Ship pipeline assembly auxiliary method based on AI and knowledge graph

Abstract

The invention relates to the technical field of ship pipeline assembly, in particular to a ship pipeline assembly auxiliary method based on AI and a knowledge graph, and a ship pipeline assembly scheme can be given more quickly and accurately through construction of a field large model and a knowledge graph network. And constructors conduct construction according to the assembly scheme, so that the accuracy and efficiency of pipeline assembly are both improved. Meanwhile, through the AR glasses, a good man-machine interaction interface is provided for the constructors, and therefore work of the constructors is facilitated. Besides, feedback judgment is added during construction, so that the system has the capability of quickly and accurately responding according to feedback, the assembly scheme can be continuously updated in the construction process, and the pipeline assembly is more accurate.

Description

Technical Field

[0001] The present invention relates to the technical field of ship pipeline assembly, and in particular to a ship pipeline assembly auxiliary method based on AI and knowledge graph. Background Art

[0002] Ship assembly is one of the most important activities in the shipbuilding process. At present, domestic ships are moving towards large-scale and high-value-added, and the number of parts they contain is also increasing. The assembly process of sections and units is becoming more and more complicated. For example, the selection of the pipeline assembly sequence is not unique, but there are multiple possibilities. For some specific parts, such as fixed-model pipes, the bolts between their connectors must be selected according to specific standards. Specific models and specifications must be selected. This requires that some links in the pipeline assembly process must be flexible enough to adapt to different situations, and must ensure strict norms to ensure safety and reliability. The traditional pipeline assembly guidance method mainly relies on experience, and manual assembly planning and execution, which makes the accuracy and efficiency of pipeline assembly low. Summary of the invention

[0003] In view of the shortcomings of the prior art mentioned above, the purpose of the present invention is to provide a ship pipeline assembly auxiliary method based on AI and knowledge graph, which is used to solve the problem of low accuracy and efficiency of pipeline assembly caused by manual planning and execution of pipeline assembly based on experience in the prior art.

[0004] To achieve the above objectives and other related objectives, the present invention provides a ship piping assembly auxiliary method based on AI and knowledge graph, comprising:

[0005] Construct a large model for ship piping assembly;

[0006] Construct a knowledge graph network for ship piping assembly;

[0007] Select the assembly object;

[0008] Acquire relevant data of the assembly object through the knowledge graph network;

[0009] According to the relevant data and in combination with the conditions of the construction site, an assembly plan is provided using the domain large model;

[0010] Instruct construction personnel to assemble pipelines according to the assembly plan;

[0011] The construction personnel provide feedback on the pipeline assembly status, and provide a new assembly plan based on the feedback, and guide the construction personnel to continue the pipeline assembly based on the new assembly plan.

[0012] Optionally, the step of the construction personnel providing feedback on the pipeline assembly situation and providing a new assembly plan based on the feedback includes:

[0013] Determining whether the feedback is an assembly sequence type feedback;

[0014] If yes, it is directly processed by the domain large model and a new assembly scheme is given;

[0015] If not, after the feedback is processed by the knowledge graph network, it is processed by the domain big model and a new assembly plan is given.

[0016] Optionally, the feedback is processed by the knowledge graph network and then processed by the domain big model, including:

[0017] Through the knowledge graph network, synonyms are searched and recombined to form "different characters but similar meanings" questions, and then pushed to the domain big model for analysis; or,

[0018] The basic conclusions are extended and inferred through the knowledge graph network, and then the basic conclusions are combined and pushed to the large domain model for processing.

[0019] Optionally, the step of constructing a large model of the ship pipeline assembly field includes:

[0020] Record the pipe dimensions, rule requirements, assembly sequence, cycle time, number of people, tooling and site resources during the assembly of different sections and units;

[0021] The incorrect combinations of the information are marked, and the large model of the pipeline assembly field is obtained through training in combination with the relevant terms of ship assembly.

[0022] Optionally, the selection of the assembly object and the guidance of the construction personnel to assemble the pipeline according to the assembly plan are both achieved by using AR glasses.

[0023] Optionally, the method of selecting the assembly object includes voice interaction, gesture selection and eye tracking.

[0024] Optionally, the method of guiding construction personnel to assemble pipelines is to use the knowledge graph network to associate the assembly plan with each work step and the corresponding part model, and guide on-site construction personnel to assemble pipelines through AR virtual-reality superposition.

[0025] Optionally, the assembly plan includes whether assembly work is currently being carried out and assembly requirements and sequence.

[0026] Optionally, the relevant data includes parts manufacturing installation drawings, three-dimensional models, process guidance documents, relevant construction standards and reference operation videos.

[0027] Optionally, the conditions at the construction site include incoming materials, personnel arrival, dynamic interference, static interference, and assembly difficulty and danger.

[0028] In a ship pipeline assembly auxiliary method based on AI and knowledge graph of the present invention, through the construction of a large domain model and a knowledge graph network, a plan for ship pipeline assembly can be given more quickly and accurately. The construction personnel carry out construction according to the assembly plan, so that the accuracy and efficiency of pipeline assembly are improved. At the same time, through AR glasses, a good human-computer interaction interface is provided for the construction personnel, thereby facilitating the work of the construction personnel. In addition, by adding feedback judgment during construction, the system has the ability to respond quickly and accurately according to feedback, so that the assembly plan can be continuously updated during the construction process to make the pipeline assembly more accurate. BRIEF DESCRIPTION OF THE DRAWINGS

[0029] Figure 1 This is a flow chart of a ship pipeline assembly auxiliary method based on AI and knowledge graph according to an embodiment of the present invention. DETAILED DESCRIPTION

[0030] Refer to the following Figure 1 To describe a ship piping assembly auxiliary method based on AI and knowledge graph of the present invention. In the description of this embodiment, it should be understood that the terms "first" and "second" are used for descriptive purposes only, and cannot be understood as indicating or implying relative importance or implicitly indicating the number of technical features indicated. Thus, the features defined as "first" and "second" may explicitly or implicitly include at least one of the features, that is, include one or more of the features. In the description of the present invention, the meaning of "multiple" is at least two, such as two, three, etc., unless otherwise clearly and specifically defined. When a feature "includes or contains" one or some of the features it covers, unless otherwise specifically described, this indicates that other features are not excluded and other features may be further included.

[0031] Unless otherwise clearly defined and limited, the terms "set", "install", "connect", "connect", "fix", "couple" and the like should be understood in a broad sense, for example, it can be a fixed connection, a detachable connection, or an integral one; it can be a mechanical connection or an electrical connection; it can be a direct connection or an indirect connection through an intermediate medium, it can be the internal connection of two elements or the interaction relationship between two elements, unless otherwise clearly defined. A person skilled in the art should be able to understand the specific meanings of the above terms in the present invention according to the specific circumstances.

[0032] In addition, in the description of this embodiment, the first feature being "above" or "below" the second feature may include the first and second features being in direct contact, or may include the first and second features not being in direct contact but being in contact through another feature between them. That is, in the description of this embodiment, the first feature being "above", "above", and "above" the second feature includes the first feature being directly above and obliquely above the second feature, or simply indicates that the first feature is higher in level than the second feature. The first feature being "below", "below", or "below" the second feature may mean that the first feature is directly below or obliquely below the second feature, or simply indicates that the first feature is lower in level than the second feature.

[0033] In the description of the present embodiment, the description with reference to the terms "one embodiment", "some embodiments", "illustrative embodiments", "examples", "specific examples", or "some examples" means that the specific features, structures, materials, or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present invention. In this specification, the schematic representation of the above terms does not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials, or characteristics described may be combined in any one or more embodiments or examples in a suitable manner.

[0034] like Figure 1 As shown, an embodiment of the present invention provides a ship pipeline assembly auxiliary method based on AI and knowledge graph, including:

[0035] Step S1: Construct a large model of the ship pipeline assembly field to provide sequence planning, process definition and auxiliary decision-making for ship pipeline assembly. The steps of constructing a large model of the ship pipeline assembly field include recording the pipe size, rule requirements, assembly sequence, cycle, number of people, tooling and site resources in different sections and unit assembly processes. Combine the various information and mark the wrong combination. Combined with the relevant terms of ship assembly, the large model of the pipeline assembly field can be obtained through training.

[0036] Step S2: Construct a ship pipeline assembly knowledge graph network to provide process push, document review, and AR model loading for ship pipeline installation.

[0037] Step S3: Select the assembly object, that is, select the section where the pipeline needs to be assembled.

[0038] Step S4: Obtain relevant data of the assembly object through the knowledge graph network. The relevant data includes parts manufacturing installation drawings, three-dimensional models, process guidance documents, relevant construction standards, and reference operation videos.

[0039] Step S5: Based on the relevant data and the situation at the construction site, the domain big model is used to give an assembly plan. Among them, the situation at the construction site includes the incoming materials, the arrival of personnel, the dynamic interference, the static interference, the difficulty and danger of assembly. The assembly plan includes whether the assembly work is currently being carried out and the assembly requirements and sequence. That is, whether the assembly work can be carried out at present is judged according to the incoming materials, the arrival of personnel, the dynamic interference, the static interference, the difficulty and danger of assembly. And, according to the situation at the construction site, the requirements and sequence of assembly are given. For example, after the analysis of the domain big model, it is found that the current assembly work will be dangerous, so the pipeline assembly work will be postponed. According to the incoming materials and the arrival of personnel, the assembly work that can be carried out at this stage is given, and the amount of materials that need to be supplemented later is given.

[0040] Step S6: Instruct construction personnel to assemble pipelines according to the assembly plan.

[0041] By building a large domain model and knowledge graph network, a plan for ship pipeline assembly can be given more quickly and accurately. Construction workers carry out construction according to the assembly plan, which improves the accuracy and efficiency of pipeline assembly.

[0042] Furthermore, AR glasses are used to select assembly objects and guide construction workers to assemble pipelines according to assembly plans. Among them, the methods of selecting assembly objects include voice interaction, gesture selection, and eye tracking. The method of guiding construction workers to assemble pipelines is to use the knowledge graph network to associate the assembly requirements and sequence of the assembly plan with each work step and the corresponding part model, and guide on-site construction workers to assemble pipelines through AR virtual-real superposition.

[0043] AR glasses can provide construction workers with a good human-computer interaction interface, making their work more convenient. Of course, other human-computer interaction methods can also be used.

[0044] Furthermore, when the construction workers assemble the pipeline, they also include:

[0045] Step S7: The construction personnel provide feedback on the pipeline assembly situation, that is, the problems encountered by the construction personnel during the construction process can be fed back to the system.

[0046] Step S8: Determine whether the feedback is of the assembly sequence type, that is, the system determines the feedback from the construction personnel.

[0047] Step S9: If yes, the domain big model directly processes the feedback and gives a new assembly plan. When the system determines that the feedback is of assembly sequence type, the domain big model directly processes the feedback and gives a new assembly plan.

[0048] Step S10: If not, after the knowledge graph network processes the feedback, it is processed by the domain model and a new assembly plan is given. When the system determines that the feedback is not an assembly sequence type, the feedback is first processed by the knowledge graph network, and then passed to the domain model for processing, and a new assembly plan is given, so that the new assembly plan given by the domain model is more accurate. After the knowledge graph network processes the feedback, it is processed by the domain model, including searching for synonyms through the knowledge graph network to recombine to form a "different words, similar meanings" problem, and then pushing it to the domain model for analysis and extending the knowledge graph network to infer the basic conclusion, and then combining the basic conclusion and pushing it to the domain model for processing.

[0049] Step S11: Instruct the construction personnel to continue assembling the pipeline according to the new assembly plan.

[0050] By adding feedback judgment during construction, the system has the ability to respond quickly and accurately based on feedback, so that the assembly plan can be continuously updated during the construction process, making the pipeline assembly more accurate.

[0051] Optionally, when the assembly object is selected as "FF1A segmented pipe system", the construction worker wears the AR glasses and speaks the object to be assembled "FF1A segmented pipe system" through voice interaction. After the AR glasses recognize the voice, they feed the information back to the knowledge graph network.

[0052] Based on the knowledge graph network, the relevant documents required for FF1A pipeline assembly are quickly retrieved, including: FF1A segment and pipeline 3D model and annotation, FF1A pipeline installation tray table, all pipe parts and accessories production drawing models, assembly process requirements related standards and documents, required matching bolt connection specifications and quantities, and AR safety precautions and other documents.

[0053] According to the pipeline information retrieved from the knowledge graph network, FF1A needs to assemble a total of 100 pipes, 5 elbows, 20 valves and the corresponding pipeline dimensions. Combined with the on-site construction personnel scanning the QR code of each pipe accessory to count the incoming materials, it was found that only 40 pipes, 2 elbows and 16 valves were delivered. After comprehensive judgment of dynamic interference, static interference, assembly difficulty and danger, the domain big model will give an assembly plan for nearly 2 days based on the existing materials. Then it is transmitted to AR glasses through the 5G network, and the on-site construction personnel are guided to assemble the pipeline through the AR virtual-real superposition method.

[0054] If the construction workers find that the next PP1B pipe to be installed does not meet the precision requirements and cannot be installed immediately after assembling 15 pipes, 1 elbow and 3 valves, they can request to re-plan the assembly sequence through voice. At this time, the system determines that it is an assembly sequence type problem and directly hands the problem to the domain big model. The domain big model determines that the subsequent pipes PP1C and PP1D connected to the PP1B pipe with substandard precision cannot be installed, so these pipes are automatically removed from the assembly sequence, and another pipe TB1A, TB1B, and TB1C that have been delivered are selected as the next assembly content, and it is ensured that the installation of this pipe first will not interfere with the subsequent installation of PP1B, PP1C, and PP1D, and the latest assembly sequence of the past two days is updated accordingly. The updated content will be updated on the AR glasses. At this time, when the construction workers say "what to install next", they can see the corresponding TB1A pipeline model highlighted in the AR glasses instead of the originally planned PP1B pipeline. The TB1A pipeline installation process requirements, annotation information, and background model are all displayed to guide the construction workers for quick assembly.

[0055] Optionally, when the assembly object is selected as "WW6B segmented pipe system", the construction worker wears the AR glasses and speaks the object to be assembled "WW6B segmented pipe system" through voice interaction. After the AR glasses recognize the voice, they feed the information back to the knowledge graph network.

[0056] Based on the knowledge graph network, the relevant documents required for WW6B pipeline assembly are quickly retrieved, including: WW6B segment and pipeline 3D model and annotation, WW6B pipeline installation pallet table, all pipe parts and accessories production drawing models, assembly process requirements related standards and documents, required matching bolt connection specifications and quantities, and AR safety precautions and other documents.

[0057] According to the pipeline information retrieved from the knowledge graph network, WW6B needs to assemble a total of 500 pipes, 79 elbows, 56 valves and the corresponding pipeline dimensions. Combined with the on-site construction personnel scanning the QR code of each pipe accessory to count the incoming materials, it was found that only 200 pipes, 36 elbows and 11 valves were delivered. After comprehensive judgment of dynamic interference, static interference, assembly difficulty and danger, the domain big model will give an assembly plan for nearly 2 days based on the existing materials. The assembly plan is transmitted to AR glasses through the 5G network, and the on-site construction personnel are guided to assemble the pipeline through the AR virtual-real superposition method.

[0058] If on the day of construction, Zhang San takes a day off due to illness. The team assembled 100 pipes, 20 elbows and 6 valves the day before. The UU21 pipe to be installed today cannot be found on the pallet on site and cannot be assembled as planned. The construction personnel say through the AR glasses voice interaction, "One person is absent today, and the next pipe to be installed cannot be found. What should I do?" At this time, the system determines that it is not a pure assembly sequence type problem, but a comprehensive problem. The system will hand over the problem to the knowledge graph network. The knowledge graph network first retrieves that Zhang San is a material sorter, who has an impact on the assembly work of the next day, mainly affecting the sorting work after the incoming materials. Therefore, the material sorting efficiency index parameters sent to the domain big model will be reduced, and the rest will remain unchanged. In addition, after the knowledge graph network obtains the name of the UU21 pipe, it can automatically retrieve its model appearance and the time of scanning the incoming materials. The domain big model automatically adjusts the production rhythm according to the new material sorting efficiency parameters input by the knowledge graph network, and also updates the assembly sequence of the next day accordingly. The system updates the updated content on the AR glasses, and the construction workers can see the UU21 tube model and scanned code information pushed by the system, and trace the source accordingly. On the other hand, they can also see the updated assembly sequence.

[0059] The above embodiments are merely illustrative of the principles and effects of the present invention, and are not intended to limit the present invention. Anyone familiar with the art may modify or alter the above embodiments without departing from the spirit and scope of the present invention. Therefore, all equivalent modifications or alterations made by a person of ordinary skill in the art without departing from the spirit and technical concept disclosed by the present invention shall still be covered by the claims of the present invention.

Claims

1. A ship pipeline assembly auxiliary method based on AI and knowledge graph, characterized in that: include: Construct a large model in the field of ship piping assembly; Construct a knowledge graph network for ship piping assembly; Select the assembly object; Acquire relevant data of the assembly object through the knowledge graph network; According to the relevant data and in combination with the conditions of the construction site, an assembly plan is provided using the domain large model; Instruct construction personnel to assemble pipelines according to the assembly plan; The construction personnel provide feedback on the pipeline assembly status, and provide a new assembly plan based on the feedback, and guide the construction personnel to continue the pipeline assembly based on the new assembly plan.

2. The ship pipeline assembly auxiliary method based on AI and knowledge graph according to claim 1 is characterized in that: The construction personnel provide feedback on the pipeline assembly situation and provide a new assembly plan based on the feedback, including: Determining whether the feedback is an assembly sequence type feedback; If yes, it is directly processed by the domain large model and a new assembly scheme is given; If not, after the feedback is processed by the knowledge graph network, it is processed by the domain big model and a new assembly plan is given.

3. The ship piping assembly auxiliary method based on AI and knowledge graph according to claim 2 is characterized in that: After the feedback is processed by the knowledge graph network, it is then processed by the domain big model, including: Through the knowledge graph network, synonyms are searched and recombined to form "different characters but similar meanings" questions, and then pushed to the domain big model for analysis; or, The basic conclusions are extended and inferred through the knowledge graph network, and then the basic conclusions are combined and pushed to the large domain model for processing.

4. The ship piping assembly auxiliary method based on AI and knowledge graph according to claim 1 is characterized in that: The steps of constructing a large model of the ship pipeline assembly field include: Record the pipe dimensions, rule requirements, assembly sequence, cycle time, number of people, tooling and site resources during the assembly of different sections and units; The incorrect combinations of the information are marked, and the large model of the pipeline assembly field is obtained through training in combination with the relevant terms of ship assembly.

5. The ship piping assembly auxiliary method based on AI and knowledge graph according to claim 1 is characterized in that: The selection of the assembly object and the guidance of the construction personnel to assemble the pipeline according to the assembly plan are both achieved by using AR glasses.

6. The ship piping assembly auxiliary method based on AI and knowledge graph according to claim 5 is characterized in that: The methods of selecting assembly objects include voice interaction, gesture selection and eye tracking.

7. The ship piping assembly auxiliary method based on AI and knowledge graph according to claim 5 is characterized in that: The method of guiding construction personnel to assemble pipelines is to use the knowledge graph network to associate the assembly plan with each work step and the corresponding part model, and guide on-site construction personnel to assemble pipelines through AR virtual-reality superposition.

8. The ship piping assembly auxiliary method based on AI and knowledge graph according to claim 1 is characterized in that: The assembly plan includes whether assembly work is currently being carried out and the assembly requirements and sequence.

9. The ship piping assembly auxiliary method based on AI and knowledge graph according to claim 1 is characterized in that: The relevant data include parts manufacturing and installation drawings, three-dimensional models, process guidance documents, relevant construction standards and reference operation videos.

10. The ship piping assembly auxiliary method based on AI and knowledge graph according to claim 1 is characterized in that: The conditions at the construction site include incoming materials, personnel arrival, dynamic interference, static interference, and assembly difficulty and danger.

Images