Ship three-dimensional design evaluation method, evaluation system and readable storage medium

Through the combination of multimodal large model and natural language large model, ship design information is automatically extracted and compared and analyzed, and the problem of manual comparison and analysis is solved, efficient design information consistency review is achieved, and the review efficiency is improved and the review is realized. Intelligent review.

CN119150558APending Publication Date: 2024-12-17JIANGNAN SHIPYARD (GRP) CO LTD
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202411284298.3
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-09-13
Publication Date
2024-12-17

AI Technical Summary

Technical Problem

The differences in the information description format of ship three-dimensional design models, two-dimensional schematic diagrams and equipment data lead to long-term and low efficiency in manual comparison and analysis.

Method used

Multimodal big model and natural language big model are used to automatically extract and compare the attribute information in three-dimensional models, two-dimensional schematic diagrams and equipment data to achieve consistency review of information.

Benefits of technology

It significantly improves the review efficiency, realizes automatic design information comparison and comprehensive analysis, and does not require manual adjustment of the information format, and realizes intelligent review.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN119150558A_ABST
    Figure CN119150558A_ABST
Patent Text Reader

Abstract

The invention provides a ship three-dimensional design review method. The method comprises the steps that a multi-modal large model and a natural language large model used for three-dimensional design review are acquired; respectively extracting attribute information and principle design information from the three-dimensional model and the two-dimensional schematic diagram, and extracting equipment information from the equipment data based on the multi-modal large model; inputting attribute information of the three-dimensional model and principle design information of the two-dimensional schematic diagram into a natural language large model for comparative analysis; inputting attribute information and equipment information of the three-dimensional model into a natural language large model for comparative analysis; summarizing to obtain a comparison result of the three-dimensional model; and reviewing the summarized comparison result to obtain a review result. According to the method, the comparison result can be automatically generated, the final three-dimensional design review result is generated, intelligent review of ship three-dimensional design is achieved, and therefore the problem that manual review efficiency is low is solved. The invention further provides a review system for executing the method and a readable storage medium.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the field of ship design, and in particular to a review method, review system and readable storage medium for three-dimensional ship design. Background Art

[0002] The number of parts and components of ship products is huge and complex, and their 3D design models need to be consistent with the information in the 2D schematics and equipment data. However, the 3D design model, 2D schematic design, and equipment data are completed by different people, and the standards and specifications they follow are not the same, resulting in differences in the description format of the same information (such as name, specification, size) in the 3D design model, 2D schematic, and equipment data. For example, "uninterruptible power supply, UPS, power UPS, marine uninterruptible power supply" describes the same device. The consistency of the information between the above three needs to be compared and analyzed during the 3D design review. Currently, designers need to manually extract different information, identify the same equipment, and complete the comparative analysis of the information. This process is time-consuming and inefficient. Summary of the invention

[0003] In view of the shortcomings of the related technologies mentioned above, the purpose of the present invention is to provide a review method, review system and readable storage medium for ship three-dimensional design, which are used to replace manual review of the consistency of information among three-dimensional models, two-dimensional schematics and equipment data, so as to improve work efficiency.

[0004] To achieve the above-mentioned and other related purposes, the present invention provides a method for reviewing three-dimensional ship design, the method comprising the following steps:

[0005] S1. Obtain multimodal big models and natural language big models for 3D design review;

[0006] S2, extracting attribute information and principle design information from the three-dimensional model and the two-dimensional schematic diagram respectively, and extracting device information from the device data based on the multimodal large model;

[0007] S3, inputting the attribute information of the three-dimensional model and the principle design information of the two-dimensional schematic diagram into the natural language large model for comparative analysis;

[0008] S4, inputting the attribute information of the three-dimensional model and the equipment information in the equipment data into the natural language large model for comparative analysis;

[0009] S5, based on the comparative analysis of steps S3 and S4, summarizing the comparison results of the three-dimensional model;

[0010] S6. Re-evaluate the comparison result in step S5 to obtain an evaluation result.

[0011] Optionally, the step of obtaining the multimodal large model is: using text and pictures of the device information as training data to train the multimodal large model.

[0012] Optionally, the step of obtaining a large natural language model is: using information of the same device in different description forms as training data to train and obtain a large natural language model.

[0013] Optionally, the attribute information includes model name, dimensions, specifications, three-dimensional coordinates, part number and weight;

[0014] The principle design information includes extraction of drawing history information, parts information, and connection relationship information;

[0015] The device information includes the device name, device dimensions and accessory specifications.

[0016] Optionally, according to the model name of the three-dimensional model and based on the material classification table, the material classification information of the three-dimensional model is obtained; according to the three-dimensional coordinates of the three-dimensional model and based on the ship space division table, the area information to which the three-dimensional model belongs is obtained; according to the starting and ending points of the pipeline model or the starting and ending points of the cable channel model in the three-dimensional model, the connection relationship information of the pipeline system or the electrical system is obtained.

[0017] Optionally, the step of inputting the attribute information of the three-dimensional model and the principle design information of the two-dimensional schematic diagram into the natural language big model for comparative analysis is: inputting the model name and material classification information of the three-dimensional model and the part information of the two-dimensional schematic diagram into the natural language big model, and the natural language big model outputs the comparison result;

[0018] The specifications of the 3D model and the parts information of the 2D schematic diagram are input into the natural language big model, and the natural language big model outputs the comparison results;

[0019] Input the 3D coordinate information, the area information of the 3D model and the drawing history information of the 2D schematic diagram into the natural language big model, and the natural language big model outputs the comparison result;

[0020] The connection relationship information of the three-dimensional model and the connection relationship information of the two-dimensional schematic diagram are input into the natural language big model, and the natural language big model outputs the comparison result.

[0021] Optionally, the step of inputting the attribute information of the three-dimensional model and the device information of the device data into the natural language big model for comparative analysis is: inputting the model name of the three-dimensional model and the device name of the device information into the natural language big model, and the natural language big model outputs the comparison result;

[0022] Input the external dimensions of the three-dimensional model and the external dimensions of the device information into the natural language large model, and the natural language large model outputs a comparison result;

[0023] The specifications of the three-dimensional model and the accessory specifications of the equipment information are input into the natural language big model, and the natural language big model outputs the comparison results.

[0024] Optionally, the re-evaluation step includes: performing interference check on the three-dimensional model, and using the interference check result as a reference to correct and supplement the review opinions.

[0025] A review system for executing the review method for three-dimensional ship design as described above, the review system comprising:

[0026] Model management module, used for training, testing and service deployment of large models;

[0027] A three-dimensional model information extraction module, used to extract information from the three-dimensional model;

[0028] A two-dimensional drawing information extraction module is used to extract information from a two-dimensional schematic diagram;

[0029] Document information extraction module, used to extract device information from device data;

[0030] Data management module, used to manage the training data of multimodal large models and natural language large models as well as business data during 3D design review;

[0031] The user interaction module is used for users to trigger review operations to re-evaluate review information, call other module functions, and output and save review results.

[0032] A readable storage medium storing computer instructions, wherein the computer instructions are used to cause the computer to execute the above-mentioned review method for three-dimensional ship design.

[0033] As described above, the review method, review system and readable storage medium of the three-dimensional design of a ship of the present invention have the following beneficial effects: the method uses a multimodal large model to automatically extract equipment information of equipment data, and uses a natural language large model to automatically compare and analyze various types of design information in different expressions. Compared with manual review, this method can automatically extract and comprehensively analyze various types of design information without manual pre-adjustment of the information format and content, and can automatically output the comparison results, significantly improving the review efficiency and realizing the intelligent review of the three-dimensional design of ships. BRIEF DESCRIPTION OF THE DRAWINGS

[0034] Figure 1 Shown is a schematic flow chart of a method for reviewing three-dimensional ship design in an embodiment of the present invention.

[0035] Figure 2 Shown is a schematic diagram of a review system in an embodiment of the present invention.

[0036] Component number description

[0037] 10 Model Management Module

[0038] 20 Data Management Module

[0039] 30 3D model information extraction module

[0040] 40 2D drawing information extraction module

[0041] 50 Document Information Extraction Module

[0042] 60 User Interaction Module DETAILED DESCRIPTION

[0043] The following describes the embodiments of the present invention through specific examples, and those skilled in the art can easily understand other advantages and effects of the present invention from the contents disclosed in this specification. The present invention can also be implemented or applied through other different specific embodiments, and the details in this specification can also be modified or changed in various ways based on different viewpoints and applications without departing from the spirit of the present invention.

[0044] For example, when describing the embodiments of the present invention in detail, for the sake of convenience, the cross-sectional view showing the device structure will not be partially enlarged according to the general scale, and the schematic view is only an example, which should not limit the scope of protection of the present invention. In addition, in actual production, the three-dimensional space dimensions of length, width and depth should be included.

[0045] For ease of description, spatial relational terms such as "under", "below", "below", "below", "above", "on", etc. may be used herein to describe the relationship of one element or feature shown in the drawings to other elements or features. It will be understood that these spatial relational terms are intended to include other directions of the device in use or operation in addition to the directions depicted in the drawings. In addition, when a layer is referred to as being "between" two layers, it can be the only layer between the two layers, or there can be one or more intervening layers. As used herein, "between..." means including the end point values.

[0046] In the context of the present application, a structure in which a first feature is described as being "above" a second feature may include embodiments in which the first and second features are formed in direct contact, and may also include embodiments in which additional features are formed between the first and second features, such that the first and second features may not be in direct contact.

[0047] It should be noted that the illustrations provided in this embodiment are only used to illustrate the basic concept of the present invention in a schematic manner, and therefore the illustrations only show components related to the present invention rather than being drawn according to the number, shape and size of components in actual implementation. In actual implementation, the type, quantity and proportion of each component may be changed arbitrarily, and the component layout may also be more complicated.

[0048] like Figure 1 As shown, the present invention provides a review method for three-dimensional design of a ship, and the specific steps of the method include:

[0049] S1. Obtain multimodal big models and natural language big models for 3D design review.

[0050] Specifically, we first conduct preparatory work and collect training data. The training data of the multimodal large model is the text and pictures of the equipment. The text and pictures can provide the corresponding equipment name information, equipment size information, and accessory specifications. For example, when extracting the length, width, and height of the equipment, the multimodal large model will identify the length, width, and height of the equipment from the picture, and obtain the size from the text at the same time. By combining the information of the picture and the information of the text, the size information of the equipment is completely extracted. By training the polymorphic large model so that the polymorphic large model can continuously adjust the weight parameters within the model, the polymorphic large model can extract information from pictures and text more accurately.

[0051] The training data of the natural language model is information in different description forms of the same device, such as model name, dimensions, specifications, coordinates, region, part number, connection relationship, etc. In one example, the name of the uninterruptible power supply can also be written as UPS, power UPS, or marine uninterruptible power supply. By training the natural language model so that the natural language model can continuously adjust the weight parameters in the model, the natural language model can establish a connection relationship between names in different description forms, and then more accurately judge whether the names in different description forms are consistent.

[0052] S2. Extract attribute information and principle design information from the three-dimensional model and the two-dimensional schematic diagram respectively, and extract device information from the device data based on the multimodal large model.

[0053] The same equipment usually corresponds to three forms of data: 3D model data, 2D schematic data and equipment data. In order to compare the consistency of the three types of data, attribute information and principle design information are extracted from 3D software and 2D software respectively, and the equipment data is input into the multimodal large model at the same time. The multimodal large model can extract the equipment information of the equipment data. Among them, the attribute information includes model name, size, specification, 3D coordinates, part number and weight.

[0054] Taking the uninterruptible power supply as an example, the attribute information extracted from the 3D software is shown in Table 1:

[0055] Table 1

[0056] Model ID 1UPS / DFG Model Name Generators and accessories - Main generator uninterruptible power supply Model material classification Generators and accessories Overall dimensions - length (mm) 600.2 Overall dimensions - width (mm) 450.2 Overall dimensions - height (mm) 1450.2 Specification - X-axis coordinate of center of gravity (mm) 41014.0 Y-axis coordinate of center of gravity (mm) -7104.3 Center of gravity Z axis coordinate (mm) 21945.0 Weight(kg) 140 Area 2 Platform (Engine Room Area) Connecting pipe segments - Connection cable channel JC-DLTD-11

[0057] The principle design information includes the extracted drawing history information, part information, and connection relationship information; the drawing history information includes the name of the region and system to which it belongs, the part information includes the part name, part number, part specification, etc., and the connection relationship information includes the equipment connection relationship information of the pipeline system and the electrical system, that is, which devices are connected. The principle design information extracted from the two-dimensional software is shown in Table 2:

[0058] Table 2

[0059]

[0060] Equipment information includes equipment name, equipment dimensions and accessory specifications. When the multimodal large model extracts equipment information, manual correction is required to ensure the accuracy of the extraction. The information extracted from the equipment information by the multimodal large model is shown in Table 3:

[0061] Table 3

[0062] Device Name Marine Uninterruptible Power Supply (CUHD-AII / 102II-2) Overall dimensions - length (mm) Width 600 Overall dimensions - width (mm) Depth 450 Overall dimensions - height (mm) Height 1450 weight 140Kg±3% Specification 220VAC±2%; 60HZ±0.1HZ Connection cable specifications 2X2.5

[0063] S3. Input the attribute information of the three-dimensional model and the principle design information of the two-dimensional schematic diagram into the natural language large model for comparative analysis.

[0064] According to the model name of the three-dimensional model and based on the material classification table, the material classification information of the three-dimensional model can be obtained. The model name of the three-dimensional model, the material classification information and the part name of the two-dimensional schematic diagram are input into the natural language big model. The natural language big model outputs the comparison result. For example, the model name 1UPS / DFG of the three-dimensional model is consistent with the part name of the two-dimensional schematic diagram, and there is no error.

[0065] The specifications of the 3D model and the part specifications of the 2D schematic are input into the natural language large model, and the natural language large model outputs the comparison results. If the specifications of the 3D model match the part specifications of the 2D schematic, there are no errors.

[0066] The three-dimensional coordinate information, area information and the area of ​​the two-dimensional schematic diagram of the three-dimensional model are input into the natural language large model, and the natural language large model outputs the comparison result. If the area of ​​the three-dimensional model matches the area of ​​the two-dimensional schematic diagram, there is no error.

[0067] The connection relationship information of the three-dimensional model and the connection relationship information of the two-dimensional schematic diagram are input into the natural language large model, and the natural language large model outputs the comparison result. For example, it determines whether the connection relationship between device A and device B in the three-dimensional model is the same as the connection relationship in the two-dimensional schematic diagram. The natural language large model will determine whether the information of device A, device B and parts used to connect device A and device B (such as cables) in the three-dimensional model and the two-dimensional schematic diagram is consistent. If the information is consistent, the connection relationship of the output three-dimensional model is consistent with the connection relationship of the two-dimensional schematic diagram without error.

[0068] S4. Input the attribute information of the three-dimensional model and the equipment information of the equipment data into the natural language large model for comparative analysis.

[0069] The model name of the 3D model and the device name of the device information are input into the natural language large model, and the natural language large model outputs a comparison result. For example, the model name 1UPS / DFG of the 3D model matches the device name of the device information and there is no error.

[0070] The external dimensions of the 3D model and the external dimensions of the device information are input into the natural language large model, and the natural language large model outputs a comparison result. If the external dimensions of the 3D model match the external dimensions of the device information, there is no error.

[0071] The specifications of the 3D model and the accessory specifications of the equipment information are input into the natural language large model, and the natural language large model outputs a comparison result. If the specifications of the 3D model match the accessory specifications of the equipment information, there is no error.

[0072] S5. Based on the comparative analysis of steps S3 and S4, the comparison results of the three-dimensional model are summarized.

[0073] The comparison results are obtained based on the natural language large model summary. For example, "After comparison, the model name, size, weight, region, and connection relationship of the 3D model are consistent with the principle design information of the 2D schematic diagram and the equipment information of the equipment, and there are no errors.

[0074] S6. Re-evaluate the comparison result in step S5 to obtain an evaluation result.

[0075] The interference check function is used to perform interference check on the 3D model, and the interference check results are used as a reference. The review experts review the summarized comparison results and interference check results, and correct and supplement the review opinions to obtain the final 3D model review results.

[0076] The review results revised by experts are added to the training data to increase the sample size of the training data, and the accuracy of the review is improved by continuing to train the large natural language model.

[0077] The present application also provides a review system for executing the above method, which includes a model management module 10, a data management module 20, a three-dimensional model information extraction module 30, a two-dimensional drawing information extraction module 40, a document information extraction module 50 and a user interaction module 60.

[0078] The model management module 10 is used for training, testing and service deployment of multimodal large models and natural language large models. The data management module 20 is used to manage the training data of multimodal large models and natural language large models and the business data during the three-dimensional design review, where the business data refers to the data generated in the entire process from the input of the three-dimensional model to the natural language large model for review to the output of the review results; the three-dimensional model information extraction module 30 is used to extract information from the three-dimensional model; the two-dimensional drawing information extraction module 40 is used to extract information from the two-dimensional schematic diagram; the document information extraction module 50 is used to extract the device information of the device; the user interaction module 60 is used for the user to trigger the review operation to re-evaluate the review information, call other module functions, and output and save the review results.

[0079] The present application also provides a readable storage medium storing computer instructions, wherein the computer instructions are used to enable the computer to execute the above-mentioned review method for three-dimensional ship design.

[0080] 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 review method for three-dimensional ship design, characterized in that: The steps of the method include: S1. Obtain multimodal big models and natural language big models for 3D design review; S2, extracting attribute information and principle design information from the three-dimensional model and the two-dimensional schematic diagram respectively, and extracting device information from the device data based on the multimodal large model; S3, inputting the attribute information of the three-dimensional model and the principle design information of the two-dimensional schematic diagram into the natural language large model for comparative analysis; S4, inputting the attribute information of the three-dimensional model and the equipment information in the equipment data into the natural language large model for comparative analysis; S5, based on the comparative analysis of steps S3 and S4, summarizing the comparison results of the three-dimensional model; S6. Re-evaluate the comparison result in step S5 to obtain an evaluation result.

2. The review method for three-dimensional ship design according to claim 1, characterized in that: The step of obtaining the multimodal large model is: using the text and pictures of the equipment data as training data to train the multimodal large model.

3. The review method for three-dimensional ship design according to claim 1, characterized in that: The steps of obtaining a large natural language model are: using information of the same device in different description forms as training data to train a large natural language model.

4. The review method for three-dimensional ship design according to claim 1, characterized in that: The attribute information includes model name, dimensions, specifications, three-dimensional coordinates, part number and weight; The principle design information includes extraction of drawing history information, parts information, and connection relationship information; The device information includes the device name, device dimensions and accessory specifications.

5. The review method for three-dimensional ship design according to claim 4, characterized in that: According to the model name of the three-dimensional model and based on the material classification table, the material classification information to which the three-dimensional model belongs is obtained; According to the three-dimensional coordinates of the three-dimensional model, based on the ship space division table, the area information to which the three-dimensional model belongs is obtained; According to the starting and ending points of the pipeline model or the starting and ending points of the cable channel model in the three-dimensional model, the connection relationship information of the pipeline system or the electrical system is obtained.

6. The review method for three-dimensional ship design according to claim 5, characterized in that: The step of inputting the attribute information of the three-dimensional model and the principle design information of the two-dimensional schematic diagram into the natural language big model for comparative analysis is: inputting the model name and material classification information of the three-dimensional model and the part information of the two-dimensional schematic diagram into the natural language big model, and the natural language big model outputs the comparison result; The specifications of the 3D model and the parts information of the 2D schematic diagram are input into the natural language big model, and the natural language big model outputs the comparison results; Input the 3D coordinate information, the area information of the 3D model and the drawing history information of the 2D schematic diagram into the natural language big model, and the natural language big model outputs the comparison result; The connection relationship information of the three-dimensional model and the connection relationship information of the two-dimensional schematic diagram are input into the natural language big model, and the natural language big model outputs the comparison result.

7. The review method for three-dimensional ship design according to claim 4, characterized in that: The step of inputting the attribute information of the three-dimensional model and the device information in the device data into the natural language big model for comparative analysis is: inputting the model name of the three-dimensional model and the device name of the device information into the natural language big model, and the natural language big model outputs the comparison result; Input the external dimensions of the three-dimensional model and the external dimensions of the device information into the natural language large model, and the natural language large model outputs a comparison result; The specifications of the three-dimensional model and the accessory specifications of the equipment information are input into the natural language big model, and the natural language big model outputs the comparison results.

8. The review method for three-dimensional ship design according to claim 1, characterized in that: The review step includes: performing interference check on the three-dimensional model, and using the interference check result as a reference to correct and supplement the review opinions.

9. A review system for executing the review method for three-dimensional ship design according to any one of claims 1 to 8, characterized in that: The review system includes: Model management module, used for training, testing and service deployment of large models; A three-dimensional model information extraction module, used to extract information from the three-dimensional model; A two-dimensional drawing information extraction module is used to extract information from a two-dimensional schematic diagram; Document information extraction module, used to extract device information from device data; Data management module, used to manage the training data of multimodal large models and natural language large models as well as business data during 3D design review; The user interaction module is used for users to trigger review operations to re-evaluate review information, call other module functions, and output and save review results.

10. A readable storage medium storing computer instructions, characterized in that: The computer instructions are used to enable the computer to execute the review method for three-dimensional ship design according to any one of claims 1-8.