Ship process rule modeling and updating method and device based on multi-source heterogeneous data, terminal equipment and storage medium

Abstract

The application discloses a ship process rule modeling and updating method and device based on multi-source heterogeneous data, terminal equipment and a storage medium, relates to the ship process technical field, and comprises the following steps: acquiring multi-source heterogeneous data, and constructing a process knowledge graph; acquiring a process update package, obtaining update entities and update relations through entity relation recognition; comparing the attribute features of the update entities with those of each process entity in the process knowledge graph; if there is a process entity with the same attribute features, the process entity is marked as a redundant entity; otherwise, the similarity of the update entities and each process entity is calculated, if the highest similarity is greater than a threshold, the update entity is marked as an associated entity; otherwise, the update entity is marked as a new entity; the process knowledge graph is updated according to the label of the update entity and the update relation; and the process rule and the process constraint are generated according to the updated process knowledge graph. Through implementation of the application, the problem of manually comparing and judging the association between new and old contents is solved, and the ship process updating efficiency is improved.

Description

Technical Field

[0001] This invention relates to the field of shipbuilding technology, and in particular to a method, apparatus, terminal equipment, and storage medium for modeling and updating shipbuilding process rules based on multi-source heterogeneous data. Background Technology

[0002] With the rapid development of intelligent shipbuilding technology, new shipbuilding materials, advanced construction techniques, and intelligent production equipment are constantly being applied, making traditional, rigid shipbuilding process rules increasingly difficult to adapt to the needs of modern shipbuilding. Outdated process rules deviate from actual on-site construction conditions, and some process parameters and construction constraints lag behind the latest industry standards and production levels. If timely updates and optimizations are not completed, it can easily lead to a disconnect between process guidance and actual production.

[0003] However, the updating of process rules in the shipbuilding industry still relies on manual operation. Process engineers rely on manual experience to compare and judge the relationship between the updated content and the existing content. The whole process has high human input, low update efficiency, and is difficult to adapt to the production needs of ship process iteration. Summary of the Invention

[0004] The present invention provides a method, apparatus, terminal equipment and storage medium for modeling and updating ship process rules based on multi-source heterogeneous data. It can solve the problem of low efficiency in ship process updating caused by relying on manual experience to compare and judge the relationship between the updated content and the existing content in the prior art, and improve the efficiency of ship process updating.

[0005] An embodiment of the present invention provides a method for modeling and updating ship process rules based on multi-source heterogeneous data, including: Acquire multi-source heterogeneous data on shipbuilding processes and process update packages to be processed; Based on the multi-source heterogeneous data, a process knowledge graph is constructed; Entity identification and relationship extraction are performed on the process update package to obtain several update entities and update relationships between them. For each updated entity, the updated entity is compared with the attributes and features of each process entity in the process knowledge graph; if there is a process entity in the process knowledge graph with the same attributes and features as the updated entity, the updated entity is treated as a redundant entity of the corresponding process entity. If no process entity in the process knowledge graph has the same attribute features as the updated entity, calculate the semantic similarity between the updated entity and each process entity in the process knowledge graph. If the highest semantic similarity is greater than the preset similarity threshold, the updated entity is regarded as the associated entity of the process entity corresponding to the highest semantic similarity. If the highest semantic similarity is less than or equal to the preset similarity threshold, the updated entity is regarded as a new entity. The process knowledge graph is updated based on the mapping labels of the updated entities and the update relationships between the updated entities to obtain the updated process knowledge graph; the mapping labels include: redundant entities, associated entities, or newly added entities. The entity attributes and relationships in the updated process knowledge graph are filled into the preset process rule template and preset process constraint template to generate new ship process rules and new process constraints.

[0006] Furthermore, based on the aforementioned multi-source heterogeneous data, a process knowledge graph is constructed, including: Multi-source heterogeneous data is input into a trained entity recognition model so that the entity recognition model can perform entity recognition on the multi-source heterogeneous data to obtain the process entity; Multi-source heterogeneous data and process entities are input into a trained relation extraction model so that the relation extraction model can identify the relationships between process entities based on the multi-source heterogeneous data. A process knowledge graph is constructed using process entities as nodes and process relationships as edges.

[0007] Furthermore, after acquiring multi-source heterogeneous data on shipbuilding processes, the following also includes: The Manhattan distance algorithm is used to remove outliers from multi-source heterogeneous data, resulting in preprocessed multi-source heterogeneous data.

[0008] Furthermore, based on the mapping labels of the updated entities and the update relationships between them, the process knowledge graph is updated to obtain the updated process knowledge graph, including: For each updated entity, perform entity node matching to obtain the process knowledge graph after entity alignment; Based on the update relationships between updated entities, the semantic relationships in the aligned process knowledge graph are added or modified to obtain the updated process knowledge graph. The entity node matching includes: When the mapping label of the updated entity is a redundant entity or a related entity, the attribute information of the corresponding process entity in the process knowledge graph is updated according to the attribute information of the updated entity. When the mapping label of the updated entity is a new entity, a new process entity is created in the process knowledge graph, and the attribute information of the updated entity is used as the attribute information of the new process entity.

[0009] Furthermore, after obtaining the updated process knowledge graph, it also includes: Traverse the updated process knowledge graph and check if there are any preset conflict paths in the updated process knowledge graph. If so, mark the updated process knowledge graph as failing the verification and record the existing preset conflict paths and their corresponding conflict types as the verification result. Otherwise, mark the updated process knowledge graph as passing the verification and use it as the verification result.

[0010] Furthermore, after obtaining the verification result, it also includes: Obtain the update time of the process knowledge graph and the source information of the process update package; A change log is generated based on the process update package, verification results, update time, and source information.

[0011] Furthermore, after generating new shipbuilding process rules and new process constraints, it also includes: In accordance with the new shipbuilding process rules and new process constraints, the equipment in the shipbuilding process is controlled.

[0012] Based on the above method embodiments, the present invention provides corresponding apparatus embodiments, including: a data acquisition module, a graph construction module, an entity relationship extraction module, an attribute feature comparison module, a similarity comparison module, a graph update module, and a process rule generation module; The data acquisition module is used to acquire multi-source heterogeneous data of ship technology and process update packages to be processed; The graph construction module is used to construct a process knowledge graph based on the multi-source heterogeneous data; The entity relationship extraction module is used to perform entity identification and relationship extraction on the process update package to obtain several update entities and update relationships between update entities. The attribute feature comparison module is used to compare the attribute features of each updated entity with those of each process entity in the process knowledge graph. If there are process entities in the process knowledge graph that have the same attribute features as the updated entity, the updated entity is treated as a redundant entity of the corresponding process entity. The similarity comparison module is used to calculate the semantic similarity between the updated entity and each process entity in the process knowledge graph if there are no process entities in the process knowledge graph that have the same attribute features as the updated entity. If the highest semantic similarity is greater than a preset similarity threshold, the updated entity is regarded as the associated entity of the process entity corresponding to the highest semantic similarity. If the highest semantic similarity is less than or equal to the preset similarity threshold, the updated entity is regarded as a new entity. The graph update module is used to update the process knowledge graph according to the mapping labels of the updated entities and the update relationships between the updated entities, so as to obtain the updated process knowledge graph; the mapping labels include: redundant entities, associated entities, or newly added entities. The process rule generation module is used to fill the entity attributes and relationships in the updated process knowledge graph into the preset process rule template and preset process constraint template to generate new ship process rules and new process constraint conditions.

[0013] Based on the above method embodiments, the present invention provides a corresponding terminal device embodiment, including: a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor. When the processor executes the computer program, it implements the steps of the ship process rule modeling and updating method based on multi-source heterogeneous data as described in the present invention.

[0014] Based on the above method embodiments, the present invention provides a corresponding computer-readable storage medium embodiment, including: a stored computer program, which, when the computer program is running, controls the device where the computer-readable storage medium is located to execute the steps of the ship process rule modeling and updating method based on multi-source heterogeneous data as described in the present invention.

[0015] Compared with the prior art, the beneficial effects of this embodiment are as follows: This invention acquires multi-source heterogeneous data on shipbuilding processes and process update packages to be processed; constructs a process knowledge graph based on the multi-source heterogeneous data; performs entity recognition and relationship extraction on the process update packages, thereby transforming unstructured process rule update text into structured update entities and update relationships between update entities; for each update entity, compares the attribute features of the update entity with those of each process entity in the process knowledge graph; if there are process entities in the process knowledge graph with the same attribute features as the update entity, the update entity is treated as a redundant entity of the corresponding process entity, thereby identifying update information that is duplicated with existing knowledge and preventing the graph from repeatedly creating the same process entity; if there are no process entities in the process knowledge graph with the same attribute features as the update entity, calculates the relationship between the update entity and the process knowledge graph... The semantic similarity of each process entity in the process knowledge graph is evaluated. If the highest semantic similarity is greater than a preset similarity threshold, the updated entity is treated as the associated entity of the process entity with the highest semantic similarity, enabling automatic association mapping of semantically similar process knowledge. If the highest semantic similarity is less than or equal to the preset similarity threshold, the updated entity is treated as a new entity, identifying new materials and processes. The process knowledge graph is updated based on the mapping labels of the updated entities and the update relationships between them. The mapping labels include redundant entities, associated entities, or new entities. The entity attributes and relationships in the updated process knowledge graph are then filled into preset process rule templates and preset process constraint templates to generate new shipbuilding process rules and new process constraint conditions.

[0016] In summary, this invention uses a two-level automated screening process—attribute feature comparison and semantic similarity calculation—to automatically determine the association between updated entities and map process entities, distinguishing between redundant entities, associated entities, and newly added entities. This solves the problem of low efficiency in ship process updates caused by relying on manual experience to compare and determine the association between updated and existing content in existing technologies, thereby improving the efficiency of ship process updates. Attached Figure Description

[0017] To more clearly illustrate the technical solution of this application, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0018] Figure 1 This is a flowchart illustrating a method for modeling and updating ship process rules based on multi-source heterogeneous data, provided in an embodiment of the present invention. Figure 2 This is a schematic diagram of the structure of a ship process rule modeling and updating device based on multi-source heterogeneous data provided in an embodiment of the present invention. Detailed Implementation

[0019] To make the objectives, technical solutions, and advantages of this application clearer, the technical solutions of this application will be clearly and completely described below with reference to the accompanying drawings of the embodiments. Obviously, the described embodiments are only some embodiments of this application, not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0020] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application pertains; the terminology used herein is for the purpose of describing particular embodiments only and is not intended to limit the application; the terms “comprising” and “having”, and any variations thereof, in the specification, claims, and foregoing description of the drawings are intended to cover non-exclusive inclusion.

[0021] In the description of the embodiments of this application, technical terms such as "first" and "second" are used only to distinguish different objects and should not be construed as indicating or implying relative importance or implicitly specifying the number, specific order, or primary and secondary relationship of the indicated technical features. In the description of the embodiments of this application, "multiple" means two or more, unless otherwise explicitly defined.

[0022] In this document, the term "embodiment" means that a particular feature, structure, or characteristic described in connection with an embodiment may be included in at least one embodiment of this application. The appearance of this phrase in various places throughout the specification does not necessarily refer to the same embodiment, nor is it a separate or alternative embodiment mutually exclusive with other embodiments. It will be explicitly and implicitly understood by those skilled in the art that the embodiments described herein can be combined with other embodiments.

[0023] In the description of the embodiments in this application, the term "and / or" is merely a description of the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A existing alone, A and B existing simultaneously, and B existing alone. Additionally, the character " / " in this document generally indicates that the preceding and following related objects have an "or" relationship.

[0024] In the description of the embodiments of this application, the term "multiple" refers to two or more (including two), similarly, "multiple sets" refers to two or more (including two sets), and "multiple pieces" refers to two or more (including two pieces).

[0025] In the description of the embodiments of this application, unless otherwise expressly specified and limited, technical terms such as "installation," "connection," "joining," and "fixing" should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral part; they can refer to a mechanical connection or an electrical connection; they can refer to a direct connection or an indirect connection through an intermediate medium; they can refer to the internal communication of two components or the interaction between two components. For those skilled in the art, the specific meaning of the above terms in the embodiments of this application can be understood according to the specific circumstances.

[0026] like Figure 1 As shown, in order to solve the problem of low efficiency in ship process updating caused by relying on manual experience to compare and judge the correlation between updated content and existing content in the prior art, an embodiment of the present invention provides a method for ship process rule modeling and updating based on multi-source heterogeneous data. The method includes at least the following steps: Step S1: Obtain multi-source heterogeneous data on shipbuilding processes and process update packages to be processed; For step S1, multi-source heterogeneous data of shipbuilding processes are collected by connecting with multiple independent business systems such as ship design system, production process system, and inspection document system. The multi-source heterogeneous data refers to data from different business systems. The data models, syntax, and semantics of various types of data differ to varying degrees. The data forms include structured process parameter tables, unstructured process drawings, process technical documents, inspection record texts, etc. Specifically, it covers parameters such as inner diameter, outer diameter, length, material, wall thickness, classification symbol, bending radius, and bending length of steel pipes, thickness, length, width, material, and classification symbol parameters of steel plates, as well as process information related to processing procedures and assembly constraints of various components. Preferably, after acquiring multi-source heterogeneous data on shipbuilding processes, the process further includes: The Manhattan distance algorithm is used to remove outliers from multi-source heterogeneous data, resulting in preprocessed multi-source heterogeneous data.

[0027] Specifically, to improve data quality, a set of historical data that has been confirmed to be valid within the previous statistical period (one day or one week) for the same data item is selected. The mean of the historical data set is used as the center point. The Manhattan distance between each historical data in the historical data set and the center point is calculated in turn. The average of the Manhattan distances corresponding to all historical data is then calculated to obtain the average distance.

[0028] For each data value in the multi-source heterogeneous data, the distance is calculated to obtain the corresponding distance value. The distance value is compared with the distance mean. If the distance value is greater than the distance mean, the data is deemed unreasonable, marked as erroneous data, and removed. If the distance value is less than or equal to the distance mean, the data is deemed reasonable and retained.

[0029] Through the above preprocessing operations, data that deviates too much from the normal average fluctuation level is filtered out and removed, resulting in a preprocessed multi-source heterogeneous data source.

[0030] Obtain the process update package to be processed, wherein the process update package contains unstructured text recording the process update content.

[0031] Step S2: Construct a process knowledge graph based on the multi-source heterogeneous data; In a preferred embodiment, a process knowledge graph is constructed based on the multi-source heterogeneous data, including: Multi-source heterogeneous data is input into a trained entity recognition model so that the entity recognition model can perform entity recognition on the multi-source heterogeneous data to obtain the process entity; Multi-source heterogeneous data and process entities are input into a trained relation extraction model so that the relation extraction model can identify the relationships between process entities based on the multi-source heterogeneous data. A process knowledge graph is constructed using process entities as nodes and process relationships as edges.

[0032] For step S2, the preprocessed multi-source heterogeneous data is input into the trained entity recognition model. The entity recognition model uses the feature extraction weights obtained from the previous ship process annotation sample training to perform semantic parsing on the multi-source heterogeneous data and automatically identify various process entities contained in the text, such as process parameter entities, constraint entities, process node entities, etc.

[0033] It should be noted that the entity recognition model uses the BERT model as its base model and is fine-tuned using annotated samples of shipbuilding process entities. Specifically, during the training phase, historical process documents, process specifications, and work instructions in the shipbuilding process field are first collected as textual data. The process entities and their attribute features contained within are manually annotated to obtain shipbuilding process entity annotation samples. Then, these annotated samples are input into the BERT model for fine-tuning and training. This allows the BERT model to learn the feature extraction weights for entity boundary recognition and attribute feature extraction in shipbuilding process texts, thereby enabling it to automatically recognize and extract entity information from shipbuilding process texts.

[0034] The preprocessed multi-source heterogeneous data and the aforementioned process entities are input into the trained relation extraction model. The relation extraction model uses the classification weights obtained from the previous training of ship process relation annotation samples to parse the semantic associations between entities and automatically identify the types of semantic relationships between each process entity. For example, the limiting relationship and adaptation relationship between process parameter entities and constraint entities, the attribution and pre-constraint relationship between process parameter entities and process node entities, the rule attachment relationship, admission verification relationship and state triggering relationship between constraint entities and process node entities, etc.

[0035] It should be noted that the relation extraction model uses the BERT model as its base model and is fine-tuned using ship process relation annotation samples. Specifically, during the training phase, historical process documents, process specifications, and work instructions in the ship process field are first collected. Process entities identified by the entity recognition model are used as annotation objects. For each pair of semantically related process entities in the text, the relation type is manually annotated. The annotated relation types include at least the limiting and adaptation relationships between process parameter entities and constraint entities, the attribution and pre-constraint relationships between process parameter entities and process node entities, and the rule-attached relationships, access verification relationships, and state triggering relationships between constraint entities and process node entities, forming ship process relation annotation samples. Then, these ship process relation annotation samples are input into the BERT model for fine-tuning training. This allows the BERT model to learn the feature extraction weights for contextual semantic relationships and relation category discrimination between entities in ship process texts, thereby enabling it to automatically identify and classify semantic relationships between entities in multi-source heterogeneous ship process texts.

[0036] Using the process entities identified by the entity recognition model as nodes, the attribute features corresponding to each process entity as the attribute information of the nodes, and the semantic relationships between the process entities identified by the relation extraction model as edges, an initial process knowledge graph is constructed.

[0037] The structured information obtained from entity recognition and relation extraction is integrated into a graph structure representation using the above method.

[0038] Step S3: Perform entity identification and relationship extraction on the process update package to obtain several update entities and update relationships between update entities; For step S3, the process update package is input into the trained entity recognition model. The entity recognition model performs semantic parsing on the text data in the process update package and automatically identifies the various update entities contained therein and their corresponding attribute features. The process update package and the aforementioned update entities are input into the trained relation extraction model. The relation extraction model parses the semantic associations between each update entity and automatically identifies the update relationships between each update entity.

[0039] Through the entity recognition and relationship extraction described above, the unstructured update text in the process update package is transformed into structured update information, namely, several update entities and the update relationships between each update entity.

[0040] Step S4: For each updated entity, compare the attribute features of the updated entity with those of each process entity in the process knowledge graph; if there are process entities in the process knowledge graph that have the same attribute features as the updated entity, treat the updated entity as a redundant entity of the corresponding process entity. For step S4, the attribute features of the updated entity are compared with the attribute features of each process entity in the process knowledge graph. If there is an existing process entity in the process knowledge graph with the same attribute features as the updated entity, it is determined that the knowledge information carried by the updated entity already exists completely in the current process knowledge graph, and the updated entity is marked as a redundant entity of the existing process entity.

[0041] By comparing the aforementioned attribute features, a rapid coarse screening of updated entities is achieved. This can accurately identify updated information that is completely duplicated with existing knowledge in the graph and mark it as redundant entities, thus avoiding the introduction of duplicate data. At the same time, it reduces the amount of data that needs to be processed for subsequent semantic similarity calculations and improves the overall update efficiency.

[0042] Step S5: If there are no process entities in the process knowledge graph that have the same attribute features as the updated entity, calculate the semantic similarity between the updated entity and each process entity in the process knowledge graph. If the highest semantic similarity is greater than the preset similarity threshold, the updated entity is regarded as the associated entity of the process entity corresponding to the highest semantic similarity. If the highest semantic similarity is less than or equal to the preset similarity threshold, the updated entity is regarded as a new entity. For step S5, if there is no existing process entity in the process knowledge graph that has exactly the same attribute features as the updated entity, it means that the updated entity is not completely duplicated with existing knowledge, and it is necessary to further determine its relationship with existing knowledge through semantic-level discrimination. At this time, the semantic similarity between the updated entity and each existing process entity in the process knowledge graph is calculated.

[0043] Specifically, the semantic similarity calculation employs a semantic representation model, mapping the updated entity and each existing process entity to a vector space to obtain the semantic vector corresponding to each entity. Then, by calculating the cosine similarity between the semantic vector of the updated entity and the semantic vector of each existing process entity, the semantic similarity value between the updated entity and each existing process entity is obtained. The semantic representation model can employ pre-trained language models such as BERT, where the entity's name, category, and attribute features are concatenated and input into the model, and the pooling layer output is extracted as the entity's semantic vector representation. The cosine similarity value ranges from 0 to 1; a higher value indicates a greater degree of semantic similarity between the two entities.

[0044] After obtaining all semantic similarity values ​​between the updated entity and each existing process entity, the maximum value is selected as the highest semantic similarity, and the existing process entity corresponding to the highest semantic similarity is selected as the candidate associated object.

[0045] The highest semantic similarity is compared with a preset similarity threshold. The preset similarity threshold is pre-set according to the actual application scenario (e.g., set to 0.85) and is used to determine whether two entities have sufficient semantic similarity to form an association relationship.

[0046] If the highest semantic similarity is greater than the preset similarity threshold, it indicates that the updated entity and the candidate associated object are highly similar at the semantic level. Although they are not completely identical in terms of attribute features (such as coming from different standard sources, using different naming methods, or different attribute description dimensions), they essentially point to the same process object or have a strong semantic relationship. In this case, the updated entity is marked as an associated entity of the candidate associated object.

[0047] If the highest semantic similarity is less than or equal to the preset similarity threshold, it indicates that the updated entity is not semantically similar to any existing process entity in the process knowledge graph. The knowledge information carried by the updated entity does not exist in the graph in exactly the same form (attribute feature comparison fails), nor does it have a strong semantic association with any existing knowledge. In this case, the updated entity is marked as a new entity.

[0048] By calculating semantic similarity and comparing thresholds as described above, we can achieve precise discrimination of updated entities that do not match attribute feature comparisons. This ensures that when faced with multi-source, heterogeneous, and inconsistently expressed ship technology data, the knowledge graph can distinguish between differences in textual expression and differences in entities, thereby improving the accuracy of dynamic updates of the technology knowledge graph.

[0049] Step S6: Update the process knowledge graph according to the mapping labels of the updated entities and the update relationships between the updated entities to obtain the updated process knowledge graph; the mapping labels include: redundant entities, associated entities, or newly added entities; In a preferred embodiment, the process knowledge graph is updated based on the mapping labels of the updated entities and the update relationships between the updated entities to obtain an updated process knowledge graph, including: For each updated entity, perform entity node matching to obtain the process knowledge graph after entity alignment; Based on the update relationships between updated entities, the semantic relationships in the aligned process knowledge graph are added or modified to obtain the updated process knowledge graph. The entity node matching includes: When the mapping label of the updated entity is a redundant entity or a related entity, the attribute information of the corresponding process entity in the process knowledge graph is updated according to the attribute information of the updated entity. When the mapping label of the updated entity is a new entity, a new process entity is created in the process knowledge graph, and the attribute information of the updated entity is used as the attribute information of the new process entity.

[0050] For step S6, the process knowledge graph is updated based on the mapping labels of the updated entities and the update relationships between the updated entities. Specifically, the above update process includes an entity node matching stage and a semantic relationship update stage.

[0051] During the entity node matching stage, for each updated entity, the corresponding process entity in the process knowledge graph is determined based on the mapping label of the updated entity, and the attribute information of the updated entity is merged into the corresponding process entity to obtain the entity-aligned process knowledge graph.

[0052] If the mapping label of the updated entity is a redundant entity, it means that the knowledge information carried by the updated entity already exists completely in the graph. In this case, the graph structure remains unchanged, and only the update time and source information of the updated entity are appended to the corresponding existing process entity as a traceability record. If the mapping label of the updated entity is a related entity, it means that the updated entity is highly related to the existing process entity at the semantic level and carries additional attribute incremental information. In this case, no new node is created, but the attribute information of the updated entity is merged into the corresponding existing process entity, and the local attributes of the existing process entity are incrementally updated without changing the other structure and relationships of the node in the graph. If the mapping label of the updated entity is a new entity, it means that the knowledge information carried by the updated entity does not exist in the existing graph at all. In this case, a new process entity node is created in the process knowledge graph, and the attribute information of the updated entity is written into the node as the attribute information of the new process entity.

[0053] During the semantic relationship update phase, for each update relationship, the starting and ending update entities involved in the update relationship are first determined to correspond to the respective process entities in the process knowledge graph after entity alignment. If an update entity is located to an existing process entity during the entity node matching phase (i.e., the mapping label is a redundant entity or an associated entity), then the existing process entity is used as the starting or ending point of the update relationship in the graph. If an update entity is created as a new process entity during the entity node matching phase (i.e., the mapping label is a newly added entity), then the new process entity is used as the starting or ending point of the update relationship in the graph.

[0054] Then, a new relation edge is created between the process entity corresponding to the starting end and the process entity corresponding to the ending end. The type of this relation edge is consistent with the relation type of the update relation. If there is already a relation edge in the graph that is the same as the update relation type and connects the same two process entities, the existing relation edge is marked as a historical version, and a new relation edge is created to replace it. In this way, the update relations between each update entity are completely migrated to the process knowledge graph after the entity alignment, and the updated process knowledge graph is obtained.

[0055] It should be noted that after obtaining the updated process knowledge graph, the updated process knowledge graph will be used as the process knowledge graph for the next update of the ship's process rules. That is, after each update, the original process knowledge graph is replaced with the currently updated process knowledge graph for storage, so that the process knowledge graph continues to accumulate in each update iteration, realizing the dynamic optimization of the ship's process knowledge system.

[0056] In a preferred embodiment, after obtaining the updated process knowledge graph, the method further includes: Traverse the updated process knowledge graph and check if there are any preset conflict paths in the updated process knowledge graph. If so, mark the updated process knowledge graph as failing the verification and record the existing preset conflict paths and their corresponding conflict types as the verification result. Otherwise, mark the updated process knowledge graph as passing the verification and use it as the verification result.

[0057] In one embodiment of the present invention, after updating the process knowledge graph, before using the graph to generate ship process rules, it is necessary to first perform a logical consistency check on the updated process knowledge graph to ensure the logical self-consistency of the nodes and relationships within the graph.

[0058] Specifically, a set of preset conflict paths is obtained, which is used to characterize the combination of node types and relationship path patterns that should not exist simultaneously in the field of shipbuilding technology.

[0059] Starting from each existing node in the updated process knowledge graph, traverse the path along the relation edges, perform pattern matching between each traversed path and the preset conflict path, and check whether there is an actual path in the graph that matches any preset conflict path.

[0060] If a matching path exists, it indicates that there is a logical conflict or contradiction in the updated process knowledge graph. In this case, the updated process knowledge graph is marked as failing the verification, and the nodes, relation edges, and conflict types involved in the matched conflict path are recorded to form the verification result. At the same time, the subsequent process rule generation process is stopped, and the verification is carried out again after manual intervention or automatic system correction.

[0061] If no path matching the preset conflict path is found after the traversal is completed, it means that the updated process knowledge graph maintains consistency and self-consistency at the logical level. At this time, the updated process knowledge graph is marked as passed and, as the verification result, the subsequent process rule generation operation is allowed.

[0062] Through the above consistency verification, logical conflicts and structural errors that may exist in the graph can be automatically identified and eliminated before generating process rules, ensuring the reliability of the process knowledge graph itself used for rule generation, thereby guaranteeing the accuracy of the final output ship process rules.

[0063] In a preferred embodiment, after obtaining the verification result, the method further includes: Obtain the update time of the process knowledge graph and the source information of the process update package; A change log is generated based on the process update package, verification results, update time, and source information.

[0064] In one embodiment of the present invention, the update time of the current update of the process knowledge graph and the source information of the process update package are obtained. The update time is used to uniquely identify the time node of the current update operation, and the source information is used to trace the origin of the update data.

[0065] Then, the preset change log template is invoked, and the process update package involved in this update, the verification result, the update time, and the source information are filled in according to the field mapping relationship of the change log template to generate a structured change log record. This realizes the auditable management of the process knowledge graph update process. When rule anomalies are found later or historical states need to be traced back, the source of the problem and the reason for the change can be quickly located by consulting the change log.

[0066] Step S7: Fill the entity attributes and relationships in the updated process knowledge graph into the preset process rule template and preset process constraint template to generate new ship process rules and new process constraints.

[0067] For step S7, the preset process rule template and the preset process constraint template are invoked. The preset process rule template defines the standard expression format and field structure of ship process rules, and the preset process constraint template defines the standard expression format and field structure of process constraint conditions.

[0068] The system extracts attribute information of each process entity and semantic relationship information between process entities from the updated process knowledge graph. According to the preset field mapping relationship, the extracted entity attributes and relationship information are filled into the corresponding field positions in the preset process rule template and preset process constraint template to generate standard ship process rules and process constraints that conform to ship process specifications.

[0069] In this embodiment, the ship process rules include process sequencing rules, process path rules, process station selection rules, and inter-process collaborative optimization rules; the process constraints include operation time process constraints, inter-process transmission time process constraints, and scheduling constraints under abnormal operating conditions.

[0070] For example, for the process parameter entity "welding current" and its associated constraint entity "upper limit of welding equipment capacity", a process constraint rule "the welding current shall not exceed the upper limit of the rated capacity of the equipment" is automatically generated based on the "limiting relationship" between the two.

[0071] In a preferred embodiment, after generating the new ship process rules and new process constraints, the method further includes: In accordance with the new shipbuilding process rules and new process constraints, the equipment in the shipbuilding process is controlled.

[0072] In one embodiment of the present invention, during the application of CNC machining equipment (such as CNC cutting machines, CNC machine tools, intelligent welding machines, etc.), the process sequencing rules and process path rules are transformed into the processing operation execution sequence and tool movement trajectory of the equipment, enabling the equipment to automatically execute processing operations according to the prescribed process sequence and processing path; the process station selection rules are transformed into station allocation instructions for the equipment, enabling the equipment to automatically match the corresponding processing station on the production line, avoiding production conflicts caused by station mismatch; and the operation time process constraints and inter-process transmission time process constraints are transformed into cycle time control parameters for the equipment. The system enables the equipment to adaptively adjust its production rhythm according to the specified operating time and waiting time between processes, ensuring a balanced production line pace. It transforms the inter-process collaborative optimization rules into synchronous control signals for multi-equipment collaborative operation, enabling multiple devices to complete material handover and process flow under the same production rhythm, achieving multi-machine collaborative operation. Furthermore, it transforms the scheduling constraints under abnormal operating conditions into equipment fault response logic and emergency shutdown strategies, allowing the equipment to automatically trigger preset emergency scheduling schemes when it detects excessive process parameters, equipment overload, or abnormal materials, performing operations such as shutdown protection, task switching, or reduced-speed operation.

[0073] By directly applying the updated process rules and constraints to the control systems of various production equipment, a seamless connection between rule updates and equipment execution is achieved, effectively improving the automation level and production consistency of the shipbuilding process, and avoiding equipment execution deviations and quality defects caused by manual parameter setting errors or rule misunderstandings.

[0074] like Figure 2 As shown, based on the above method embodiments, corresponding apparatus embodiments are provided; An embodiment of the present invention provides a ship process rule modeling and updating device based on multi-source heterogeneous data, including: a data acquisition module 101, a graph construction module 102, an entity relationship extraction module 103, an attribute feature comparison module 104, a similarity comparison module 105, a graph updating module 106, and a process rule generation module 107. The data acquisition module 101 is used to acquire multi-source heterogeneous data of ship technology and process update packages to be processed. The graph construction module 102 is used to construct a process knowledge graph based on the multi-source heterogeneous data; The entity relationship extraction module 103 is used to perform entity identification and relationship extraction on the process update package to obtain a number of update entities and update relationships between update entities. The attribute feature comparison module 104 is used to compare the attribute features of each updated entity with those of each process entity in the process knowledge graph for each updated entity; if there is a process entity in the process knowledge graph with the same attribute features as the updated entity, the updated entity is regarded as a redundant entity of the corresponding process entity. The similarity comparison module 105 is used to calculate the semantic similarity between the updated entity and each process entity in the process knowledge graph if there are no process entities in the process knowledge graph that have the same attribute features as the updated entity. If the highest semantic similarity is greater than a preset similarity threshold, the updated entity is regarded as the associated entity of the process entity corresponding to the highest semantic similarity. If the highest semantic similarity is less than or equal to the preset similarity threshold, the updated entity is regarded as a new entity. The graph update module 106 is used to update the process knowledge graph according to the mapping labels of the updated entities and the update relationship between the updated entities, so as to obtain the updated process knowledge graph; the mapping labels include: redundant entities, associated entities or newly added entities. The process rule generation module 107 is used to fill the entity attributes and relationships in the updated process knowledge graph into the preset process rule template and the preset process constraint template to generate new ship process rules and new process constraint conditions.

[0075] It is understood that the above-described device embodiments correspond to the method embodiments of the present invention, and can implement the ship process rule modeling and updating method based on multi-source heterogeneous data provided by any of the above-described method embodiments of the present invention.

[0076] It should be noted that the device embodiments described above are merely illustrative, and some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs. Furthermore, in the accompanying drawings of the device embodiments provided by this invention, the connection relationships between modules indicate that they have communication connections, which can specifically be implemented as one or more communication buses or signal lines. Those skilled in the art can understand and implement this without any creative effort.

[0077] Based on the above embodiments of the ship process rule modeling and updating method based on multi-source heterogeneous data, another embodiment of the present invention provides a terminal device, which includes a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor. When the processor executes the computer program, it implements the ship process rule modeling and updating method based on multi-source heterogeneous data according to any embodiment of the present invention.

[0078] For example, in this embodiment, the computer program can be divided into one or more modules, which are stored in the memory and executed by the processor to complete the present invention. The one or more modules may be a series of computer program instruction segments capable of performing a specific function, which describe the execution process of the computer program in the terminal device.

[0079] The terminal device may be a desktop computer, laptop, handheld computer, or cloud server, etc. The terminal device may include, but is not limited to, a processor and a memory.

[0080] The processor can be a Central Processing Unit (CPU), or other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. A general-purpose processor can be a microprocessor or any conventional processor. The processor is the control center of the terminal device, connecting all parts of the terminal device via various interfaces and lines.

[0081] Based on the above-described method embodiments, another embodiment is provided: another embodiment of the present invention provides a computer-readable storage medium, including a stored computer program, wherein, when the computer program is running, it controls the device where the computer-readable storage medium is located to execute the ship process rule modeling and updating method based on multi-source heterogeneous data as described in any of the above-described method embodiments of the present invention.

[0082] The module / unit integrated into the ship process rule modeling and updating device / terminal equipment based on multi-source heterogeneous data, if implemented as a software functional unit and sold or used as an independent product, can be stored in a computer-readable storage medium. Based on this understanding, all or part of the processes in the above embodiments of the present invention can also be implemented by a computer program instructing related hardware. The computer program can be stored in a computer-readable storage medium, and when executed by a processor, it can implement the steps of the various method embodiments described above. The computer program includes computer program code, which can be in the form of source code, object code, executable files, or certain intermediate forms. The computer-readable medium can include: any entity or device capable of carrying the computer program code, recording media, USB flash drives, portable hard drives, magnetic disks, optical disks, computer memory, read-only memory (ROM), random access memory (RAM), electrical carrier signals, telecommunication signals, and software distribution media, etc.

[0083] The above description represents the preferred embodiments of the present invention. It should be noted that those skilled in the art can make various improvements and modifications without departing from the principles of the present invention, and these improvements and modifications are also considered to be within the scope of protection of the present invention.

Claims

1. A method for modeling and updating ship process rules based on multi-source heterogeneous data, characterized in that, include: Acquire multi-source heterogeneous data on shipbuilding processes and process update packages to be processed; Based on the multi-source heterogeneous data, a process knowledge graph is constructed; Entity identification and relationship extraction are performed on the process update package to obtain several update entities and update relationships between them. For each updated entity, the attribute features of the updated entity are compared with those of each process entity in the process knowledge graph. If there are process entities and updated entities with the same attribute characteristics in the process knowledge graph, the updated entity will be treated as a redundant entity of the corresponding process entity. If no process entity in the process knowledge graph has the same attribute features as the updated entity, calculate the semantic similarity between the updated entity and each process entity in the process knowledge graph. If the highest semantic similarity is greater than the preset similarity threshold, the updated entity is regarded as the associated entity of the process entity corresponding to the highest semantic similarity. If the highest semantic similarity is less than or equal to the preset similarity threshold, the updated entity is regarded as a new entity. The process knowledge graph is updated based on the mapping labels of the updated entities and the update relationships between the updated entities to obtain the updated process knowledge graph. The mapping labels include: redundant entities, associated entities, or newly added entities; The entity attributes and relationships in the updated process knowledge graph are filled into the preset process rule template and preset process constraint template to generate new ship process rules and new process constraints.

2. The method for modeling and updating ship process rules based on multi-source heterogeneous data according to claim 1, characterized in that, Based on the aforementioned multi-source heterogeneous data, a process knowledge graph is constructed, including: Multi-source heterogeneous data is input into a trained entity recognition model so that the entity recognition model can perform entity recognition on the multi-source heterogeneous data to obtain the process entity; Multi-source heterogeneous data and process entities are input into a trained relation extraction model so that the relation extraction model can identify the relationships between process entities based on the multi-source heterogeneous data. A process knowledge graph is constructed using process entities as nodes and process relationships as edges.

3. The method for modeling and updating ship process rules based on multi-source heterogeneous data according to claim 1, characterized in that, After acquiring multi-source heterogeneous data on shipbuilding processes, the following is also included: The Manhattan distance algorithm is used to remove outliers from multi-source heterogeneous data, resulting in preprocessed multi-source heterogeneous data.

4. The method for modeling and updating ship process rules based on multi-source heterogeneous data according to claim 1, characterized in that, Based on the mapping labels of the updated entities and the update relationships between them, the process knowledge graph is updated to obtain the updated process knowledge graph, including: For each updated entity, perform entity node matching to obtain the process knowledge graph after entity alignment; Based on the update relationships between updated entities, the semantic relationships in the aligned process knowledge graph are added or modified to obtain the updated process knowledge graph. The entity node matching includes: When the mapping label of the updated entity is a redundant entity or a related entity, the attribute information of the corresponding process entity in the process knowledge graph is updated according to the attribute information of the updated entity. When the mapping label of the updated entity is a new entity, a new process entity is created in the process knowledge graph, and the attribute information of the updated entity is used as the attribute information of the new process entity.

5. The method for modeling and updating ship process rules based on multi-source heterogeneous data according to claim 1, characterized in that, After obtaining the updated process knowledge graph, it also includes: Traverse the updated process knowledge graph and check if there are any preset conflict paths in the updated process knowledge graph. If so, mark the updated process knowledge graph as failing the verification and record the existing preset conflict paths and their corresponding conflict types as the verification result. Otherwise, mark the updated process knowledge graph as passing the verification and use it as the verification result.

6. The method for updating ship process rules according to claim 5, characterized in that, After obtaining the verification result, it also includes: Obtain the update time of the process knowledge graph and the source information of the process update package; A change log is generated based on the process update package, verification results, update time, and source information.

7. The method for modeling and updating ship process rules based on multi-source heterogeneous data according to claim 1, characterized in that, After generating new shipbuilding process rules and new process constraints, the following also includes: In accordance with the new shipbuilding process rules and new process constraints, the equipment in the shipbuilding process is controlled.

8. A device for modeling and updating ship process rules based on multi-source heterogeneous data, characterized in that, include: The system includes a data acquisition module, a graph construction module, an entity relationship extraction module, an attribute feature comparison module, a similarity comparison module, a graph update module, and a process rule generation module. The data acquisition module is used to acquire multi-source heterogeneous data of ship technology and process update packages to be processed; The graph construction module is used to construct a process knowledge graph based on the multi-source heterogeneous data; The entity relationship extraction module is used to perform entity identification and relationship extraction on the process update package to obtain several update entities and update relationships between update entities. The attribute feature comparison module is used to compare the attribute features of each updated entity with those of each process entity in the process knowledge graph for each updated entity. If there are process entities and updated entities with the same attribute characteristics in the process knowledge graph, the updated entity will be treated as a redundant entity of the corresponding process entity. The similarity comparison module is used to calculate the semantic similarity between the updated entity and each process entity in the process knowledge graph if there are no process entities in the process knowledge graph that have the same attribute features as the updated entity. If the highest semantic similarity is greater than a preset similarity threshold, the updated entity is regarded as the associated entity of the process entity corresponding to the highest semantic similarity. If the highest semantic similarity is less than or equal to the preset similarity threshold, the updated entity is regarded as a new entity. The graph update module is used to update the process knowledge graph according to the mapping labels of the updated entities and the update relationship between the updated entities, so as to obtain the updated process knowledge graph. The mapping labels include: redundant entities, associated entities, or newly added entities; The process rule generation module is used to fill the entity attributes and relationships in the updated process knowledge graph into the preset process rule template and preset process constraint template to generate new ship process rules and new process constraint conditions.

9. A terminal device, characterized in that, The system includes a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor, wherein the processor executes the computer program to implement the ship process rule modeling and updating method based on multi-source heterogeneous data as described in any one of claims 1-7.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium includes a stored computer program, wherein, when the computer program is executed, it controls the device where the computer-readable storage medium is located to perform the ship process rule modeling and updating method based on multi-source heterogeneous data as described in any one of claims 1-7.

Images