Ship hull hydrodynamic configuration optimization design process architecture based on artificial intelligence technology

Abstract

This invention relates to a process architecture for optimizing ship hydrodynamic configurations based on artificial intelligence technology. The architecture includes modules such as expert experience, historical sample data of ship hydrodynamic configurations, design input for ship hydrodynamic configurations, a knowledge base for ship hydrodynamic configuration design and evaluation, a parametric model of ship hydrodynamic configurations, knowledge-driven reconstruction of the ship hydrodynamic model, AI-based optimization design of ship hydrodynamic configurations, rapid evaluation of ship hydrodynamic performance, and a comprehensive evaluation index model for ship hydrodynamic performance. This invention fully leverages the design experience contained in historical sample data, extracting the experience of numerous experts into design knowledge and integrating it into the ship hydrodynamic configuration optimization design process. The process architecture drives intelligent reconstruction of the ship hydrodynamic model through design knowledge, and improves upon the shortcomings of traditional ship hydrodynamic performance prediction methods, such as time consumption and low efficiency, by establishing a rapid prediction proxy model for hydrodynamic performance based on machine learning algorithms.

Description

TECHNICAL FIELD

[0001] The present application relates to a process architecture of hull hydrodynamic configuration optimization design based on artificial intelligence technology, in particular to a process architecture design method suitable for developing hull hydrodynamic configuration intelligent optimization design software or related optimization design platform. BACKGROUND

[0002] The traditional hull hydrodynamic configuration design relies on the transformation of the parent ship and the experience of the designer, and with the aid of CAD, CAE and CFD tools, the performance of the hull hydrodynamic configuration scheme is evaluated, and whether to further optimize the design of the hull hydrodynamic configuration is determined according to the comparison of the performance evaluation results and the technical indicators. This is iterated until the scheme meets the technical indicator requirements. The above-mentioned hull hydrodynamic configuration optimization design mode mainly relies on the experience and level of the designer, which is low in efficiency, time-consuming and labor-intensive. The level of hull hydrodynamic configuration is limited by the experience of the current designer and team, and it is impossible to absorb and utilize the excellent hull hydrodynamic configuration experience in history vertically, and it is also difficult to learn from the successful experience and failure lessons of other design teams horizontally. The designer's brain cannot consider the matching and trade-off between multiple dimensional hull hydrodynamic configuration design constraints and multiple hydrodynamic performance indicators, which limits the improvement of the level of hull hydrodynamic configuration. In addition, the large number of process schemes and final schemes of hydrodynamic configuration optimization are scattered in the hands of various design teams, which are isolated information islands, affecting the basis of historical sample data mining analysis.

[0003] The traditional hull hydrodynamic configuration optimization design mode does not fully utilize the historical design sample data and expert experience, and the optimization design knowledge contained in the historical samples has not been fully mined, refined and utilized. The design and evaluation knowledge formed by expert experience and historical sample data mining and refinement has not been applied in the hull hydrodynamic configuration optimization design decision. The hull hydrodynamic performance prediction means is usually numerical simulation or model test, which takes a long time and affects the efficiency of hull hydrodynamic configuration optimization design. The comprehensive evaluation index model of hydrodynamic performance is not integrated into the optimization design process, and it is impossible to judge the matching and trade-off between multiple performance indicators in time. The hull hydrodynamic configuration optimization design needs to switch between multiple independent functional modules, which is time-consuming and labor-intensive, affecting the efficiency and quality of the optimization design. In addition, the traditional hull hydrodynamic configuration optimization design process only focuses on the final scheme that meets the index requirements, and a large number of process schemes are not valued. The optimization design knowledge and strategies contained in the process schemes have not been subjected to mining analysis and utilization process conventions.

[0004] With the development of computer computing power and artificial intelligence technology, an opportunity is provided to solve the above problems. Artificial intelligence technologies such as machine learning and data mining analysis are used to process and analyze historical sample data of ship hull hydrodynamic configuration and expert experience to form design knowledge and evaluation knowledge for guiding the design of hydrodynamic configuration. Meanwhile, parameterization of the ship hull hydrodynamic configuration model, a fast prediction method of the hydrodynamic performance of the ship hull configuration, and a comprehensive evaluation index model of the hydrodynamic performance of the ship hull are combined into the ship hull hydrodynamic configuration optimization design process based on artificial intelligence technology. A process architecture of the ship hull hydrodynamic configuration optimization design based on artificial intelligence technology is proposed, which is different from the traditional ship hull hydrodynamic configuration optimization design mode. SUMMARY

[0005] Therefore, the application provides a process architecture of ship hull hydrodynamic configuration optimization design based on artificial intelligence technology.

[0006] The application adopts the following technical scheme:

[0007] The ship hull hydrodynamic configuration optimization design process architecture based on artificial intelligence technology includes the following modules: expert experience, ship hull hydrodynamic configuration historical sample data, ship hull hydrodynamic configuration design input, ship hull hydrodynamic configuration design knowledge and evaluation knowledge base, ship hull hydrodynamic configuration parameterization model, knowledge-driven ship hull hydrodynamic model reconstruction, ship hull hydrodynamic configuration optimization design based on artificial intelligence technology, ship hull hydrodynamic performance rapid evaluation, and ship hull hydrodynamic performance comprehensive evaluation index model. The ship hull hydrodynamic configuration design knowledge and evaluation knowledge base are jointly constructed by the knowledge formed through expert experience and ship hull hydrodynamic configuration historical sample data mining analysis. The process of the ship hull hydrodynamic configuration optimization design based on artificial intelligence technology starts from the ship hull hydrodynamic configuration design input. According to the design input parameters, the artificial intelligence engine correlates and matches the design requirements with the ship hull hydrodynamic configuration design knowledge base for analysis, forms the process of driving the ship hull hydrodynamic configuration parameterization model construction, generates the parameterization model based on the design knowledge, and enters the optimization space constituted by the ship hull hydrodynamic configuration parameterization model. The ship hull hydrodynamic configuration design optimization direction is calculated in the optimization space through the artificial intelligence algorithm, and the configuration scheme meeting the technical index requirements is explored. After each ship hull hydrodynamic configuration scheme is generated, the resistance, propulsion, wave resistance, seakeeping performance, maneuvering performance, and flow field performance are rapidly predicted through the application of the ship hull hydrodynamic performance rapid prediction module. On one hand, the ship hull hydrodynamic performance prediction results are stored and updated into the ship hull hydrodynamic configuration historical sample database together with the parameters and geometric information of the configuration scheme. On the other hand, the ship hull hydrodynamic performance prediction results are transmitted to the ship hull hydrodynamic performance comprehensive evaluation index module. According to the scene in the design input, the algorithm selects the appropriate evaluation index model to evaluate and analyze the comprehensive performance index of the ship hull hydrodynamic configuration scheme. When all the index requirements are met, the ship hull hydrodynamic configuration scheme meeting the requirements is output. If the index does not meet the requirements, the evaluation conclusions given by the evaluation index model and the performance and flow field characteristics of the ship hull hydrodynamic configuration scheme are comprehensively considered. Through the deep learning algorithm, the further optimization direction of the hydrodynamic configuration is learned and explored in the design knowledge and evaluation knowledge base. The knowledge driving the ship hull hydrodynamic model reconstruction is formed, the reconstruction of the knowledge-driven ship hull hydrodynamic parameterization model is completed, and the reconstructed hydrodynamic parameterization model combined with the expert experience and historical sample data experience knowledge starts a new round of intelligent optimization design. The new ship hull hydrodynamic configuration scheme generated by the optimization is iteratively calculated and analyzed until the optimization design obtains the scheme meeting the technical index requirements.

[0008] Further, the expert experience includes successful cases and failed lessons of ship hull hydrodynamic configuration and optimization design knowledge accumulated in practice.

[0009] Further, the expert experience adopts a process between the expert experience and the ship hull hydrodynamic configuration design knowledge and evaluation knowledge base to form a systematic ship hull hydrodynamic configuration optimization design knowledge base from the tangible, or intangible, or rough form experience summarized and refined in the long-term practice of experts through knowledge engineering and semantic technology.

[0010] Further, the ship hull hydrodynamic configuration historical sample data includes test data and simulation data of the ship hull hydrodynamic configuration.

[0011] Further, the test data and simulation data of the ship hull hydrodynamic configuration include ship hull hydrodynamic configuration geometric parameters and performance data.

[0012] Further, the ship hull hydrodynamic configuration historical sample data adopts a process between the ship hull hydrodynamic configuration historical sample data and the ship hull hydrodynamic configuration design knowledge and evaluation knowledge base to establish a multi-dimensional, multi-granularity mapping and association between the ship hull hydrodynamic configuration parameters and performance, and between the data, information, and knowledge fragments through a machine learning algorithm, and the expert knowledge formed by the process between the expert experience and the ship hull hydrodynamic configuration design knowledge and evaluation knowledge base together constitute the ship hull hydrodynamic configuration design knowledge and evaluation knowledge base.

[0013] Further, the process between the expert experience and the ship hull hydrodynamic configuration design knowledge and evaluation knowledge base and the process between the ship hull hydrodynamic configuration historical sample data and the ship hull hydrodynamic configuration design knowledge and evaluation knowledge base are previously based on the existing ship hull hydrodynamic configuration historical sample data and expert experience to construct a basic knowledge base of hydrodynamic configuration optimization design through multivariate statistical learning, support vector machine, deep learning, knowledge engineering, and other machine learning algorithms, and continuously updated through intelligent technology of continuous self-learning.

[0014] Further, the ship hull hydrodynamic configuration parameters, geometric information, and performance data generated by each optimization are saved to the historical sample database through the process, and when the cumulative sample data reaches a certain amount, a data mining algorithm is triggered to analyze and mine the historical sample data for a new round of analysis and mining, and the newly mined or adjusted ship hull hydrodynamic configuration design knowledge and evaluation knowledge is updated to the ship hull hydrodynamic configuration design knowledge and evaluation knowledge base, and the hydrodynamic configuration level is automatically upgraded.

[0015] Further, the ship hull hydrodynamic performance rapid prediction is a proxy model of the ship hull hydrodynamic performance rapid prediction constructed through the training and parameter adjustment of the machine learning algorithm based on a large number of simulation data sets and test data sets of different ship types.

[0016] Further, the ship hull hydrodynamic performance rapid prediction includes resistance performance, propulsion performance, wave-induced resistance, seakeeping performance, maneuvering performance, and analysis and evaluation of flow field characteristics.

[0017] The beneficial effects of the present application are:

[0018] The present application can fully tap the design experience contained in historical sample data, refine the experience of numerous experts into design knowledge, and integrate the design knowledge into the ship hull hydrodynamic configuration optimization design process. The process architecture drives the intelligent reconstruction of the ship hull hydrodynamic model through design knowledge. The process architecture improves and improves the time-consuming and low-efficiency defects of the traditional ship hull hydrodynamic performance prediction method by establishing a machine learning algorithm-based hydrodynamic performance rapid prediction proxy model. The process architecture collects a large amount of intermediate scheme and final scheme information in the optimization process into the historical sample library in a timely manner, expands the capacity of the sample, and regularly mines and updates the design evaluation knowledge base according to the corresponding strategy through intelligent algorithms. At the same time, the process architecture integrates a ship hull hydrodynamic performance comprehensive evaluation index model into the optimization design process, overcoming the difficulty that the traditional optimization process can only judge a single target. The ship hull hydrodynamic configuration optimization design process architecture based on artificial intelligence technology is more targeted and has higher optimization efficiency. BRIEF DESCRIPTION OF DRAWINGS

[0019] Figure 1 The present application is a principle diagram. DETAILED DESCRIPTION

[0020] The technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor fall within the scope of protection of the present application.

[0021] As shown in Figure 1 The process architecture of the ship hull hydrodynamic configuration optimization design based on artificial intelligence technology provided by the embodiments of the present application includes the modules of expert experience, ship hull hydrodynamic configuration historical sample data, ship hull hydrodynamic configuration design input, ship hull hydrodynamic configuration design knowledge and evaluation knowledge base, ship hull hydrodynamic configuration parameterization model, knowledge-driven ship hull hydrodynamic model reconstruction, ship hull hydrodynamic configuration optimization design based on artificial intelligence technology, ship hull hydrodynamic performance rapid evaluation, and ship hull hydrodynamic performance comprehensive evaluation index model.

[0022] Figure 1 The present application is a principle diagram, which shows the process architecture of the ship hull hydrodynamic configuration optimization design based on artificial intelligence technology.

[0023] The tangible, or intangible, or rough form experience summarized and refined by the expert long-term practice is formed into a systematic ship hull hydrodynamic configuration optimization design knowledge base through knowledge engineering and semantic technology in Process 1; the multi-source and heterogeneous ship hull hydrodynamic configuration historical sample data is established through machine learning algorithm to form the mapping association between the ship hull hydrodynamic configuration parameters and performance, between the data, information and knowledge fragments in multiple dimensions and multiple granularities in Process 2, and the expert knowledge formed in Process 1 together constitute the ship hull hydrodynamic configuration design knowledge and evaluation knowledge base; Process 1 and Process 2 form the basic knowledge base of the hydrodynamic configuration optimization design in advance based on the existing ship hull hydrodynamic configuration historical sample data and expert experience through machine learning algorithms such as multivariate statistical learning, support vector machine, deep learning and knowledge engineering, and the basic knowledge base is continuously updated through intelligent technology of continuous self-learning.

[0024] The ship hull hydrodynamic configuration optimization design process based on artificial intelligence technology starts from 3, according to the design input parameters, the artificial intelligence engine analyzes the design requirements and the hydrodynamic configuration design knowledge base, forms Process 4 to guide the construction of the ship hull hydrodynamic configuration parameterized model, and the design knowledge drives the generation of the parameterized model.

[0025] After entering the optimization space constituted by the parametric model through process 5, the direction of optimization of the hydrodynamic configuration design is calculated in the optimization space by an artificial intelligence algorithm, and a configuration scheme that meets the technical index requirements is explored. After each hydrodynamic configuration scheme is generated, process 6 is applied to apply a ship hydrodynamic performance rapid prediction module to rapidly predict the resistance, propulsion, wave added resistance, seakeeping performance, maneuvering performance, and flow field performance. On the one hand, the hydrodynamic performance prediction results are stored and updated to the ship hydrodynamic configuration historical sample database through process 7, together with the parameters, geometric information of the configuration scheme. When the cumulative sample data reaches a certain amount, a data mining algorithm is triggered to analyze and mine the historical sample data, and the newly mined or adjusted hydrodynamic configuration design knowledge and evaluation knowledge are updated to the knowledge base. On the other hand, the hydrodynamic performance prediction results enter the ship hydrodynamic performance comprehensive evaluation index module through process 8, and the algorithm selects the appropriate evaluation index model according to the scene in the design input to evaluate and analyze the comprehensive performance index of the hydrodynamic configuration scheme. When all the indexes meet the requirements, the ship hydrodynamic configuration scheme that meets the requirements is output through process 9. When the indexes do not meet the requirements, the performance of the hydrodynamic configuration scheme, the flow field characteristics, and the evaluation conclusions given by the evaluation index model are comprehensively analyzed through process 10, and a deep learning algorithm is used to learn and explore the further optimization direction of the hydrodynamic configuration in the design knowledge and evaluation knowledge base, and form the knowledge that drives the reconstruction of the ship hydrodynamic model. The knowledge-driven reconstruction of the ship hydrodynamic parametric model is completed through process 11, and the reconstructed hydrodynamic parametric model combined with the expert experience and historical sample data experience knowledge starts a new round of intelligent optimization design through process 12. The new hydrodynamic configuration scheme generated by optimization starts iterative calculation and analysis through process 6 until the optimization design obtains a scheme that meets the technical index requirements.

[0026] The ship hydrodynamic performance rapid prediction module is a proxy model for rapid prediction of hydrodynamic performance based on experimental data sets and simulation data sets, and obtained by adjusting and training the hydrodynamic performance rapid prediction algorithm through artificial intelligence.

Claims

1. A hull hydrodynamic configuration optimization design method based on artificial intelligence technology, characterized in that: The system comprises expert experience, hull hydrodynamic configuration historical sample data, hull hydrodynamic configuration design input, hull hydrodynamic configuration design knowledge and evaluation knowledge base, hull hydrodynamic configuration parameterization model, knowledge-driven hull hydrodynamic model, hull hydrodynamic configuration optimization design based on artificial intelligence technology, hull hydrodynamic performance rapid evaluation, and hull hydrodynamic performance comprehensive evaluation index module. The hull hydrodynamic configuration design knowledge and evaluation knowledge base are formed by expert experience and hull hydrodynamic configuration historical sample data mining analysis. The process of the hull hydrodynamic configuration optimization design based on artificial intelligence technology starts from the hull hydrodynamic configuration design input. According to the design input parameters, the artificial intelligence engine matches and analyzes the design requirements and the hull hydrodynamic configuration design knowledge base to form the process of driving the hull hydrodynamic configuration parameterization model construction. The design knowledge drives the parameterization model, and enters the optimization space formed by the hull hydrodynamic configuration parameterization model. The hull hydrodynamic configuration design optimization direction is calculated in the optimization space through the artificial intelligence algorithm, and the configuration scheme meeting the technical index requirements is explored. After each hull hydrodynamic configuration scheme is generated, the resistance, propulsion, wave resistance, seakeeping performance, maneuvering performance and flow field performance are rapidly predicted through the application of the hull hydrodynamic performance rapid evaluation module. On one hand, the hull hydrodynamic performance prediction results are stored and updated into the hull hydrodynamic configuration historical sample database together with the parameters and geometric information of the configuration scheme. On the other hand, the hull hydrodynamic performance prediction results are transmitted to the hull hydrodynamic performance comprehensive evaluation index module. According to the scene in the design input, the algorithm selects the appropriate evaluation index model to evaluate and analyze the comprehensive performance index of the hull hydrodynamic configuration scheme. When all the index requirements are met, the hull hydrodynamic configuration scheme meeting the requirements is output. If the index does not meet the requirements, the evaluation conclusions of the hull hydrodynamic configuration scheme in various performances, flow field characteristics and evaluation index models are comprehensively considered. Through the deep learning algorithm, the further optimization direction of the water dynamic configuration is learned and explored in the design knowledge and evaluation knowledge base. The knowledge driving the hull hydrodynamic model reconstruction is formed. The reconstruction of the knowledge-driven hull hydrodynamic parameterization model is completed. The hull hydrodynamic parameterization model reconstructed by combining the expert experience and historical sample data experience knowledge starts a new round of intelligent optimization design. The new hull hydrodynamic configuration scheme generated by optimization is iteratively calculated and analyzed until the optimization design obtains a scheme meeting the technical index requirements. 2.The ship hull hydrodynamic configuration optimization design method based on artificial intelligence technology according to claim 1, characterized in that: The expert experience comprises successful cases and failed lessons of hull hydrodynamic configuration and optimization design knowledge accumulated in practice.

3. The ship hull hydrodynamic configuration optimization design method based on artificial intelligence technology according to claim 2, characterized in that: The expert experience adopts the process between the expert experience and the hull hydrodynamic configuration design knowledge and evaluation knowledge base to form the systematic hull hydrodynamic configuration optimization design knowledge base through knowledge engineering and semantic technology from the tangible, intangible or rough experience summarized and refined in the long-term practice of experts. 4.The ship hull hydrodynamic configuration optimization design method based on artificial intelligence technology according to claim 1, characterized in that: The hull hydrodynamic configuration historical sample data comprises test data and simulation data of hull hydrodynamic configuration.

5. The ship hull hydrodynamic configuration optimization design method based on artificial intelligence technology according to claim 4, characterized in that: The test data and simulation data of the ship hull hydrodynamic configuration include ship hull hydrodynamic configuration geometric parameters and performance data. 6.The ship hull hydrodynamic configuration optimization design method based on artificial intelligence technology according to claim 1, characterized in that: The ship hull hydrodynamic configuration historical sample data adopts a process between ship hull hydrodynamic configuration historical sample data and a ship hull hydrodynamic configuration design knowledge and evaluation knowledge base to establish a multi-dimensional, multi-granularity mapping relationship between ship hull hydrodynamic configuration parameters and performance, between data, between information, and between knowledge fragments of the ship hull hydrodynamic configuration historical sample data through a machine learning algorithm, and the expert knowledge formed by a process between expert experience and the ship hull hydrodynamic configuration design knowledge and evaluation knowledge base to jointly constitute the ship hull hydrodynamic configuration design knowledge and evaluation knowledge base.

7. The ship hull hydrodynamic configuration optimization design method based on artificial intelligence technology according to claim 6, characterized in that: The process between the expert experience and the ship hull hydrodynamic configuration design knowledge and evaluation knowledge base and the process between the ship hull hydrodynamic configuration historical sample data and the ship hull hydrodynamic configuration design knowledge and evaluation knowledge base are previously based on existing ship hull hydrodynamic configuration historical sample data and expert experience to construct a basic knowledge base of hydrodynamic configuration optimization design through multivariate statistical learning, support vector machines, deep learning, and knowledge engineering machine learning algorithms, and are continuously updated through intelligent technology of continuous self-learning. 8.The ship hull hydrodynamic configuration optimization design method based on artificial intelligence technology according to claim 1, characterized in that: The ship hull hydrodynamic configuration parameters, geometric information, and performance data generated by each optimization are saved to the historical sample database through the process, and when the accumulated sample data reaches a certain amount, a data mining algorithm is triggered to analyze and mine the historical sample data for a new round of analysis and mining, the newly mined or adjusted ship hull hydrodynamic configuration design knowledge and evaluation knowledge is updated to the ship hull hydrodynamic configuration design knowledge and evaluation knowledge base, and the hydrodynamic configuration level is automatically upgraded. 9.The ship hull hydrodynamic configuration optimization design method based on artificial intelligence technology according to claim 1, characterized in that: The ship hull hydrodynamic performance rapid prediction is a proxy model of the ship hull hydrodynamic performance rapid prediction constructed through training and parameter adjustment of a machine learning algorithm based on a large number of simulation data sets and test data sets of different ship types.

Images