Manufacturing method of ship operation and maintenance manual

Abstract

The invention belongs to the technical field of ship operation and maintenance, and provides a manufacturing method of a ship operation and maintenance manual, comprising the following steps: constructing a natural language model; inputting a user language, and converting the user language into a structured process language by using the natural language model; constructing a multi-modal knowledge graph network in which modal data of an operation object is stored; selecting a manual mode type; obtaining target object modal data from the operation object modal data according to the manual modal type; and forming a ship operation and maintenance manual according to the structured process language and the modal data of the target object. According to the manufacturing method, the natural language model and the multi-modal knowledge graph network are combined for use, the working efficiency of manufacturing the ship operation and maintenance manual is effectively improved, and based on the generated ship operation and maintenance manual, the construction difficulty can be reduced, the construction efficiency can be improved, and the time cost and the material cost of actual production can be saved.

Description

Technical Field

[0001] This application relates to the technical field of ship operation and maintenance, and particularly to a method for making a ship operation and maintenance manual. Background Art

[0002] With the rapid development of the ship industry, ships are moving towards the direction of being larger, more complex, and of higher added value. The development trend of the ship industry has led to a sharp increase in the number of parts contained in ships, resulting in a significant increase in the difficulty of ship operation and maintenance. In the past, when shipyards delivered ships, they mainly provided the shipowners with traditional forms of materials such as drawings, documents, and manuals to guide the use, repair, and operation and maintenance of the ships.

[0003] In the context of the rapid development of digital technology, traditional paper materials have gradually shown their limitations. To better meet the needs of ship operation and maintenance, IETM (Interactive Electronic Technical Manual) has emerged. Through an electronic method, IETM structurally integrates multi-modal information such as text, models, pictures, videos, and animations in the operation and maintenance manuals of traditional paper or technical materials, and constructs a standardized structure tree. This structure tree can effectively organize and present information, making it more suitable for various aspects of work such as the use, repair, and training of equipment.

[0004] However, although IETM has many advantages in information presentation and retrieval, in the actual editing process, for senior ship system debugging personnel, although these professionals have a deep understanding and rich experience in the operation process of ships, due to the high requirements and strong standardization of IETM manual editing, the editing process of IETM manuals is time-consuming and laborious, reducing the efficiency of professionals in editing IETM manuals, and to a certain extent restricting the wide application and further development of IETM in the field of ship operation and maintenance. Summary of the Invention

[0005] The purpose of this application is to provide a method for making a ship operation and maintenance manual to solve problems such as the low production efficiency of ship operation and maintenance manuals in the prior art.

[0006] To achieve the above and other related purposes, this application provides a method for making a ship operation and maintenance manual, including the following steps:

[0007] S1. Construct a natural language model;

[0008] S2. Input the user's language, and use the natural language model to convert the user's language into a structured process language;

[0009] S3. Construct a multi-modal knowledge graph network, where the multi-modal knowledge graph network has operation object modal data;

[0010] S4. Select the manual modal type for generating the ship operation and maintenance manual;

[0011] S5. According to the manual modal type, obtain the target object modal data from the operation object modal data;

[0012] S6. Form a ship operation and maintenance manual according to the structured process language and the target object modal data.

[0013] Optionally, the user language is a natural language, and the structured process language includes operation object information and object action information; the steps of converting the user language into the structured process language in step S2 include:

[0014] Use the natural language model to convert the user language into natural language text;

[0015] Parse the natural language text to obtain operation object information and object action information;

[0016] Form the structured process language according to the operation object information and the object action information.

[0017] Optionally, after converting the user language into a structured language, the following steps are further included:

[0018] Perform a logical check on the structured process language to determine whether there are data defects in the structured process language;

[0019] When there are data defects in the structured process language, generate and output inquiry information.

[0020] Optionally, after converting the user language into a structured process language, the following steps are further included:

[0021] Input user supplementary language according to the inquiry information;

[0022] Use the natural language model to convert the user supplementary language into a structured supplementary language;

[0023] Adjust the structured process language according to the structured supplementary language.

[0024] Optionally, constructing a multi-modal knowledge graph network includes the following steps:

[0025] Build a multi-modal knowledge graph network;

[0026] Obtain the object name data of the operation object;

[0027] Obtain the image dataset, video dataset, and model dataset of the operating object according to the object name data to form the operating object modal data;

[0028] Input the object name data and the operating object modal data into the multi-modal knowledge graph network;

[0029] Associate the object name data with the operating object modal data.

[0030] Optionally, the manual modal type includes one or more of video, graphic and text electronic manual, and document.

[0031] Optionally, the target object modal data includes target image data, target video data, and target modal data.

[0032] Optionally, the structured bold language includes a number of operation steps, and each operation has operation object information and / or object action information; Step S6 includes the following steps:

[0033] Insert the target image data, the target video data, and the target modal data into the number of operation steps according to the operation object information and object action information in each operation step.

[0034] Optionally, the operating object modal data further includes a text dataset of the operating object, and the target object modal data further includes target text data; Step S6 further includes the following steps:

[0035] Add the target text data to the number of operation steps according to the operation object information and object action information in each operation step.

[0036] Optionally, performing step S6 further includes the following steps:

[0037] Detect whether there is target video data in the operating object modal data according to the structured process language;

[0038] When the result is "no", obtain the target modal data from the operating object modal data;

[0039] Call an external editing software to produce an action video clip as the target video data according to the target modal data and the target text data.

[0040] As described above, compared with the prior art, the method for producing a ship operation and maintenance manual provided by the present application has at least the following beneficial effects:

[0041] In the manufacturing method of the present application, the natural language model can automatically convert the input user language into a structured process language that conforms to the specifications for manufacturing a ship operation and maintenance manual, eliminating the need for manual conversion by the user, reducing the workload of the staff, and improving the manufacturing efficiency of the ship operation and maintenance manual. The multimodal knowledge graph network stores the modal data of the operating object. According to the structured process language, the modal data of the target object is directly obtained from the modal data of the operating object, and then the ship operation and maintenance manual is constructed, simplifying the manufacturing process of the ship operation and maintenance manual and further improving the manufacturing efficiency of the ship operation and maintenance manual. It can be seen that the manufacturing method of this embodiment effectively improves the work efficiency of manufacturing the ship operation and maintenance manual by combining the natural language model and the multimodal knowledge graph network. Brief Description of the Drawings

[0042] To more clearly illustrate the technical solutions of the embodiments of the present application, the drawings required for use in the embodiments will be briefly introduced below. It should be understood that the following drawings only show some embodiments of the present application and should not be regarded as limiting the scope. For those of ordinary skill in the art, other relevant drawings can also be obtained based on these drawings without creative efforts.

[0043] Figure 1 It shows a schematic flow chart of a method for manufacturing a ship operation and maintenance manual provided by an embodiment of the present application. Detailed Embodiments

[0044] To make the technical objectives, technical solutions, and technical effects of the present application clearer, the technical solutions in the present application will be clearly and completely described below in conjunction with the embodiments. Obviously, the described embodiments are some, but not all, of the embodiments of the present application. Usually, the components of the embodiments of the present application described and shown in the drawings here can be arranged and designed in various different configurations.

[0045] Therefore, the following detailed description of the embodiments of the present application is not intended to limit the scope of the present application to be protected, but only represents the selected embodiments of the present application. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present application without creative efforts fall within the scope of protection of the present application. In addition, the terms "first" and "second" are only used for descriptive purposes and cannot be understood as indicating or implying relative importance.

[0046] In the description of the present application, it should be noted that the orientation or positional relationship indicated by the terms "center", "longitudinal", "transverse", "upper", "lower", "front", "rear", "left", "right", "vertical", "horizontal", "top", "bottom", "inner", "outer", etc. is based on the orientation or positional relationship shown in the drawings. It is only for the convenience of describing the present application and simplifying the description, rather than indicating or implying that the device referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore should not be construed as a limitation to the present application.

[0047] In the description of the present application, unless otherwise clearly specified and limited, the terms "installation", "connection", and "coupling" should be understood in a broad sense. For example, it can be a fixed connection or a detachable connection. Additionally, the description referring to terms such as "one embodiment", "some embodiments", "schematic embodiments", "examples", "specific examples", or "some examples" means that the specific features, structures, materials, or characteristics described in connection with the embodiments or examples are included in at least one embodiment or example of the present application. In this specification, the schematic descriptions of the above terms do not necessarily refer to the same embodiments or examples. Moreover, the specific features, structures, materials, or characteristics described can be combined in a suitable manner in any one or more embodiments or examples.

[0048] This embodiment provides a method for creating a ship operation and maintenance manual, referring to Figure 1 , including steps S1 to S6, specifically including:

[0049] S1. Construct a natural language model;

[0050] S2. Input the user's language, and use the natural language model to convert the user's language into a structured process language;

[0051] S3. Construct a multi-modal knowledge graph network, where the multi-modal knowledge graph network has operation object modal data;

[0052] S4. Select the manual modal type for generating the ship operation and maintenance manual;

[0053] S5. According to the manual modal type, obtain the target object modal data from the operation object modal data;

[0054] S6. According to the structured process language and the target object modal data, form a ship operation and maintenance manual.

[0055] In this embodiment, the user language is natural language. Through the natural language model, the input user language can be automatically converted into a structured process language that conforms to the production specification of the ship operation and maintenance manual, without the need for the user to manually convert it, reducing the user's workload, saving the user's time cost, and improving the production efficiency of the ship operation and maintenance manual. The multimodal knowledge graph network stores the modal data of the operation objects related to the ship. According to the structured process language, the modal data of the target object is directly obtained from the modal data of the operation object, and then the ship operation and maintenance manual is constructed, simplifying the production process of the ship operation and maintenance manual and further improving the production efficiency of the ship operation and maintenance manual. It can be seen that the production method of this embodiment effectively improves the work efficiency of producing the ship operation and maintenance manual and saves labor costs and time costs by combining the natural language model with the multimodal knowledge graph network.

[0056] In step S1, the natural language model can be an NLP natural language processing model, which can convert the input natural language into a structured process language that conforms to the production specification of the ship operation and maintenance manual.

[0057] In an alternative embodiment, the steps of constructing the natural language model in step S1 include: obtaining a training language data set; building a natural language model; and training the natural language model using the training language data set. Among them, the training language data set can be understood as all possible natural languages related to ship operation and maintenance in the ship field.

[0058] In this embodiment, the user language is natural language. The ways of inputting the user language in step S2 include keyboard input, voice input, or other appropriate input methods. After the staff inputs the user language, the trained natural language model outputs a structured process language according to the input natural language. After the user's staff inputs the user language, the trained natural language model is directly used to convert the user language, which can quickly and effectively convert the natural language into a structured process language without the need for the staff to manually perform the structured conversion of the natural language, effectively reducing the workload of the staff, improving the conversion efficiency of the natural language, saving the user's labor cost and time cost, and improving the work efficiency of producing the ship operation and maintenance manual.

[0059] In an alternative embodiment, the structured process language includes operation object information and object action information; the steps of converting the user language into a structured process language in step S2 include: using the natural language model to convert the user language into natural language text; parsing the natural language text to obtain operation object information and object action information; and forming a structured process language according to the operation object information and object action information.

[0060] Among them, the operation object information is related information such as the identification information of the operation object, and the operation object is the ship structure, components, etc. involved in the specific ship operation and maintenance process. For example, the operation object information includes "fuel valve", "lubricating oil valve", "fan power supply", "control room", "second control console", "generator", "button", etc., and the object action information is the action information that needs to be performed on the operation object during the specific operation and maintenance process. For example, the object action information includes "open", "arrive at", "press", etc. After obtaining the operation object information and the object action information, integrate the operation object information and the object action information, and structure and process them to obtain a structured process language. Optionally, the structured process language may also include other appropriate information.

[0061] In an alternative embodiment, the structured process language may include several operation steps. The number of operation steps may be one or more, and each operation step has operation object information and / or object action information. For example, an operation step has both operation object information and object action information at the same time.

[0062] In an alternative embodiment, taking the example of editing the generator starting operation process, when the user inputs the user language by wearing AR glasses and speaking, during the execution of step S2, the natural language model is used to convert the user language of the voice type into natural language text. For example, for the operation process of editing the generator starting, the natural language text after the voice conversion of the user language is "First, go to the purifier room to open the fuel valve, then open the lubricating oil valve, and then go to the engine room to turn on the power supply of the fan electric control box, and finally click the start button on the second control console in the control room". Then, use the natural language model to parse the natural language text, and form a structured process language according to the obtained operation object information and object action information. For example, after parsing the above natural language text, the obtained structured process language is "(1) Go to the purifier room; (2) Open the fuel valve; (3) Open the lubricating oil valve; (4) Go to the engine room; (5) Turn on the fan power supply; (6) Arrive beside the second control console in the control room; (7) Press the generator start button".

[0063] In an alternative embodiment, after converting the user language into a structured process language in step S2, the following steps may further be included: performing a logical detection on the structured process language to determine whether there are data defects in the structured process language; when there are data defects in the structured process language, generating and outputting inquiry information. Among them, the data defects may include, for example, logical defects, data missing, and other possible defects, such as logical discontinuities and other logical defects, and data missing defects such as missing necessary information. The inquiry information includes a reminder message and a defect detection result. The reminder message is used to remind the user that there are data defects, and the defect detection result is the result of the logical detection of the structured process language. The user can, based on the defect detection result, improve and supplement the structured process language.

[0064] Further, after converting the user language into a structured process language, the following steps may further be included: inputting the user's supplementary language according to the inquiry information; using a natural language model to convert the user's supplementary language into a structured supplementary language; and adjusting the structured process language according to the structured supplementary language to improve and supplement the structured process language. Specifically, after inputting the user's supplementary language, use a natural language model to convert the user's supplementary language into natural language text, and parse the obtained natural language text to obtain the structured supplementary language. According to the structured supplementary language, improve and supplement the structured process language. Among them, the structured supplementary language may include operation object supplementary information and / or object action supplementary information. According to the operation object supplementary information and the object action supplementary information, adjust and optimize the structured process language to improve and supplement the structured process language and ensure the information integrity of the structured process language.

[0065] In step S3 of this embodiment, the multimodal knowledge graph network has modal data of various related operating objects, that is, operating object modal data. The operating object modal data includes at least an image data set, a video data set, a model data set of the operating object, and other suitable related data of the operating object. Optionally, the operating object modal data may further include a text data set. The text data set is text description information related to the operation process, operating object, etc., and may include, for example, action descriptions of the operating object. Specifically, for example, the text data set records information such as the fuel valve needs to be rotated counterclockwise 9 times, the lubricating oil valve needs to be rotated counterclockwise 3 times to open, and turning the knob to the start position is to turn on the fan power supply. Using the image data set, video data set, and model data set of the operating object, various types of ship operation and maintenance manuals can be generated, realizing a more specific and accurate description of the ship operation and maintenance process, not only saving the time cost and material cost in the operation preparation process, but also reducing the construction difficulty and improving the work efficiency.

[0066] In an alternative embodiment, constructing the multimodal knowledge network graph in step S3 may include the following steps: building a multimodal knowledge graph network; obtaining the object name data of the operating object; according to the object name data, obtaining the image data set, video data set, and model data set of the operating object to form the operating object modal data; inputting the object name data and the operating object modal data into the multimodal knowledge graph network; and associating the object name data with the operating object modal data.

[0067] Among them, the object name data is used to identify the operating object to distinguish different operating objects. The image data set is the graphic-related data of the operating object, the video data set is the usage video and other related data of the operating object during the ship operation process, and the model data set is the design model (such as a 3D design model) and other related data of the operating object. In the multimodal knowledge graph network, the object name data is associated with the operating object modal data to facilitate subsequent obtaining of the target object modal data of the corresponding operating object from the multimodal knowledge graph network. The image data set, video data set, and model data set of the operating object can be extracted from the enterprise production management system, or manually input by the staff, or can also be obtained by other appropriate means.

[0068] In step S4 of this embodiment, the manual modal types of the ship operation and maintenance manual include at least one or several of video, graphic and electronic manual, animation, and document. The manual modal types can also include other appropriate manual types. Optionally, the manual modal types can be, for example, VR video, AR video, text document, graphic and electronic manual, etc.

[0069] In step S5, the target object model data includes target image data, target video data, and target modal data. According to the manual modal type, the object modal data is obtained from the operating object modal data. Specifically, for example, when the manual modal type is video, at least the target video data is obtained from the operating object modal data. When the manual modal type is a graphic and electronic manual, at least the target image data is obtained from the operating object modal data.

[0070] In an alternative embodiment, the structured process language includes operation object information. The step of performing step S5 to obtain the target object modal data from the operation object modal data includes: obtaining the target name data of the target object according to the operation object information; and respectively obtaining the target image data, target video data, and target modal data from the operation object modal data by using a multimodal knowledge graph network according to the target name data and the manual modal type, so as to form the target object modal data. Among them, the target name data of the target object can be obtained according to the operation object information, and the target image data, target video data, and target modal data can be respectively obtained from the image data set, video data set, and model data set. Specifically, taking the example of editing the starting operation process of a generator, a three-dimensional model of the fuel oil separator room, a cabin model, a control room model, and a video clip of opening a valve are retrieved by using a multimodal knowledge graph network.

[0071] In an alternative embodiment, the operation object modal data further includes a text data set of the operation object, the target object modal data further includes target text data, and the step of performing step S5 further includes the following steps: obtaining the target text data from the operation object modal data by using a multimodal knowledge graph network according to the target name data. Among them, the target text data can be obtained from the text data set of the operation object modal data, and the target text data is text description information related to the operation process, target object, etc.

[0072] In step S6, the obtained structured process language and target object modal data are integrally processed to form a ship operation and maintenance manual. Optionally, the structured process language includes a number of operation steps, and each operation step has operation object information and / or object action information. The step of performing step S6 may include the following steps: inserting the target image data, target video data, and target modal data into the number of operation steps according to the operation object information and object action information in each operation step, so as to form a ship operation and maintenance manual. Specifically, according to the logical order of the operation steps, the target graphic data, target video data, and target modal data corresponding to the logical step are inserted into the logical step in chronological order until the data insertion of all logical steps is completed, so as to form a ship operation and maintenance manual.

[0073] In an alternative embodiment, the operation object modal data further includes a text data set of the operation object, the target object modal data further includes target text data, and the step of performing step S6 further includes the following steps: adding the target text data to the number of operation steps according to the operation object information and object action information in each operation step, so as to form a ship operation and maintenance manual. Specifically, according to the logical order of the operation steps, a text description is added to the number of operation steps according to the target text data.

[0074] In an alternative embodiment, performing step S6 may further include the following steps: Detecting whether the operation object modal data has target video data according to the structured process language; when the result is "no", obtaining the target modal data from the operation object modal data; according to the target modal data and the target text data, calling an external editing software to produce an action video clip as the target video data. Specifically, when there is no inventory video clip of the target object in the multi-modal knowledge graph network, that is, there is no target video data in the operation object modal data, and the structured process language has operation object information, and the target video data cannot be obtained from the operation object modal data according to the operation object information. Then, the target modal data is extracted, and according to the operation requirements of the target object in the target text data, an action video clip is produced using the external editing software and the design model of the target object as the target video data, and inserted into the corresponding operation step to realize a more specific and accurate description of the ship operation and maintenance process, thereby reducing the construction difficulty and improving the work efficiency.

[0075] In an alternative embodiment, taking the editing of the generator starting operation process as an example, when the target video data of the fuel valve and the lubricating oil valve cannot be obtained through the multi-modal knowledge graph network, the target modal data of the fuel valve and the lubricating oil valve is obtained, and the target text data of the fuel valve and the lubricating oil valve is obtained by using the multi-modal knowledge graph network. The target text data records that the fuel valve needs to be rotated counterclockwise 9 times, and the lubricating oil valve needs to be rotated counterclockwise 3 times. Using the external editing software, the design models of the fuel valve and the lubricating oil valve are programmed to perform the above actions respectively to generate the corresponding action video clips, and then the target video data is obtained.

[0076] In summary, the production method of this embodiment can directly convert the user's language using the natural language model, quickly and effectively obtain the structured process language, without the need for staff to manually perform the structured conversion of natural language, effectively reducing the workload of the staff and improving the work efficiency of producing the ship operation and maintenance manual; the preparation method of this embodiment can also prepare various types of ship operation and maintenance manuals. When the target video data is missing in the inventory, the corresponding action video clip can be produced using the target modal data as the target video data, realizing a more specific and accurate description of the ship operation and maintenance process. The ship operation and maintenance manual obtained based on this production method can effectively save the time cost and material cost in the operation preparation link, and can also reduce the construction difficulty and improve the work efficiency. It can be seen that the production method of this embodiment effectively improves the work efficiency of producing the ship operation and maintenance manual by combining the natural language model with the multi-modal knowledge graph network, and based on the produced ship operation and maintenance manual, it can also reduce the construction difficulty, improve the construction efficiency, and save the time cost and material cost.

[0077] The above embodiments are only illustrative of the principles and effects of the present application and are not intended to limit the present application. Any person familiar with this technology can modify, change or combine the above embodiments without departing from the spirit and scope of the present application. Therefore, all equivalent modifications or changes made by those with ordinary knowledge in the technical field without departing from the spirit and technical idea disclosed in the present application should still be covered by the claims of the present application.

Claims

1. A method for producing a ship operation and maintenance manual, characterized in that It includes the following steps: S1. Construct a natural language model; S2. Input the user language and use the natural language model to convert the user language into a structured process language; S3. Construct a multimodal knowledge graph network, in which there is operation object modal data; S4. Select the manual modal type for generating the ship operation and maintenance manual; S5. According to the manual modal type, obtain the target object modal data from the operation object modal data; S6. According to the structured process language and the target object modal data, form a ship operation and maintenance manual.

2. The method for making a ship operation and maintenance manual according to claim 1, wherein The user language is a natural language, and the structured process language includes operation object information and object action information; The steps of converting the user language into a structured process language in step S2 include: Use the natural language model to convert the user language into natural language text; Parse the natural language text to obtain operation object information and object action information; According to the operation object information and the object action information, form the structured process language.

3. The method for producing a ship operation and maintenance manual according to claim 1, characterized in that After converting the user language into a structured process language, the following steps are further included: Perform a logical check on the structured process language to determine whether there are data defects in the structured process language; When there are data defects in the structured process language, generate and output inquiry information.

4. The method for producing a ship operation and maintenance manual according to claim 3, wherein After converting the user language into a structured process language, the following steps are further included: According to the inquiry information, input user supplementary language; Use the natural language model to convert the user supplementary language into a structured supplementary language; According to the structured supplementary language, adjust the structured process language.

5. The method for producing a ship operation and maintenance manual according to claim 1, characterized in that, Constructing a multimodal knowledge graph network includes the following steps: Build a multimodal knowledge graph network; Obtain the object name data of the operation object; According to the object name data, obtain the image data set, video data set, and model data set of the operation object to form operation object modal data; Input the object name data and the operation object modal data into the multimodal knowledge graph network; Associate the object name data with the operation object modal data.

6. The method for producing a ship operation and maintenance manual according to claim 1, wherein The manual modal type includes one or several of video, graphic and electronic manual, and document.

7. The method for producing a ship operation and maintenance manual according to claim 1, wherein The target object modal data includes target image data, target video data, and target modal data.

8. The method for producing a ship operation and maintenance manual according to claim 7, characterized in that, The structured process language includes several operation steps, and each operation step has operation object information and / or object action information; Performing step S6 includes the following steps: According to the operation object information and object action information in each operation step, insert the target image data, the target video data, and the target modal data into the several operation steps.

9. The method for producing a ship operation and maintenance manual according to claim 8, wherein, The operation object modal data further includes a text data set of the operation object, and the target object modal data further includes target text data; Performing step S6 further includes the following steps: According to the operation object information and object action information in each operation step, add the target text data to the several operation steps.

10. The method for producing a ship operation and maintenance manual according to claim 9, characterized in that, Performing step S6 further includes the following steps: Detecting whether there is target video data in the operation object modal data according to the structured process language; When the result is "no", obtaining target modal data from the operation object modal data; Calling an external editing software to produce an action video clip as the target video data according to the target modal data and the target text data.

Images