068] Step 204: obtaining a ship maintainability design strategy based on the ship maintainability design ontology and entities and their corresponding classes.

[0069] Based on the ship maintainability design ontology and entities and their corresponding classes, the ship maintainability design strategy is obtained, which specifically includes:

[0070] Step 2041, based on the ship maintainability design requirements and DFMEA methodology, SWRL (Semantic WebRule Language, a language that presents rules in a semantic way) language is used to create SWRL reasoning rules, and close cooperation with domain experts is carried out to ensure the accuracy and practicality of the rules. These rules will be used to identify and analyze potential failure modes and their impacts in ship design, and provide corresponding maintainability design knowledge, and update the ship maintainability design knowledge and ship DFMEA knowledge in the ship maintainability design ontology through reasoning.

[0071] SWRL rules consist of an antecedent and a consequent, usually written in the form of If-Then. The SWRL reasoning rules can be divided into three categories: functional rules, structural rules, and reliability rules.

[0072] Functional rules: This type of reasoning rules focuses on the functional failures and functional defects that may occur in ships. They are based on the DFMEA methodology and identify functional failure modes that may cause maintenance difficulties or complexity by analyzing different system functions and design elements. Examples include rules such as failure mode identification and failure cause identification.

[0073] hasFunction(?c,?x)^relatesToFailureMode(?c,?y)->hasFailureMode(?c,?y), the meaning of this rule is that if a component c has a specific function x and the specific function x may have failure y, then the component c has failure mode y.

[0074] hasFunction(?c,?x)^relatesToFailureMode(?c,?y)^hasCause(?y,?z)->hasRootCause(?c,?z), the meaning of this rule is that if a component c has a specific function x and the specific function x may have a failure y, and the failure mode y has a corresponding failure cause z, then the failure cause corresponding to the failure mode y of the component c is z.

[0075] Structural rules: This type of reasoning rules focuses on potential problems and impacts on ship structure design. They are based on the DFMEA methodology, taking into account the mutual influence and dependencies between different components and structural elements, and inferring different levels of fault impacts, including local impacts, higher-level impacts, and final impacts. Examples include structural hierarchy, risk priority number calculation, failure mode priority assessment, failure impact reasoning and other rules.

[0076] partOf(?x,?y)^partOf(?y,?z)->partOf(?x,?z), the meaning of this rule is that if part x is part of y and y is part of z, then it is inferred that x is part of z.

[0077] hasSeverity(?fm,?s)^hasOccurrence(?fm,?o)^hasDetection(?fm,?d)^multiply(?rpn,?s,?o)^multiply(?rpn,?d)->hasRPN(?fm,?rpn). The meaning of this rule is that if a fault mode fm has severity s, occurrence frequency o and detection difficulty d, then the risk priority number rpn of the fault mode fm can be calculated by multiplication.

[0078] hasRPN(?x,?z)^greaterThan(?z,?y)->isHighPriorityFailureMode(?x,true), the meaning of this rule is that if the risk priority number z of a failure mode x is higher than a specific threshold y, then the failure mode x is of high priority.

[0079] hasFailureMode(?c,?y)^hasEffect(?y,?e)->hasSystemImpact(?c,?e), the meaning of this rule is that if component c has failure mode y and the impact of failure mode y is e, then it can be inferred that component c has system-level impact e.

[0080] Maintainability rules: This type of reasoning rules focuses on the maintainability design of ship equipment and systems. Through the DFMEA methodology, design elements that may cause frequent equipment failures or require highly technical maintenance are identified, so that corresponding maintainability improvement measures can be formulated. Examples include failure mode-related maintenance design, identification of frequent maintenance caused by failure modes, and safety improvement rules for failure modes and maintenance activities.

[0081] relatesToFailureMode(?c,?y)^hasMaintenanceAction(?y,?a)^improvesMaintenance(?a,?z)->hasMaintenanceDesign(?c,?z), which means that component c has failure mode y, failure mode y corresponds to maintenance activity a, maintenance activity a can improve the maintenance process through maintainability design measure z, and the maintenance design measure corresponding to failure mode y of component c is z.

[0082] hasFailureMode(?c,?y)^causesFrequentMaintenance(?y,true)->requiresMaintenanceDesign(?c,?z), the meaning of this rule is that if component c has failure mode y, and failure mode y leads to frequent maintenance, then maintainability design measure z is required.

[0083] hasFailureMode(?c,?y)^hasMaintenanceAction(?y,?a)^hasSafetyRisk(?a,true)^improvesSafety(?z,true)->shouldImplementSafetyDesign(?c,?z), the meaning of this rule is that if maintenance activity a of failure mode y poses a safety hazard and maintainability design measure z can improve safety, then z needs to be implemented.

[0084] Step 2042, develop a reasoning engine based on SWRL to reason about the ship maintainability design ontology, responsible for processing SWRL rules, inferring new knowledge and maintaining consistency within the ontology. Based on the SWRL reasoning rules constructed in the previous step, select a reasoning engine that supports the SWRL language. This application uses the HermiT reasoning engine integrated into the Protégé software, uses the ontology editing tool Protégé to import the SWRL reasoning rules into the ship maintainability design ontology, and uses the HermiT reasoning engine to process the SWRL reasoning rules to establish the context of reasoning, update the ship maintainability design ontology, and obtain the updated ship maintainability design ontology.

[0085] Step 2043, based on the entity and its corresponding class, query the updated ship maintainability design ontology using SPARQL language to obtain the ship maintainability design strategy. The relevant process is as follows:

[0086] a. Retrieval of associated knowledge: Ontology queries are usually queried using SPARQL. Based on the definition of ontology O=(C, P, R, I), where C is a class set, P is a data attribute set, R is an object attribute set, and I is an instance set, the SPARQL statement for the query is constructed. Through the defined namespace and specific query template, qualified entities and relationships can be accurately retrieved from the ontology to obtain relevant knowledge information. For example, SELECT?failureMode?failureCauseWHERE{?CoolingSystemrdf:type ex:Component.?CoolingSystemex:hasFailureMode?failureMode.?failureMode ex:hasFailureCause?failureCause.} indicates that the failure modes of the cooling system components are queried and the failure causes of each failure mode are obtained.

[0087] b. Visual display of knowledge: Through the retrieval of associated knowledge, the failure modes and the failure causes, failure consequences, detection methods, risk priority numbers and maintenance design strategies of these failure modes can be obtained, and visualized through Protégé software for reference by designers, such as Figure 6 Demonstrated knowledge of cooling system failure modes and corresponding repair design strategies.

[0088] According to the query results, designers can obtain relevant knowledge and cases of similar failure modes, as well as corresponding maintainability design improvement measures. They can make maintainability design improvements based on this knowledge and recommended measures to improve the reliability and maintainability of the system.

[0089] By implementing the above steps 201 to 204, and constructing a ship maintainability design ontology, effective management and structured representation of ship maintainability design knowledge and DFMEA knowledge are achieved, so that rich design knowledge can be integrated into the new model design process, improving the availability and effectiveness of knowledge. Moreover, this application uses natural language processing technology to perform entity recognition on text data input by designers, which can automatically extract relevant entities and their classes, and combined with the ship maintainability design ontology, it can generate maintainability design strategies for specific ships. This process can reduce manual intervention and subjective influence, thereby improving the efficiency and reliability of maintainability design.

[0090] The present application also provides an application scenario, which applies the above-mentioned ship maintainability design recommendation method. Specifically: the ship maintainability design recommendation method provided in this embodiment can be applied in the ship maintainability design recommendation scenario. The ship maintainability design recommendation scenario includes a ship maintainability design generation link and a ship maintainability design display link. The ship maintainability design generation link is used to automatically generate the ship's maintainability design strategy, and the ship maintainability design display link is used to display the ship's maintainability design strategy; the ship maintainability design recommendation method provided in this embodiment belongs to the ship maintainability design generation link and the ship maintainability design display link.

[0091] Based on the same inventive concept, the embodiment of the present application also provides a ship maintainability design recommendation system for implementing the ship maintainability design recommendation method involved above. The implementation scheme for solving the problem provided by the system is similar to the implementation scheme recorded in the above method, so the specific limitations in one or more ship maintainability design recommendation system embodiments provided below can refer to the limitations of the ship maintainability design recommendation method above, and will not be repeated here.

[0092] In an exemplary embodiment, Figure 7 As shown, a ship maintainability design recommendation system is provided, comprising: a knowledge base module, a knowledge representation module and a maintainability design recommendation module;

[0093] The knowledge base module is used to acquire and store ship knowledge data, and process the ship knowledge data to construct a ship maintainability design knowledge base; the ship knowledge data includes ship maintainability design knowledge and ship DFMEA knowledge;

[0094] The knowledge representation module is used to construct a ship maintainability design ontology according to the ship maintainability design knowledge base; the ship maintainability design ontology is a structured representation of the ship maintainability design knowledge base;

[0095] The maintainability design recommendation module is used to obtain text data input by designers, use natural language processing technology to perform entity recognition on the text data, obtain entities and their corresponding classes, and obtain the ship's maintainability design strategy based on the ship maintainability design ontology and the entities and their corresponding classes.

[0096] The ship maintainability design recommendation system further includes: a knowledge reasoning module;

[0097] The knowledge reasoning module is used to create SWRL reasoning rules based on ship maintainability design requirements and DFMEA methodology using SWRL language, import the SWRL reasoning rules into the ship maintainability design ontology using the ontology editing tool Protégé, and process the SWRL reasoning rules using the HermiT reasoning engine to update the ship maintainability design ontology to obtain an updated ship maintainability design ontology; the SWRL reasoning rules include functional rule classes, structural rule classes and reliability rule classes.

[0098] The technical features of the above embodiments may be combined arbitrarily. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0099] This article uses specific examples to illustrate the principles and implementation methods of this application. The description of the above embodiments is only used to help understand the method and core ideas of this application. At the same time, for those skilled in the art, according to the ideas of this application, there will be changes in the specific implementation methods and application scope. In summary, the content of this specification should not be understood as limiting this application.

Claims

1. A ship maintainability design recommendation method, characterized in that: The recommended method for ship maintainability design includes: Acquire ship knowledge data, process the ship knowledge data, and construct a ship maintainability design knowledge base; the ship knowledge data includes ship maintainability design knowledge and ship DFMEA knowledge; According to the ship maintainability design knowledge base, a ship maintainability design ontology is constructed; the ship maintainability design ontology is a structured representation of the ship maintainability design knowledge base; Obtain text data input by the designer, use natural language processing technology to perform entity recognition on the text data, and obtain entities and their corresponding classes, specifically including: Get text data input by the designer; Input the text data into the trained NER model to obtain entities and their corresponding classes; The determination process of the trained NER model is as follows: Constructing a sample data set; each sample of the sample data set includes historical sample data and corresponding label data; the historical sample data is data obtained after removing noise, segmenting and marking parts of speech on historical text data, and the label data includes an entity label and its corresponding class label; Input each sample of the sample data set into the BERT model, fine-tune the parameters of the BERT model to obtain a fine-tuned model, and use the fine-tuned model as the trained NER model; Based on the ship maintainability design ontology and entities and their corresponding classes, the ship maintainability design strategy is obtained, which specifically includes: Based on the ship maintainability design requirements and DFMEA methodology, SWRL reasoning rules are created using SWRL language; the SWRL reasoning rules include functional rule class, structural rule class and reliability rule class; Among them, the functional rule category is based on the DFMEA methodology, which identifies functional failure modes that may cause maintenance difficulties or complexity by analyzing different system functions and design elements; the structural rule category is based on the DFMEA methodology, which considers the mutual influence and dependency between different components and structural elements, and infers different levels of failure impact, including local impact, higher-level impact and final impact; the reliability rule category uses the DFMEA methodology to identify design elements that may cause frequent equipment failures or require highly technical maintenance, so as to formulate corresponding maintainability improvement measures; The SWRL reasoning rules are imported into the ship maintainability design ontology by using the ontology editing tool Protégé, and the SWRL reasoning rules are processed by using the HermiT reasoning engine to update the ship maintainability design ontology to obtain an updated ship maintainability design ontology; Based on the updated ship maintainability design ontology and entities and their corresponding classes, a ship maintainability design strategy is obtained.

2. The ship maintainability design recommendation method according to claim 1, characterized in that: Acquiring ship knowledge data, processing the ship knowledge data, and constructing a ship maintainability design knowledge base, specifically including: Acquiring ship knowledge data; the formats of the ship knowledge data include structured data, semi-structured data and unstructured data; Performing data cleaning and format conversion operations on the ship knowledge data to obtain intermediate format data; Performing data mapping and conversion operations on the intermediate format data to obtain RDF triple data; the data mapping and conversion operations include entity extraction, attribute extraction and relationship extraction; The RDF triple data is integrated to obtain the ship maintainability design knowledge base; the integration operation includes ambiguity elimination and coreference resolution.

3. The ship maintainability design recommendation method according to claim 1, characterized in that: According to the ship maintainability design knowledge base, a ship maintainability design ontology is constructed, which specifically includes: Determine the domain and scope of ship maintainability design ontology; Based on the domain and scope of the ship maintainability design ontology, core concepts related to constructing the ship maintainability design ontology are extracted from the ship maintainability design knowledge base; According to the attributes and functions of the core concepts, define classes and attribute structures; the attribute structures include data attributes and object attributes; wherein the data attributes are attributes of the classes, and the object attributes are associations between classes; The classes, data attributes and object attributes are created by ontology editing tools, and the ship maintainability design knowledge base is described according to the classes, data attributes and object attributes by ontology language to obtain the ship maintainability design ontology.

4. The ship maintainability design recommendation method according to claim 3, characterized in that: The ontology editing tool is Protégé software; the ontology language is OWL language.

5. The ship maintainability design recommendation method according to claim 1, characterized in that: Based on the updated ship maintainability design ontology and entity and their corresponding classes, a ship maintainability design strategy is obtained, which specifically includes: According to the entities and their corresponding classes, the SPARQL language is used to query in the updated ship maintainability design ontology to obtain the ship maintainability design strategy.

6. A ship maintainability design recommendation system, characterized in that: Applying the ship maintainability design recommendation method as described in any one of claims 1 to 5, the ship maintainability design recommendation system comprises: a knowledge base module, a knowledge representation module and a maintainability design recommendation module; The knowledge base module is used to obtain ship knowledge data, process the ship knowledge data, and construct a ship maintainability design knowledge base; the ship knowledge data includes ship maintainability design knowledge and ship DFMEA knowledge; The knowledge representation module is used to construct a ship maintainability design ontology according to the ship maintainability design knowledge base; the ship maintainability design ontology is a structured representation of the ship maintainability design knowledge base; The maintainability design recommendation module is used to obtain text data input by the designer, perform entity recognition on the text data using natural language processing technology, obtain entities and their corresponding classes, and obtain a ship maintainability design strategy based on the ship maintainability design ontology and the entities and their corresponding classes; Also includes: knowledge reasoning module; The knowledge reasoning module is used to create SWRL reasoning rules based on ship maintainability design requirements and DFMEA methodology using SWRL language, import the SWRL reasoning rules into the ship maintainability design ontology using the ontology editing tool Protégé, and process the SWRL reasoning rules using the HermiT reasoning engine to update the ship maintainability design ontology to obtain an updated ship maintainability design ontology; the SWRL reasoning rules include functional rule classes, structural rule classes and reliability rule classes.

Images