Artificial intelligence-based method, system, medium and equipment for designing man-hour unification

Abstract

The invention provides a man-hour unified design method, system, medium and equipment based on artificial intelligence, and the method comprises the steps: extracting related data from historical design man-hour filling to form a training database, constructing an artificial intelligence learning model according to the training database, and carrying out the man-hour design of a specific ship. And an artificial intelligence learning model is utilized to generate design tasks and distribution working hours, and plan dates of the corresponding tasks are scheduled according to manpower loads, so that intelligent unification of ship design working hours is realized. According to the method provided by the invention, the design task plan time can be automatically scheduled, manual participation is not needed, and the working efficiency is remarkably improved; mistakes easily occurring during manual unification are avoided, task splitting is more reasonable, and man-hour distribution is more accurate; according to the method, the design plan is directly arranged, design work can be arranged according to the human resource load condition, and intelligence of the ship design man-hour unified type is achieved.

Description

Technical Field

[0001] The present invention belongs to the technical field of shipbuilding, and specifically relates to a design time standardization method, system, medium and equipment based on artificial intelligence. Background Art

[0002] Ship design is a complex business involving many processes. It is divided into overall design, detailed design, production design, etc. according to the design stage, hull, piping, outfitting, electrical, etc. according to the design specialty, and superstructure, engine room, bow, stern, etc. according to the design area. In the process of splitting the design business and generating design hours, it is necessary to coordinate and allocate the design hours of each stage, each specialty, and each area according to the difficulty of the ship type, and it is necessary to consider the relationship between related design tasks before and after the design stage, inside and outside the design specialty, and inside and outside the design area.

[0003] The existing design time classification method is generally classified according to the above dimensions, and the design business is split according to the design experience. The design tasks are generated from top to bottom and the design time is allocated. This allocation method has many disadvantages: 1) The time division is difficult, and the allocation is very demanding. The allocator may not be clear about the design characteristics of each type of ship, and may not understand the design difficulty of each stage, each profession, and each area, so it may not be able to reasonably generate tasks according to the process; 2) It is difficult to accurately assign tasks and it is difficult to make timely adjustments to unreasonable feedback from top to bottom; 3) The simple generation of design tasks and time cannot be combined with human resources, and is not enough to meet the needs of production planning. Summary of the invention

[0004] In view of the shortcomings of the prior art described above, the present invention provides a method, system, medium and equipment for design work time standardization based on artificial intelligence. The method extracts relevant data from the historical design work time report to form a training database, builds an artificial intelligence learning model based on the training database, and uses the artificial intelligence learning model to generate design tasks and allocate work time when designing the work time of a specific ship, and schedules the corresponding tasks according to the manpower load, so as to realize the intelligent standardization of ship design work time. The method provided by the present application can automatically schedule the design task plan time without manual participation, and significantly improves work efficiency; it avoids the errors that are easy to occur during manual standardization, makes the task splitting more reasonable, and the work time distribution more accurate; the method directly schedules the design plan, which is conducive to arranging the design work according to the human resource load, and realizes the intelligent standardization of ship design work time.

[0005] To achieve the above-mentioned purpose and other related purposes, the present invention provides a design time standardization method based on artificial intelligence, comprising the following steps:

[0006] Establish structured feature templates and label templates for each ship type;

[0007] Filling historical data into the structured feature template and the label template to form a training database;

[0008] Building an artificial intelligence learning model based on the training database;

[0009] Inputting structured feature data of a specific ship type into the learning model to obtain first label data;

[0010] Iteratively optimizing the learning model according to the first label data until the output label data meets expectations, thereby obtaining second label data;

[0011] According to the second tag data, a design task plan and human resource allocation are formulated.

[0012] Optionally, the parameters in the structured feature template include ship type parameters and key design requirements.

[0013] Optionally, the parameters in the label template include design area, design stage, design specialty, and rated working hours.

[0014] Optionally, forming a training database includes:

[0015] Filling historical data in the structured feature template and the label template;

[0016] The data in the above template is cleaned to form a training database.

[0017] Optionally, the data cleaning includes: processing missing values, deleting duplicates, processing outliers, data format conversion, and data consistency checking.

[0018] Optionally, iteratively optimizing the learning model to obtain the second label data includes:

[0019] Correcting the first label data, and re-inputting the corrected label data into the training database;

[0020] Optimizing the learning model according to the updated training database;

[0021] The structured feature data of the specific ship type is re-input into the optimized learning model. If the output label data meets expectations, it is the second label data; otherwise, the above steps are repeated until the output label data meets expectations.

[0022] Optionally, develop a design task plan and human resource allocation including:

[0023] Link design tasks to personnel positions;

[0024] Formulate a single ship design task plan based on the single ship design delivery nodes and design business sequencing;

[0025] Based on the available human resources for corresponding positions within the design cycle, calculate the task start time, task end time, and number of people to be deployed.

[0026] The present invention also provides a system for the design time standardization method based on any one of the above-mentioned artificial intelligence, comprising:

[0027] Data sorting module, used to extract structured feature data, label data, and perform data cleaning from historical design tasks and reported working hours;

[0028] A data storage module, used for storing the extracted structured feature data and the label data;

[0029] Learning and training module, used for algorithm selection, model training, model verification, and model testing for artificial intelligence learning;

[0030] A data output module is used to output the first label data.

[0031] The present invention also provides a computer-readable storage medium, wherein the computer-readable storage medium stores a computer program, and when the computer program runs on a processor, the processor executes any of the above methods.

[0032] The present invention also provides an electronic device, comprising:

[0033] at least one memory for storing a program;

[0034] At least one processor is used to execute the program stored in the memory. When the program stored in the memory is executed, the processor is used to execute any of the methods described above.

[0035] The artificial intelligence-based design time standardization method, system, medium and device provided by the present invention have at least the following beneficial effects:

[0036] The method provided in this application can automatically schedule the design task plan time without human intervention, and work efficiency is significantly improved; it avoids the errors that are easy to occur during manual planning, makes the task splitting more reasonable, and the working time distribution more accurate; this method directly schedules the design plan, which is conducive to arranging the design work according to the human resource load and realizing the intelligent planning of ship design working hours.

[0037] The system, medium and device provided in this application are formed based on the above method and also have the above beneficial effects. BRIEF DESCRIPTION OF THE DRAWINGS

[0038] Figure 1Shown is a flow chart of the artificial intelligence-based design time standardization method provided in Example 1. DETAILED DESCRIPTION

[0039] 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.

[0040] It should be noted that the illustrations provided in this embodiment only illustrate the basic concept of the present invention in a schematic manner. Although 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, the form, quantity, positional relationship and proportion of each component in actual implementation can be changed at will under the premise of realizing the technical solution of this party, and the component layout form may also be more complicated.

[0041] Embodiment 1

[0042] This embodiment provides a design time standardization method based on artificial intelligence, such as Figure 1 As shown, the following steps are included:

[0043] Step S1: Establish structured feature templates and label templates for each ship type;

[0044] As an example, the parameters in the structured feature template include ship type parameters (such as container ships, liquefied gas ships), key design requirements (such as tonnage, quantity), which provide a basic framework and standards for ship design to ensure that the design meets specific specifications and requirements. As an example, the parameters in the label template include design area (such as engine room area, superstructure area), design stage (such as detailed design, production design), design specialty (such as electrical equipment, hull), and rated working hours. These parameters are used to clarify the specific ownership and schedule of design tasks, which facilitates the management and coordination of the entire design process.

[0045] Step S2: filling historical data into the structured feature template and the label template to form a training database;

[0046] First, the historical data is reported in the above structured feature template and label template. As an example, based on the actual reported design hours of each ship type and the design requirements of the ship type itself, the structured feature data and label data are extracted and reported in the structured feature template and label template, as shown in Table 1 and Table 2.

[0047]

[0048] Table 1 Structured feature data

[0049]

[0050] Table 2 Label data

[0051] Next, the data in the above template is cleaned to form a training database. As an example, data cleaning includes: 1) processing missing values, for example, directly deleting rows or columns containing missing values; filling missing values ​​through statistical methods (such as mean, median, mode) or prediction models (such as regression, kNN); treating missing values ​​as part of the data distribution law, processing missing values ​​through transformation methods, so that they can participate in subsequent calculations, deleting duplicates, processing outliers, data format conversion, and data consistency checks; 2) deleting duplicates, which may be due to data collection errors, data merging operations, system problems or other reasons. These duplicates need to be identified and deleted to ensure the uniqueness of data in the data warehouse; 3) processing outliers, which refers to data with large deviations, which can be identified and processed through statistical models (such as the Laida criterion, the Dixon criterion, the Grubbs criterion, the T test, etc.); 4) data format conversion, converting data in different formats into a unified format for subsequent analysis and reporting; 5) data consistency check to ensure that data metrics between different systems and entities are consistent.

[0052] Next, after completing data cleaning, screening and format conversion, the processed data is stored in the database. After integration and optimization, a high-quality training database is finally formed to provide reliable data support for subsequent model training.

[0053] Step S3: constructing an artificial intelligence learning model based on the training database;

[0054] First, choose a suitable machine learning algorithm. For example, you need to consider the characteristics of the data, the type of problem, and the expected model performance. For example, for classification problems, you can consider logistic regression, support vector machine, or decision tree; for regression problems, linear regression or random forest may be a better choice; and for complex nonlinear relationships, deep learning algorithms such as neural networks may be more suitable; in addition, you also need to consider the complexity of the algorithm, training time, and interpretability of the model to ensure that the algorithm finally selected can solve practical problems efficiently and accurately.

[0055] Next, the model is trained based on the training database obtained in step S2. Since different ship types have their own unique design features and structural requirements, for example, liquefied gas ships have liquid tank areas, while container ships have unique shelf areas, when training the model, the corresponding training data should be used to construct the machine learning model according to the different ship types. In addition, even some common areas between ship types, such as the upper construction area, have different design difficulties depending on the ship type. Therefore, in order to improve the accuracy and applicability of the model, it is necessary to classify the training data according to the ship type and form targeted artificial intelligence learning models.

[0056] Step S4: inputting structured feature data of a specific ship type into the learning model to obtain first label data;

[0057] As an example, the structured feature data of a specific ship type is input into the learning model obtained in step S3, as shown in Table 1, including, for example, ship type parameters (such as container ship, liquefied gas ship), key design requirements (such as tonnage, quantity), and the learning model can output corresponding label data, as shown in Table 2, including, for example, design area (such as engine room area, superstructure area), design stage (such as detailed design, production design), design specialty (such as electrical equipment, hull), and rated working hours, which are recorded as the first label data.

[0058] Step S5: iteratively optimizing the learning model according to the first label data until the output label data meets expectations, thereby obtaining second label data;

[0059] As an example, in order to improve the accuracy and reliability of artificial intelligence learning models, the learning models need to be optimized and verified.

[0060] First, the first label data outputted in step S4 may contain errors or deviations, and therefore needs to be corrected by professionals to ensure its accuracy and completeness. The corrected label data is re-entered into the training database to update and enrich the content of the database.

[0061] Next, the AI ​​learning model is optimized based on the updated training database. This process aims to use more accurate data to adjust the model's parameters, thereby improving the model's performance and predictive capabilities.

[0062] Finally, the structured feature data of the specific ship type is re-input into the optimized artificial intelligence learning model for verification. If the output label data meets expectations, then the label data is regarded as the second label data, indicating that the model optimization is successful; conversely, if the output result does not meet expectations, repeat the above steps, that is, re-correct the data, update the database, optimize the model, and verify again until the output label data is completely in line with expectations.

[0063] Through the above-mentioned iterative process, it is ensured that the artificial intelligence learning model can be continuously improved, ultimately achieving efficient and accurate prediction results.

[0064] Step S6: Formulate a design task plan and human resource allocation based on the second tag data.

[0065] First, associate the design tasks with the personnel positions. As an example, the personnel positions are first classified according to the "design stage" in Table 2, and each design stage includes multiple personnel positions; then the design tasks are associated with the personnel positions. In this embodiment, the design tasks and the personnel positions are in a many-to-one relationship, that is, multiple design tasks can be completed by one personnel position, for example, design tasks A and B can both be completed by personnel position A.

[0066] Next, a single ship design task plan is developed based on the design delivery nodes and design business sequencing of the single ship. For example, the design business sequencing includes but is not limited to: the production design of the same area and section is arranged after the detailed design and before the on-site construction; there are also design sequence requirements within the detailed design and production design, such as hull design before outfitting design, etc. As an example, when developing a single ship design task plan, the rated working hours for the design task are 8 hours per person per day, and the required manpower is calculated based on this.

[0067] Finally, the task start time, task end time, and number of personnel are calculated based on the available human resources of the corresponding positions within the design cycle. As an example, the above single ship design task plan is integrated into the already scheduled design tasks of other ship types, and the remaining available human resources of the corresponding positions within the design cycle are combined to calculate the exact start time, end time, and number of personnel to complete the task.

[0068] Embodiment 2

[0069] This embodiment provides a design time prototyping system based on artificial intelligence, which is used to perform the design time prototyping method based on artificial intelligence provided in Example 1, including a data sorting module, a data storage module, a learning and training module, and a data output module.

[0070] As an example, the data sorting module is used to extract structured feature data, label data, and perform data cleaning from historical design tasks and reported working hours. Specifically, this module can mine key information from historical design tasks, extract structured feature data and label data, provide strong support for subsequent data analysis and model building, and provide an important basis for working hour forecasting and task management; in addition, after the data extraction is completed, the data sorting module will clean the extracted data, remove noise data, duplicate data, and missing data, ensure the accuracy and completeness of the data, and provide a high-quality data foundation for subsequent data analysis and application.

[0071] As an example, the data storage module is used to store the extracted structured feature data and label data. Specifically, the above data is stored in the form of tables or relational databases to facilitate subsequent query, analysis and processing; in order to ensure the security and reliability of the data, the data storage module adopts advanced data backup and recovery mechanisms to regularly back up the stored data to prevent data loss or damage.

[0072] As an example, the learning and training module is used for algorithm selection, model training, model verification, and model testing of artificial intelligence learning. Specifically, it provides users with a rich algorithm library covering multiple types such as supervised learning, unsupervised learning, and deep learning. Users can flexibly choose according to task requirements and data characteristics; in the model training stage, the module uses structured feature data and label data, and through automated parameter optimization and iterative training, the model gradually adapts to the data and improves performance; in the model verification stage, cross-validation and other methods are used to comprehensively evaluate the generalization ability and performance indicators of the model; in the model testing stage, an independent test data set is used to conduct rigorous performance testing of the model, and a detailed test report is generated to ensure that the model has high accuracy and reliability in practical applications.

[0073] As an example, the data output module is used to output the first label data. In order to ensure the accuracy and security of the data, the data output module will perform strict data verification and encryption during the output process, and will also record the output log to facilitate users to trace the flow and use of data. Through these functions, the data output module can not only efficiently complete the data output task, but also provide users with reliable and secure data support.

[0074] Embodiment 3

[0075] The present embodiment provides a computer-readable storage medium, on which a program of a design time prototyping method based on artificial intelligence is stored, so that the design time prototyping method based on artificial intelligence provided in Embodiment 1 is executed. As for the aforementioned computer-readable storage medium, those skilled in the art can understand that the implementation example of the functions of the aforementioned system and each unit can be completed by hardware related to the computer program. The aforementioned computer program can be stored in a computer-readable storage medium. When the program is executed, it executes an implementation example including the functions of the aforementioned system and each unit. The aforementioned storage medium includes: various media that can store program codes, such as ROM, RAM or optical disk.

[0076] Embodiment 4

[0077] The present embodiment provides an electronic device, comprising: at least one memory for storing programs; at least one processor for executing the programs stored in the memory. The electronic device may be a server. The electronic device comprises a processor, a memory, a network interface and a database connected via a system bus. Among them, the processor of the electronic device is used to provide computing and control capabilities. The memory of the electronic device comprises a non-volatile and / or volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program and a database. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The network interface of the electronic device is used to communicate with an external client via a network connection. When the computer program is executed by the processor, the design time standardization method based on artificial intelligence provided in Example 1 is implemented.

[0078] 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 design time standardization method based on artificial intelligence, characterized in that: The steps include: Establish structured feature templates and label templates for each ship type; Filling historical data into the structured feature template and the label template to form a training database; Building an artificial intelligence learning model based on the training database; Inputting structured feature data of a specific ship type into the learning model to obtain first label data; Iteratively optimizing the learning model according to the first label data until the output label data meets expectations, thereby obtaining second label data; According to the second tag data, a design task plan and human resource allocation are formulated.

2. The design time standardization method based on artificial intelligence according to claim 1 is characterized in that: The parameters in the structured feature template include ship type parameters and key design requirements.

3. The design time standardization method based on artificial intelligence according to claim 1 is characterized in that: The parameters in the label template include design area, design stage, design specialty, and rated working hours.

4. The design time standardization method based on artificial intelligence according to claim 1 is characterized in that: The training database includes: Filling historical data in the structured feature template and the label template; The data in the above template is cleaned to form a training database.

5. The design time standardization method based on artificial intelligence according to claim 4 is characterized in that: The data cleaning includes: processing missing values, deleting duplicates, processing outliers, data format conversion, and data consistency checking.

6. The design time standardization method based on artificial intelligence according to claim 1 is characterized in that: The learning model is iteratively optimized to obtain the second label data including: Correcting the first label data, and re-inputting the corrected label data into the training database; Optimizing the learning model according to the updated training database; The structured feature data of the specific ship type is re-input into the optimized learning model. If the output label data meets expectations, it is the second label data; otherwise, the above steps are repeated until the output label data meets expectations.

7. The design time standardization method based on artificial intelligence according to claim 1 is characterized in that: Develop design task plans and human resource allocation including: Link design tasks to personnel positions; Formulate a single ship design task plan based on the single ship design delivery nodes and design business sequencing; Based on the available human resources for corresponding positions within the design cycle, calculate the task start time, task end time, and number of people to be deployed.

8. A system for design time standardization method based on artificial intelligence according to any one of claims 1 to 7, characterized in that: include: Data sorting module, used to extract structured feature data, label data, and perform data cleaning from historical design tasks and reported working hours; A data storage module, used for storing the extracted structured feature data and the label data; Learning and training module, used for algorithm selection, model training, model verification, and model testing for artificial intelligence learning; A data output module is used to output the first label data.

9. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, and when the computer program is executed on a processor, the processor is caused to execute the method according to any one of claims 1 to 7.

10. An electronic device, characterized in that: include: at least one memory for storing a program; At least one processor is used to execute the program stored in the memory, and when the program stored in the memory is executed, the processor is used to execute the method according to any one of claims 1 to 7.

Images