CN118133437B - Ship local structural strength analysis method, device, computer equipment and medium - Google Patents
Ship local structural strength analysis method, device, computer equipment and medium Download PDFInfo
- Publication number
- CN118133437B CN118133437B CN202410571886.9A CN202410571886A CN118133437B CN 118133437 B CN118133437 B CN 118133437B CN 202410571886 A CN202410571886 A CN 202410571886A CN 118133437 B CN118133437 B CN 118133437B
- Authority
- CN
- China
- Prior art keywords
- neural network
- physical information
- network model
- loss
- displacement
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
- 238000004458 analytical method Methods 0.000 title claims abstract description 47
- 238000003062 neural network model Methods 0.000 claims abstract description 158
- 238000012549 training Methods 0.000 claims abstract description 124
- 238000006073 displacement reaction Methods 0.000 claims abstract description 86
- 238000000034 method Methods 0.000 claims abstract description 52
- 238000012795 verification Methods 0.000 claims abstract description 47
- 238000013461 design Methods 0.000 claims abstract description 32
- 239000007787 solid Substances 0.000 claims abstract description 20
- 230000006870 function Effects 0.000 claims description 100
- 230000036961 partial effect Effects 0.000 claims description 42
- 238000004590 computer program Methods 0.000 claims description 22
- 238000013528 artificial neural network Methods 0.000 claims description 13
- 238000004422 calculation algorithm Methods 0.000 claims description 13
- 230000004069 differentiation Effects 0.000 claims description 11
- 238000010276 construction Methods 0.000 claims description 9
- 238000013459 approach Methods 0.000 claims description 8
- 230000000452 restraining effect Effects 0.000 claims description 8
- 238000005516 engineering process Methods 0.000 claims description 6
- 238000012216 screening Methods 0.000 claims description 5
- 238000011478 gradient descent method Methods 0.000 claims description 4
- 238000010200 validation analysis Methods 0.000 claims description 3
- 230000008569 process Effects 0.000 description 7
- 238000010586 diagram Methods 0.000 description 6
- 238000004891 communication Methods 0.000 description 4
- 238000012545 processing Methods 0.000 description 3
- 238000004364 calculation method Methods 0.000 description 2
- 230000008859 change Effects 0.000 description 2
- 238000013500 data storage Methods 0.000 description 2
- 230000000670 limiting effect Effects 0.000 description 2
- 230000003068 static effect Effects 0.000 description 2
- OKTJSMMVPCPJKN-UHFFFAOYSA-N Carbon Chemical compound [C] OKTJSMMVPCPJKN-UHFFFAOYSA-N 0.000 description 1
- 230000001413 cellular effect Effects 0.000 description 1
- 238000006243 chemical reaction Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 229910021389 graphene Inorganic materials 0.000 description 1
- 239000004973 liquid crystal related substance Substances 0.000 description 1
- 239000000463 material Substances 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 238000005457 optimization Methods 0.000 description 1
- 229920000642 polymer Polymers 0.000 description 1
- 230000002829 reductive effect Effects 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
- G06F30/10—Geometric CAD
- G06F30/15—Vehicle, aircraft or watercraft design
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T17/00—Three dimensional [3D] modelling, e.g. data description of 3D objects
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2119/00—Details relating to the type or aim of the analysis or the optimisation
- G06F2119/14—Force analysis or force optimisation, e.g. static or dynamic forces
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Geometry (AREA)
- General Physics & Mathematics (AREA)
- Evolutionary Computation (AREA)
- General Engineering & Computer Science (AREA)
- Software Systems (AREA)
- Health & Medical Sciences (AREA)
- Computing Systems (AREA)
- Biomedical Technology (AREA)
- Biophysics (AREA)
- Computational Linguistics (AREA)
- Data Mining & Analysis (AREA)
- Life Sciences & Earth Sciences (AREA)
- General Health & Medical Sciences (AREA)
- Molecular Biology (AREA)
- Artificial Intelligence (AREA)
- Computer Graphics (AREA)
- Mathematical Physics (AREA)
- Automation & Control Theory (AREA)
- Aviation & Aerospace Engineering (AREA)
- Computational Mathematics (AREA)
- Mathematical Analysis (AREA)
- Mathematical Optimization (AREA)
- Pure & Applied Mathematics (AREA)
- Computer Hardware Design (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The application relates to a method, a device, computer equipment and a medium for analyzing the local structural strength of a ship, wherein the method comprises the following steps: acquiring a three-dimensional entity model corresponding to the analysis of the local structural strength of the ship; the discrete three-dimensional solid model is a three-dimensional coordinate point; generating a training set and a verification set according to the three-dimensional coordinate points, constructing a loss function based on the training set, displacement and stress boundary conditions corresponding to ship design, a preset punishment coefficient and an elastic mechanics theory, taking the training set as input, taking the displacement after structural stress as output, constructing a physical information neural network model based on the loss function, training the physical information neural network model by using the training set, and iteratively updating model parameters of the physical information neural network model to obtain a target physical information neural network model; and carrying out structural strength analysis on the ship part based on the target physical information neural network model.
Description
Technical Field
The present application relates to the technical field of ship design, and in particular, to a method, an apparatus, a computer device, a storage medium and a computer program product for analyzing local structural strength of a ship.
Background
The ship structural design adopts an iterative optimization mode on the basis of strength analysis, so that the functionality of the structure can be effectively ensured, but the design efficiency is greatly affected by a single iteration period.
The traditional mainstream intensity analysis method is a finite element method, whether the calculation result is accurate and closely related to the grid quality is judged, and the finite element model is created in a single iteration period and occupies a large proportion.
With the development of three-dimensional design platforms by more shipyards and ship design companies, how to fully utilize three-dimensional models and avoid complicated finite element modeling to efficiently realize the analysis of the structural strength of ships is a current urgent problem to be solved.
Disclosure of Invention
In view of the foregoing, it is desirable to provide an efficient method, apparatus, computer device, storage medium and computer program product for analyzing the local structural strength of a ship.
In a first aspect, the application provides a method for analyzing local structural strength of a ship. The method comprises the following steps:
Acquiring a three-dimensional entity model corresponding to the analysis of the local structural strength of the ship;
dispersing the three-dimensional solid model to obtain three-dimensional coordinate points;
Generating a training set and a verification set according to the three-dimensional coordinate points, and constructing a loss function based on the training set, displacement and stress boundary conditions corresponding to ship design, a preset penalty coefficient and an elastic mechanics theory, wherein the preset penalty coefficient is used for restraining balance of loss items in the constructed loss function;
taking the training set as input and the displacement after the structure is stressed as output, constructing a physical information neural network model based on the loss function, training the physical information neural network model by utilizing the training set, and iteratively updating model parameters of the physical information neural network model so that the displacement after the structure is stressed, which is output by the updated physical information neural network model, approximates to an elastic mechanical solution of the local strength of the structure;
verifying the physical information neural network model updated by each iteration according to the verification set and the loss function until the updated physical information neural network model reaches a preset convergence condition, so as to obtain a target physical information neural network model;
and carrying out structural strength analysis on the ship local based on the target physical information neural network model.
In one embodiment, the generating a training set and a verification set according to the three-dimensional coordinate points, and constructing the loss function based on the training set, the displacement and stress boundary conditions corresponding to the ship design, a preset penalty coefficient, and an elastic mechanics theory includes:
generating a training set and a verification set according to the three-dimensional coordinate points;
Obtaining displacement and stress boundary conditions corresponding to ship design, and screening target data points meeting the displacement and stress boundary conditions from the three-dimensional coordinate points;
Generating a boundary condition loss term according to the target data point, and generating a partial differential equation set loss term according to non-target data points in the three-dimensional coordinate point;
calculating a loss value of a boundary condition loss term by adopting an elastic mechanical balance equation based on a displacement solving method;
calculating a loss value of the partial differential equation set loss term according to the stress boundary condition;
And constructing a loss function based on the boundary condition loss term and the partial differential equation set loss term, and obtaining a total loss value based on the loss value of the boundary condition loss term and the loss value of the partial differential equation set loss term.
In one embodiment, generating a partial differential equation set loss term from non-target data points in the three-dimensional coordinate points includes:
Obtaining a corresponding Einstein tensor representation by adopting automatic differentiation according to non-target data points in the three-dimensional coordinate points;
Based on the einstein tensor representation, a partial differential equation set loss term is generated.
In one embodiment, iteratively updating model parameters of the physical information neural network model includes:
And iteratively updating model parameters of the physical information neural network model by adopting a gradient descent algorithm and a back propagation algorithm.
In one embodiment, the constructing a physical information neural network model based on the loss function by taking the training set as input and taking the displacement after the structure is stressed as output, training the physical information neural network model by using the training set, and iteratively updating model parameters of the physical information neural network model, so that the updated elastic mechanical solution of the displacement after the structure is stressed, which is output by the physical information neural network model, approximates to the local strength of the structure comprises:
Inputting the training set into a fully-connected neural network to generate a physical information neural network model taking displacement of the structure after being stressed as output;
Solving the loss function according to the output of the physical information neural network model by utilizing an automatic differentiation technology;
Iteratively updating model parameters of the physical information neural network model by adopting a gradient descent method until a loss value corresponding to the loss function of the updated physical information neural network model approaches to 0; and the loss value corresponding to the loss function approaches to 0to represent the elastic mechanical solution of the local strength of the displacement approximation structure after the structure output by the updated physical information neural network model is stressed.
In one embodiment, the generating the training set and the verification set according to the three-dimensional coordinate points includes:
Randomly dividing the three-dimensional coordinate point into first partial data and second partial data, wherein the data volume of the first partial data is larger than that of the second partial data;
the first part of data is used as a training set, and the second part of data is used as a verification set.
In one embodiment, the preset convergence condition includes a preset loss function value condition or an increment condition of the loss function in each preset iteration.
In a second aspect, the application further provides a device for analyzing the local structural strength of the ship. The device comprises:
The model acquisition module is used for acquiring a three-dimensional entity model corresponding to the analysis of the local structural strength of the ship;
the discrete module is used for dispersing the three-dimensional solid model into three-dimensional coordinate points;
The loss function construction module is used for generating a training set and a verification set according to the three-dimensional coordinate points, constructing a loss function based on the training set, displacement and stress boundary conditions corresponding to ship design, a preset penalty coefficient and an elastic mechanics theory, wherein the preset penalty coefficient is used for restraining balance of loss items in the constructed loss function;
The model training module is used for taking the training set as input and taking the displacement of the structure after being stressed as output, constructing a physical information neural network model based on the loss function, training the physical information neural network model by utilizing the training set, and iteratively updating model parameters of the physical information neural network model so as to enable the displacement of the structure after being stressed, which is output by the updated physical information neural network model, to approximate to an elastic mechanical solution of the local strength of the structure;
the model verification module is used for verifying the physical information neural network model updated by each iteration according to the verification set and the loss function until the updated physical information neural network model reaches a preset convergence condition, so as to obtain a target physical information neural network model;
and the analysis module is used for analyzing the structural strength of the ship part based on the target physical information neural network model.
In a third aspect, the present application also provides a computer device. The computer device comprises a memory storing a computer program and a processor which when executing the computer program performs the steps of:
Acquiring a three-dimensional entity model corresponding to the analysis of the local structural strength of the ship;
dispersing the three-dimensional solid model to obtain three-dimensional coordinate points;
Generating a training set and a verification set according to the three-dimensional coordinate points, and constructing a loss function based on the training set, displacement and stress boundary conditions corresponding to ship design, a preset penalty coefficient and an elastic mechanics theory, wherein the preset penalty coefficient is used for restraining balance of loss items in the constructed loss function;
taking the training set as input and the displacement after the structure is stressed as output, constructing a physical information neural network model based on the loss function, training the physical information neural network model by utilizing the training set, and iteratively updating model parameters of the physical information neural network model so that the displacement after the structure is stressed, which is output by the updated physical information neural network model, approximates to an elastic mechanical solution of the local strength of the structure;
verifying the physical information neural network model updated by each iteration according to the verification set and the loss function until the updated physical information neural network model reaches a preset convergence condition, so as to obtain a target physical information neural network model;
and carrying out structural strength analysis on the ship local based on the target physical information neural network model.
In a fourth aspect, the present application also provides a computer-readable storage medium. The computer readable storage medium having stored thereon a computer program which when executed by a processor performs the steps of:
Acquiring a three-dimensional entity model corresponding to the analysis of the local structural strength of the ship;
dispersing the three-dimensional solid model to obtain three-dimensional coordinate points;
Generating a training set and a verification set according to the three-dimensional coordinate points, and constructing a loss function based on the training set, displacement and stress boundary conditions corresponding to ship design, a preset penalty coefficient and an elastic mechanics theory, wherein the preset penalty coefficient is used for restraining balance of loss items in the constructed loss function;
taking the training set as input and the displacement after the structure is stressed as output, constructing a physical information neural network model based on the loss function, training the physical information neural network model by utilizing the training set, and iteratively updating model parameters of the physical information neural network model so that the displacement after the structure is stressed, which is output by the updated physical information neural network model, approximates to an elastic mechanical solution of the local strength of the structure;
verifying the physical information neural network model updated by each iteration according to the verification set and the loss function until the updated physical information neural network model reaches a preset convergence condition, so as to obtain a target physical information neural network model;
and carrying out structural strength analysis on the ship local based on the target physical information neural network model.
In a fifth aspect, the present application also provides a computer program product. The computer program product comprises a computer program which, when executed by a processor, implements the steps of:
Acquiring a three-dimensional entity model corresponding to the analysis of the local structural strength of the ship;
dispersing the three-dimensional solid model to obtain three-dimensional coordinate points;
Generating a training set and a verification set according to the three-dimensional coordinate points, and constructing a loss function based on the training set, displacement and stress boundary conditions corresponding to ship design, a preset penalty coefficient and an elastic mechanics theory, wherein the preset penalty coefficient is used for restraining balance of loss items in the constructed loss function;
taking the training set as input and the displacement after the structure is stressed as output, constructing a physical information neural network model based on the loss function, training the physical information neural network model by utilizing the training set, and iteratively updating model parameters of the physical information neural network model so that the displacement after the structure is stressed, which is output by the updated physical information neural network model, approximates to an elastic mechanical solution of the local strength of the structure;
verifying the physical information neural network model updated by each iteration according to the verification set and the loss function until the updated physical information neural network model reaches a preset convergence condition, so as to obtain a target physical information neural network model;
and carrying out structural strength analysis on the ship local based on the target physical information neural network model.
The method, the device, the computer equipment, the storage medium and the computer program product for analyzing the local structural strength of the ship acquire a three-dimensional entity model corresponding to the analysis of the local structural strength of the ship; the discrete three-dimensional solid model is a three-dimensional coordinate point; generating a training set and a verification set according to the three-dimensional coordinate points, constructing a loss function based on the training set, displacement and stress boundary conditions corresponding to ship design, a preset punishment coefficient and an elastic mechanics theory, taking the training set as input, taking the displacement after structural stress as output, constructing a physical information neural network model based on the loss function, training the physical information neural network model by using the training set, and iteratively updating model parameters of the physical information neural network model so that the displacement after structural stress output by the updated physical information neural network model approximates to the elastic mechanics solution of the structural local strength; verifying the physical information neural network model updated by each iteration according to the verification set and the loss function until the updated physical information neural network model reaches a preset convergence condition to obtain a target physical information neural network model; and carrying out structural strength analysis on the ship part based on the target physical information neural network model. In the whole process, a three-dimensional design model is used as a direct analysis object, discrete point coordinates are used as model input, the output on each discrete point is guaranteed to meet an elastic theoretical equilibrium equation, complicated finite element modeling is not needed, a target physical information neural network model is obtained based on the data and an iterative training mode, and finally structural strength analysis can be effectively carried out on a ship part based on the target physical information neural network model.
Drawings
FIG. 1 is a diagram of an application environment of a method for analyzing the local structural strength of a ship in one embodiment;
FIG. 2 is a flow chart of a method of analyzing local structural strength of a ship in one embodiment;
FIG. 3 is a schematic diagram of a sub-process of S300 in one embodiment;
FIG. 4 is a schematic diagram of a model architecture;
FIG. 5 is a flow chart of a method for analyzing the local structural strength of a ship in one embodiment;
FIG. 6 is a block diagram of a device for analyzing the local structural strength of a ship according to an embodiment;
Fig. 7 is an internal structural diagram of a computer device in one embodiment.
Detailed Description
The present application will be described in further detail with reference to the drawings and examples, in order to make the objects, technical solutions and advantages of the present application more apparent. It should be understood that the specific embodiments described herein are for purposes of illustration only and are not intended to limit the scope of the application.
The method for analyzing the local structural strength of the ship provided by the embodiment of the application can be applied to an application environment shown in figure 1. Wherein the terminal 102 communicates with the server 104 via a network. The data storage system may store data that the server 104 needs to process. The data storage system may be integrated on the server 104 or may be located on a cloud or other network server. The terminal 102 sends a ship local structural strength analysis request to the server 104, and the server 104 responds to the ship local structural strength analysis request to acquire a three-dimensional entity model corresponding to the ship local structural strength analysis; the discrete three-dimensional solid model is a three-dimensional coordinate point; generating a training set and a verification set according to the three-dimensional coordinate points, constructing a loss function based on the training set, displacement and stress boundary conditions corresponding to ship design, a preset punishment coefficient and an elastic mechanics theory, taking the training set as input, taking the displacement after structural stress as output, constructing a physical information neural network model based on the loss function, training the physical information neural network model by using the training set, and iteratively updating model parameters of the physical information neural network model so that the displacement after structural stress output by the updated physical information neural network model approximates to the elastic mechanics solution of the structural local strength; verifying the physical information neural network model updated by each iteration according to the verification set and the loss function until the updated physical information neural network model reaches a preset convergence condition to obtain a target physical information neural network model; and carrying out structural strength analysis on the ship part based on the target physical information neural network model. The terminal 102 may be, but not limited to, various personal computers, notebook computers, smart phones, tablet computers, internet of things devices, and portable wearable devices, where the internet of things devices may be smart speakers, smart televisions, smart air conditioners, smart vehicle devices, and the like. The portable wearable device may be a smart watch, smart bracelet, headset, or the like. The server 104 may be implemented as a stand-alone server or as a server cluster of multiple servers. It should be noted that the method for analyzing the local structural strength of the ship can also be directly applied to the terminal, namely, the user can operate on the terminal side, the user can interact with the terminal, and the terminal generates a final analysis result of the local structural strength of the ship and feeds back (displays) the analysis result to the user.
In one embodiment, as shown in fig. 2, a method for analyzing local structural strength of a ship is provided, and the method is applied to the server 104 in fig. 1 for illustration, and includes the following steps:
s100: and obtaining a three-dimensional entity model corresponding to the analysis of the local structural strength of the ship.
The ship part is intercepted according to the actual situation. Specifically, for example, when we need to perform structural strength analysis on the bow part, the three-dimensional solid model of the bow part is cut out from the whole three-dimensional solid model of the ship. After the three-dimensional solid model corresponding to the analysis of the local structural strength of the ship is obtained, further displacement boundary conditions and stress boundary conditions can be defined according to engineering problems.
S200: the discrete three-dimensional solid model is a three-dimensional coordinate point.
And discretizing the intercepted three-dimensional solid model into three-dimensional coordinate points. Specifically, the three-dimensional solid model can be discretized into three-dimensional coordinate points x= (X 1、X2、……、Xn) by using a stereolithography rapid prototyping technology. Here, the stereolithography rapid prototyping technique successfully achieves the conversion from virtual design to physical entity by discretizing a three-dimensional solid model into a series of three-dimensional coordinate points.
S300: generating a training set and a verification set according to the three-dimensional coordinate points, and constructing a loss function based on the training set, displacement and stress boundary conditions corresponding to ship design, a preset penalty coefficient and an elastic mechanics theory, wherein the preset penalty coefficient is used for restraining balance of loss items in the constructed loss function.
Here, the three-dimensional coordinate points may be converted into a data set for physical information neural network training, and then the data set may be divided into a training set and a verification set. Specifically, the duty ratio of the training set and the verification set can be set according to the actual situation, and since the application only includes the coordinate point data of the result, namely belongs to the data set without the label, in order to enable the model obtained by training to be converged subsequently, the verification set is generally not smaller than the training set, specifically, 1/3 three-dimensional coordinate point can be used as the training set, and 2/3 three-dimensional coordinate point can be used as the verification set; it is also possible to use 1/2 three-dimensional coordinate points as training sets and 1/2 three-dimensional coordinate points as verification sets.
In general, displacement and stress boundary conditions are generally known, and are design parameters specified by specifications or determined in practice, based on specific vessel local structural strength analysis problems. Therefore, the displacement and stress boundary conditions corresponding to the ship design are parameters which can be directly obtained. The preset penalty coefficient is used for balancing loss terms in the loss function obtained by constraint construction. In particular, a special loss function is constructed here, which consists of a plurality of loss terms; the magnitude of each loss term may be different, for example, the boundary condition loss of the MSE u term includes a displacement boundary and a stress boundary loss, the displacement value may be 6, the stress value may be 500, at this time, the two values are different by two magnitudes, which may lead to the stress boundary condition to be dominant in the model training process, so that the model is more biased to find a solution in a certain direction, and finally, a locally optimal solution may be obtained, and the prediction effect on the local structural strength of the ship may be greatly reduced. Therefore, penalty coefficients need to be set to make the initial individual loss terms of the model close in magnitude, i.e., to constrain the model.
And constructing a loss function based on the elastic theory according to the training set, the displacement and stress boundary conditions corresponding to the ship design and the preset punishment coefficient. Specifically, the loss function includes two parts of a boundary condition loss term MSE u and a partial differential equation set loss term MSE i.
S400: the training set is taken as input, the displacement after the structure is stressed is taken as output, a physical information neural network model is constructed based on a loss function, the physical information neural network model is trained by the training set, and model parameters of the physical information neural network model are iteratively updated, so that the displacement after the structure is stressed, which is output by the updated physical information neural network model, approximates to an elastomehc solution of the local strength of the structure.
The physical information neural network model is constructed with the training set as input and the displacement after structural stress as output and the aggregate loss function. Specifically, the training set includes three-dimensional coordinate points, the three-dimensional coordinate points are input into a preset fully connected neural network, the output of the fully connected neural network is directly used as the displacement of the structure after being stressed to construct an obtained physical information neural network model, then the training set is used for training the physical information neural network model, model parameters of the physical information neural network model are iteratively updated in the training process, a loss function corresponding to the updated physical information neural network model is continuously calculated, and when the loss function is infinitely close to 0, the fact that the displacement of the structure after being stressed output by the updated physical information neural network model approximates to an elasto-mechanical solution of the local strength of the structure is indicated.
S500: and verifying the physical information neural network model updated by each iteration according to the verification set and the loss function until the updated physical information neural network model reaches a preset convergence condition, so as to obtain the target physical information neural network model.
The preset convergence condition is a preset condition for judging that the physical information neural network model has reached convergence, and specifically may include a preset loss function value condition, or a loss function increment condition of each time in preset iteration, and the like. For example, a loss function value of < 10 -4; and/or the absolute value of each loss function increment in subsequent iterations < 10 -12, etc. And when the updated physical information neural network model is judged to reach the preset convergence condition, obtaining the target physical information neural network model. In the verification process, if the updated physical information neural network model is found to fail to reach the preset convergence condition, the initial model parameters are adjusted, the penalty coefficients are reset, the new initial physical information neural network model is reconstructed in S300, and a new round of iterative training is started. Specifically, the penalty factor reset may be adjusted according to the magnitudes of different terms in the loss function to ensure that each loss term is at the same magnitude to accelerate model convergence.
S600: and carrying out structural strength analysis on the ship part based on the target physical information neural network model.
And (5) analyzing the structural strength of the ship part by utilizing the target physical information neural network model obtained by training in the step (S500).
According to the method for analyzing the local structural strength of the ship, the three-dimensional entity model corresponding to the analysis of the local structural strength of the ship is obtained; the discrete three-dimensional solid model is a three-dimensional coordinate point; generating a training set and a verification set according to the three-dimensional coordinate points, constructing a loss function based on the training set, displacement and stress boundary conditions corresponding to ship design, a preset punishment coefficient and an elastic mechanics theory, taking the training set as input, taking the displacement after structural stress as output, constructing a physical information neural network model based on the loss function, training the physical information neural network model by using the training set, and iteratively updating model parameters of the physical information neural network model so that the displacement after structural stress output by the updated physical information neural network model approximates to the elastic mechanics solution of the structural local strength; verifying the physical information neural network model updated by each iteration according to the verification set and the loss function until the updated physical information neural network model reaches a preset convergence condition to obtain a target physical information neural network model; and carrying out structural strength analysis on the ship part based on the target physical information neural network model. In the whole process, a three-dimensional design model is used as a direct analysis object, discrete point coordinates are used as model input, the output on each discrete point is guaranteed to meet an elastic theoretical equilibrium equation, complicated finite element modeling is not needed, a target physical information neural network model is obtained based on the data and an iterative training mode, and finally structural strength analysis can be effectively carried out on a ship part based on the target physical information neural network model.
As shown in fig. 3, in one embodiment, S300 includes:
S310: and generating a training set and a verification set according to the three-dimensional coordinate points.
Converting three-dimensional coordinate points into a data set for training a physical information neural networkWherein a smaller portion of the data is used as a training setA larger portion of the data is used as a validation set。
S320: and obtaining displacement and stress boundary conditions corresponding to the ship design, and screening target data points meeting the displacement and stress boundary conditions from the three-dimensional coordinate points.
The displacement and stress boundary conditions for a ship design are conditions that are directly determined based on the ship design requirements. From training setPoetry anthology, which satisfy the displacement and stress boundary conditions, are used as target data points for the next step of loss term calculation.
S330: and generating a boundary condition loss term according to the target data point, and generating a partial differential equation set loss term according to the non-target data point in the three-dimensional coordinate point.
The loss term MSE u is calculated based on the target data point obtained in S320. In addition, a penalty coefficient is set for calculating a partial differential equation set loss term MSE i for non-target data points in the three-dimensional coordinate points. Specifically, when generating partial differential equation set loss term MSE i, automatic differentiation may be used to find the correspondingAndThe partial differential equation set loss term MSE i is then calculated. In this context,It is an einstein tensor representation, a simplified representation of the computation between partial derivatives of the displacement.
S340: and calculating the loss value of the boundary condition loss term by adopting an elastic mechanical balance equation based on a displacement solving method.
The loss term is calculated by adopting an elastic mechanical balance equation based on a displacement solution method, and the Einstein tensor expression form is as follows:
s350: and calculating the loss value of the partial differential equation set loss term according to the stress boundary condition.
And calculating the loss value of the partial differential equation set loss term by adopting the corresponding stress boundary condition.
In the two formulas described above,In order for the displacement to be a function of,In order to be strained the material is,Is of volume, a,In order to be a force of a face,，In order to achieve a shear modulus, the polymer is,In order to pull Mei Jishu the wire,Is the direction cosine of the direction of the normal outside the boundary.
S360: and constructing a loss function based on the boundary condition loss term and the partial differential equation set loss term, and obtaining a total loss value based on the loss value of the boundary condition loss term and the loss value of the partial differential equation set loss term.
Based on the boundary condition loss term MSE u and the partial differential equation set loss term MSE i obtained as described above, a loss function is constructed, and based on the loss values corresponding to the two loss terms, a total loss value is calculated. Specifically, the total LOSS value LOSS is in the form:
In the above-mentioned method, the step of, Is a penalty coefficient; for the number of boundary points used for training, A loss term for boundary conditions; for the number of points within the physical field used for training, A loss term for the partial differential equation set; Is a boundary displacement constraint.
In one embodiment, iteratively updating model parameters of the physical information neural network model includes:
and iteratively updating model parameters of the physical information neural network model by adopting a gradient descent algorithm and a back propagation algorithm.
In the process of training the physical information neural network model by using the training set, the model parameters are iteratively updated through gradient descent and back propagation algorithms, so that the output displacement of the model is realizedAnd gradually approaching the elastic mechanical solution of the local strength of the structure.
In one embodiment, with the training set as input and the displacement after the structure is stressed as output, constructing a physical information neural network model based on the loss function, training the physical information neural network model by using the training set, and iteratively updating model parameters of the physical information neural network model, so that the elastic mechanical solution of the displacement after the structure is stressed, which is output by the updated physical information neural network model, approximates the local strength of the structure comprises:
Inputting the training set into a fully-connected neural network to generate a physical information neural network model taking displacement of the structure after being stressed as output; solving a loss function according to the output of the physical information neural network model and by utilizing an automatic differentiation technology; iteratively updating model parameters of the physical information neural network model by adopting a gradient descent method until a loss value corresponding to a loss function of the updated physical information neural network model approaches to 0; and the loss value corresponding to the loss function approaches to 0 to represent the elastic mechanical solution of the local strength of the structure approximated by the displacement after the structure output by the updated physical information neural network model is stressed.
The model principle as shown in fig. 4 is explained as follows:
the physical field is assumed to satisfy the following equation:
；
；
。
Wherein, Is the original function to be solved and is characterized in that,AndIs a function of the satisfaction of the physical field initial value condition. The step of constructing a physical information neural network model to solve u is as follows: first, data in training setInputting the fully-connected neural network to generate an initial network; Then, by means of automatic differentiation、AndThen calculating two loss functions; finally, the weight parameters of the neural network are adjusted by utilizing a gradient descent algorithmMultiple iterative trainingApproximation is made。
Specifically, the weight parameters of the neural network are adjusted by using the gradient descent algorithmThe processing procedure of (1) comprises the following steps:
1. randomly initializing weight parameters of a neural network (I= 0,1,3 … n) and setting an initial step sizeAnd algorithm termination distance；
2. Calculating current loss function versus parameter from training setGradient values of (2) then using step sizeMultiplying the gradient value to obtain the descending distance S of the current position;
3. If all parameters are The gradient drops by a distance less thanEnding the algorithm, otherwise, entering step 4;
4. Updating all parameters Will be currentThe value minus S is taken as a new weight parameter for the neural network and then step 1 is entered.
In order to describe the method for analyzing the local structural strength of the ship according to the present application in detail, a specific application example will be described below with reference to fig. 5. The method for analyzing the local structural strength of the ship comprises the following steps of:
1. intercepting a three-dimensional solid model in a local intensity analysis range;
2. Defining displacement and stress boundary conditions;
3. the discrete entity model is a three-dimensional coordinate point;
4. Generating a training set based on the three-dimensional coordinate points, and separating out a verification set;
5. Adding penalty coefficients Constructing a loss function based on an elasticity theory;
6. constructing PINNs models;
7. Calculating a loss value by using automatic differentiation;
8. model accuracy verification is carried out by utilizing a verification set, if yes, step 9 is carried out, and if not, penalty coefficients are modified Returning to the step 5;
9. based on the validated model, the local structural strength of the solid model is calculated.
It should be understood that, although the steps in the flowcharts related to the above embodiments are sequentially shown as indicated by arrows, these steps are not necessarily sequentially performed in the order indicated by the arrows. The steps are not strictly limited to the order of execution unless explicitly recited herein, and the steps may be executed in other orders. Moreover, at least some of the steps in the flowcharts described in the above embodiments may include a plurality of steps or a plurality of stages, which are not necessarily performed at the same time, but may be performed at different times, and the order of the steps or stages is not necessarily performed sequentially, but may be performed alternately or alternately with at least some of the other steps or stages.
Based on the same inventive concept, the embodiment of the application also provides a ship local structural strength analysis device for realizing the ship local structural strength analysis method. The implementation of the solution provided by the device is similar to the implementation described in the above method, so the specific limitation in the embodiments of the device for analyzing the local structural strength of a ship provided below may be referred to the limitation of the method for analyzing the local structural strength of a ship hereinabove, and will not be repeated herein.
In one embodiment, as shown in fig. 6, there is provided a ship local structural strength analysis apparatus, comprising:
The model acquisition module 100 is used for acquiring a three-dimensional entity model corresponding to the analysis of the local structural strength of the ship;
The discrete module 200 is used for dispersing the three-dimensional entity model into three-dimensional coordinate points;
The loss function construction module 300 is configured to generate a training set and a verification set according to the three-dimensional coordinate points, and construct a loss function based on the training set, displacement and stress boundary conditions corresponding to the ship design, a preset penalty coefficient and an elastic mechanics theory, wherein the preset penalty coefficient is used for constraining balance of loss items in the constructed loss function;
The model training module 400 is configured to construct a physical information neural network model based on a loss function by taking a training set as input and taking displacement after the structure is stressed as output, train the physical information neural network model by using the training set, and iteratively update model parameters of the physical information neural network model, so that the displacement after the structure is stressed, which is output by the updated physical information neural network model, approximates to an elastic mechanical solution of the local strength of the structure;
The model verification module 500 is configured to verify the physical information neural network model updated in each iteration according to the verification set and the loss function until the updated physical information neural network model reaches a preset convergence condition, thereby obtaining a target physical information neural network model;
The analysis module 600 is used for analyzing the structural strength of the ship part based on the target physical information neural network model.
In one embodiment, the loss function construction module 300 is further configured to generate a training set and a verification set according to the three-dimensional coordinate points; obtaining displacement and stress boundary conditions corresponding to ship design, and screening target data points meeting the displacement and stress boundary conditions from the three-dimensional coordinate points; generating a boundary condition loss term according to the target data point, and generating a partial differential equation set loss term according to non-target data points in the three-dimensional coordinate point; calculating a loss value of a boundary condition loss term by adopting an elastic mechanical balance equation based on a displacement solving method; calculating a loss value of a partial differential equation set loss term according to the stress boundary condition; and constructing a loss function based on the boundary condition loss term and the partial differential equation set loss term, and obtaining a total loss value based on the loss value of the boundary condition loss term and the loss value of the partial differential equation set loss term.
In one embodiment, the loss function construction module 300 is further configured to obtain a corresponding einstein tensor representation according to non-target data points in the three-dimensional coordinate points and using automatic differentiation; based on the einstein tensor representation, partial differential equation set loss terms are generated.
In one embodiment, the model training module 400 is further configured to iteratively update model parameters of the physical information neural network model using a gradient descent algorithm and a back propagation algorithm.
In one embodiment, the model training module 400 is further configured to input the training set to a fully connected neural network, and generate a physical information neural network model using the displacement of the structure after being stressed as an output; solving a loss function according to the output of the physical information neural network model and by utilizing an automatic differentiation technology; iteratively updating model parameters of the physical information neural network model by adopting a gradient descent method until a loss value corresponding to a loss function of the updated physical information neural network model approaches to 0; and the loss value corresponding to the loss function approaches to 0 to represent the elastic mechanical solution of the local strength of the structure approximated by the displacement after the structure output by the updated physical information neural network model is stressed.
The loss function construction module 300 is further configured to randomly divide the three-dimensional coordinate point into a first part of data and a second part of data, where the data size of the first part of data is greater than the data size of the second part of data; the first portion of data is used as a training set and the second portion of data is used as a validation set.
In one embodiment, the predetermined convergence condition includes a predetermined loss function value condition or an increment of the loss function condition for each of the predetermined iterations.
The above-described modules in the ship local structural strength analysis apparatus may be implemented in whole or in part by software, hardware, and a combination thereof. The above modules may be embedded in hardware or may be independent of a processor in the computer device, or may be stored in software in a memory in the computer device, so that the processor may call and execute operations corresponding to the above modules.
In one embodiment, a computer device is provided, which may be a terminal, and the internal structure of which may be as shown in fig. 7. The computer device includes a processor, a memory, a communication interface, a display screen, and an input device connected by a system bus. Wherein the processor of the computer device is configured to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for the operation of the operating system and computer programs in the non-volatile storage media. The communication interface of the computer device is used for carrying out wired or wireless communication with an external terminal, and the wireless mode can be realized through WIFI, a mobile cellular network, NFC (near field communication) or other technologies. The computer program is executed by the processor to implement a method of analyzing the local structural strength of a vessel. The display screen of the computer equipment can be a liquid crystal display screen or an electronic ink display screen, and the input device of the computer equipment can be a touch layer covered on the display screen, can also be keys, a track ball or a touch pad arranged on the shell of the computer equipment, and can also be an external keyboard, a touch pad or a mouse and the like.
It will be appreciated by those skilled in the art that the structure shown in FIG. 7 is merely a block diagram of some of the structures associated with the present inventive arrangements and is not limiting of the computer device to which the present inventive arrangements may be applied, and that a particular computer device may include more or fewer components than shown, or may combine some of the components, or have a different arrangement of components.
In one embodiment, a computer device is provided, comprising a memory and a processor, the memory storing a computer program, the processor implementing the above-described method of analyzing the local structural strength of a vessel when executing the computer program.
In one embodiment, a computer readable storage medium is provided, on which a computer program is stored, which when executed by a processor implements the above-described method of analyzing the local structural strength of a vessel.
In an embodiment, a computer program product is provided, comprising a computer program which, when executed by a processor, implements the above-described method of analyzing the local structural strength of a vessel.
Those skilled in the art will appreciate that implementing all or part of the above-described methods may be accomplished by way of a computer program, which may be stored on a non-transitory computer readable storage medium and which, when executed, may comprise the steps of the above-described embodiments of the methods. Any reference to memory, database, or other medium used in embodiments provided herein may include at least one of non-volatile and volatile memory. The nonvolatile Memory may include Read-Only Memory (ROM), magnetic tape, floppy disk, flash Memory, optical Memory, high density embedded nonvolatile Memory, resistive random access Memory (ReRAM), magneto-resistive random access Memory (Magnetoresistive Random Access Memory, MRAM), ferroelectric Memory (Ferroelectric Random Access Memory, FRAM), phase change Memory (PHASE CHANGE Memory, PCM), graphene Memory, and the like. Volatile memory can include random access memory (Random Access Memory, RAM) or external cache memory, and the like. By way of illustration, and not limitation, RAM can be in various forms such as static random access memory (Static Random Access Memory, SRAM) or dynamic random access memory (Dynamic Random Access Memory, DRAM), etc. The databases referred to in the embodiments provided herein may include at least one of a relational database and a non-relational database. The non-relational database may include, but is not limited to, a blockchain-based distributed database, and the like. The processor referred to in the embodiments provided in the present application may be a general-purpose processor, a central processing unit, a graphics processor, a digital signal processor, a programmable logic unit, a data processing logic unit based on quantum computing, or the like, but is not limited thereto.
The technical features of the above embodiments may be arbitrarily combined, and all possible combinations of the technical features in the above embodiments are not described for brevity of description, however, as long as there is no contradiction between the combinations of the technical features, they should be considered as the scope of the description.
The foregoing examples illustrate only a few embodiments of the application, which are described in detail and are not to be construed as limiting the scope of the application. It should be noted that it will be apparent to those skilled in the art that several variations and modifications can be made without departing from the spirit of the application, which are all within the scope of the application. Accordingly, the scope of the application should be assessed as that of the appended claims.
Claims (10)
1. A method for analyzing local structural strength of a ship, the method comprising:
Acquiring a three-dimensional entity model corresponding to the analysis of the local structural strength of the ship;
dispersing the three-dimensional solid model to obtain three-dimensional coordinate points;
Generating a training set and a verification set according to the three-dimensional coordinate points, and constructing a loss function based on the training set, displacement and stress boundary conditions corresponding to ship design, a preset penalty coefficient and an elastic mechanics theory, wherein the preset penalty coefficient is used for restraining balance of loss items in the constructed loss function;
taking the training set as input and the displacement after the structure is stressed as output, constructing a physical information neural network model based on the loss function, training the physical information neural network model by utilizing the training set, and iteratively updating model parameters of the physical information neural network model so that the displacement after the structure is stressed, which is output by the updated physical information neural network model, approximates to an elastic mechanical solution of the local strength of the structure;
verifying the physical information neural network model updated by each iteration according to the verification set and the loss function until the updated physical information neural network model reaches a preset convergence condition, so as to obtain a target physical information neural network model;
carrying out structural strength analysis on the ship part based on the target physical information neural network model;
Generating a training set and a verification set according to the three-dimensional coordinate points, and constructing a loss function based on the training set, displacement and stress boundary conditions corresponding to ship design, a preset punishment coefficient and an elastic mechanics theory comprises: generating a training set and a verification set according to the three-dimensional coordinate points; obtaining displacement and stress boundary conditions corresponding to ship design, and screening target data points meeting the displacement and stress boundary conditions from the three-dimensional coordinate points; generating a boundary condition loss term according to the target data point, and generating a partial differential equation set loss term according to non-target data points in the three-dimensional coordinate point; calculating a loss value of a boundary condition loss term by adopting an elastic mechanical balance equation based on a displacement solving method; calculating a loss value of the partial differential equation set loss term according to the stress boundary condition; and constructing a loss function based on the boundary condition loss term and the partial differential equation set loss term, and obtaining a total loss value based on the loss value of the boundary condition loss term and the loss value of the partial differential equation set loss term.
2. The method of claim 1, wherein generating a partial differential equation set loss term from non-target data points in the three-dimensional coordinate points comprises:
Obtaining a corresponding Einstein tensor representation by adopting automatic differentiation according to non-target data points in the three-dimensional coordinate points;
Based on the einstein tensor representation, a partial differential equation set loss term is generated.
3. The method of claim 1, wherein iteratively updating model parameters of the physical information neural network model comprises:
And iteratively updating model parameters of the physical information neural network model by adopting a gradient descent algorithm and a back propagation algorithm.
4. The method of claim 1, wherein constructing a physical information neural network model based on the loss function with the training set as input and the displacement after the structure is stressed as output, training the physical information neural network model with the training set, and iteratively updating model parameters of the physical information neural network model such that the displacement after the structure is stressed output by the updated physical information neural network model approximates an elastomehc solution of the local strength of the structure comprises:
Inputting the training set into a fully-connected neural network to generate a physical information neural network model taking displacement of the structure after being stressed as output;
Solving the loss function according to the output of the physical information neural network model by utilizing an automatic differentiation technology;
Iteratively updating model parameters of the physical information neural network model by adopting a gradient descent method until a loss value corresponding to the loss function of the updated physical information neural network model approaches to 0; and the loss value corresponding to the loss function approaches to 0to represent the elastic mechanical solution of the local strength of the displacement approximation structure after the structure output by the updated physical information neural network model is stressed.
5. The method of claim 1, wherein the generating a training set and a validation set from the three-dimensional coordinate points comprises:
Randomly dividing the three-dimensional coordinate point into first partial data and second partial data, wherein the data volume of the first partial data is larger than that of the second partial data;
the first part of data is used as a training set, and the second part of data is used as a verification set.
6. The method of claim 1, wherein the predetermined convergence condition comprises a predetermined loss function value condition or an incremental loss function condition for each of a predetermined iteration.
7. A device for analyzing the local structural strength of a ship, said device comprising:
The model acquisition module is used for acquiring a three-dimensional entity model corresponding to the analysis of the local structural strength of the ship;
the discrete module is used for dispersing the three-dimensional solid model into three-dimensional coordinate points;
The loss function construction module is used for generating a training set and a verification set according to the three-dimensional coordinate points, constructing a loss function based on the training set, displacement and stress boundary conditions corresponding to ship design, a preset penalty coefficient and an elastic mechanics theory, wherein the preset penalty coefficient is used for restraining balance of loss items in the constructed loss function;
The model training module is used for taking the training set as input and taking the displacement of the structure after being stressed as output, constructing a physical information neural network model based on the loss function, training the physical information neural network model by utilizing the training set, and iteratively updating model parameters of the physical information neural network model so as to enable the displacement of the structure after being stressed, which is output by the updated physical information neural network model, to approximate to an elastic mechanical solution of the local strength of the structure;
the model verification module is used for verifying the physical information neural network model updated by each iteration according to the verification set and the loss function until the updated physical information neural network model reaches a preset convergence condition, so as to obtain a target physical information neural network model;
The analysis module is used for analyzing the structural strength of the ship part based on the target physical information neural network model;
The loss function construction module is further used for generating a training set and a verification set according to the three-dimensional coordinate points; obtaining displacement and stress boundary conditions corresponding to ship design, and screening target data points meeting the displacement and stress boundary conditions from the three-dimensional coordinate points; generating a boundary condition loss term according to the target data point, and generating a partial differential equation set loss term according to non-target data points in the three-dimensional coordinate point; calculating a loss value of a boundary condition loss term by adopting an elastic mechanical balance equation based on a displacement solving method; calculating a loss value of the partial differential equation set loss term according to the stress boundary condition; and constructing a loss function based on the boundary condition loss term and the partial differential equation set loss term, and obtaining a total loss value based on the loss value of the boundary condition loss term and the loss value of the partial differential equation set loss term.
8. The apparatus of claim 7, wherein the loss function construction module is further configured to derive a corresponding einstein tensor representation from non-target data points in the three-dimensional coordinate points and using automatic differentiation; based on the einstein tensor representation, partial differential equation set loss terms are generated.
9. A computer device comprising a memory and a processor, the memory storing a computer program, characterized in that the processor implements the steps of the method of any of claims 1 to 6 when the computer program is executed.
10. A computer readable storage medium, on which a computer program is stored, characterized in that the computer program, when being executed by a processor, implements the steps of the method of any of claims 1 to 6.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410571886.9A CN118133437B (en) | 2024-05-10 | 2024-05-10 | Ship local structural strength analysis method, device, computer equipment and medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410571886.9A CN118133437B (en) | 2024-05-10 | 2024-05-10 | Ship local structural strength analysis method, device, computer equipment and medium |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN118133437A CN118133437A (en) | 2024-06-04 |
| CN118133437B true CN118133437B (en) | 2024-07-23 |
Family
ID=91244520
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202410571886.9A Active CN118133437B (en) | 2024-05-10 | 2024-05-10 | Ship local structural strength analysis method, device, computer equipment and medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN118133437B (en) |
Families Citing this family (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114443796A (en) * | 2022-02-08 | 2022-05-06 | 云控智行(上海)汽车科技有限公司 | Data processing method and system based on neural network |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111619755A (en) * | 2020-06-09 | 2020-09-04 | 中国船舶科学研究中心 | Hull profile design method based on convolutional neural network |
| CN115730501A (en) * | 2022-12-02 | 2023-03-03 | 兰州大学 | Soil body elastic modulus prediction method based on three-dimensional discrete element model and convolutional neural network |
Family Cites Families (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20230062600A1 (en) * | 2021-08-30 | 2023-03-02 | Accenture Global Solutions Limited | Adaptive design and optimization using physics-informed neural networks |
| WO2023150352A1 (en) * | 2022-02-07 | 2023-08-10 | Nanyang Technological Unversity | Methods for analyses of internal structures and defects in materials using physics-informed neural networks |
| CN117875504A (en) * | 2024-01-15 | 2024-04-12 | 中国人民解放军国防科技大学 | Dam break prediction and model training method and device, electronic equipment and storage medium |
| CN117951850A (en) * | 2024-02-06 | 2024-04-30 | 山东大学 | Hydrogen transport pipeline modeling method and system based on physical information neural network |
-
2024
- 2024-05-10 CN CN202410571886.9A patent/CN118133437B/en active Active
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111619755A (en) * | 2020-06-09 | 2020-09-04 | 中国船舶科学研究中心 | Hull profile design method based on convolutional neural network |
| CN115730501A (en) * | 2022-12-02 | 2023-03-03 | 兰州大学 | Soil body elastic modulus prediction method based on three-dimensional discrete element model and convolutional neural network |
Also Published As
| Publication number | Publication date |
|---|---|
| CN118133437A (en) | 2024-06-04 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Lee et al. | CNN-based image recognition for topology optimization | |
| CN118133437B (en) | Ship local structural strength analysis method, device, computer equipment and medium | |
| Nguyen et al. | Three-dimensional topology optimization of auxetic metamaterial using isogeometric analysis and model order reduction | |
| Wang et al. | IH-GAN: A conditional generative model for implicit surface-based inverse design of cellular structures | |
| Imediegwu et al. | Multiscale structural optimization towards three-dimensional printable structures | |
| Park et al. | Structural optimization based on CAD–CAE integration and metamodeling techniques | |
| Praveen et al. | Low cost PSO using metamodels and inexact pre-evaluation: Application to aerodynamic shape design | |
| US20180330029A1 (en) | Structural optimization of additively manufactured parts accounting for manufacturing-induced states | |
| US20200209832A1 (en) | Method and system for hierarchical multi-scale part design with the aid of a digital computer | |
| US10102316B2 (en) | Virtual reality authoring method | |
| JP7431030B2 (en) | Multi-instance simulation for large-scale environments | |
| Sieger et al. | A comprehensive comparison of shape deformation methods in evolutionary design optimization | |
| CN112861374B (en) | Multi-physical coupling simulation processing method, device and equipment based on pre-controller | |
| Li et al. | Proper orthogonal decomposition with SUPG-stabilized isogeometric analysis for reduced order modelling of unsteady convection-dominated convection-diffusion-reaction problems | |
| JP2020119189A (en) | Fluid analysis system, method for analyzing fluid, and fluid analysis program | |
| Le-Duc et al. | A finite-element-informed neural network for parametric simulation in structural mechanics | |
| Nobari et al. | Range-constrained generative adversarial network: Design synthesis under constraints using conditional generative adversarial networks | |
| Xiang et al. | Real-time stress-based topology optimization via deep learning | |
| Liu et al. | Cluster-based optimization of cellular materials and structures for crashworthiness | |
| Rüttgers et al. | Prediction of acoustic fields using a lattice-Boltzmann method and deep learning | |
| Wang et al. | Functionally graded non-periodic cellular structure design and optimization | |
| EP3647972A1 (en) | A method of training a convolutional neural network to carry out simulation of a physcial product, a corresponding method of carrying out simulation of a physical product and a convolutional neural network | |
| CN116956671A (en) | Finite element simulation method, system, medium and equipment based on deep learning neural network | |
| Gao et al. | Predicting fluid–structure interaction with graph neural networks | |
| CN114627971B (en) | Data processing method and device for solid system |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |