CN115270645A - Design method and system based on ERNIE-DPCNN model - Google Patents
Design method and system based on ERNIE-DPCNN model Download PDFInfo
- Publication number
- CN115270645A CN115270645A CN202211205559.9A CN202211205559A CN115270645A CN 115270645 A CN115270645 A CN 115270645A CN 202211205559 A CN202211205559 A CN 202211205559A CN 115270645 A CN115270645 A CN 115270645A
- Authority
- CN
- China
- Prior art keywords
- model
- ernie
- design
- dpcnn
- target
- 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.)
- Granted
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
- G06F30/20—Design optimisation, verification or simulation
- G06F30/27—Design optimisation, verification or simulation using machine learning, e.g. artificial intelligence, neural networks, support vector machines [SVM] or training a model
-
- 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
- G06N3/08—Learning methods
-
- 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
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Evolutionary Computation (AREA)
- Software Systems (AREA)
- General Physics & Mathematics (AREA)
- Geometry (AREA)
- Artificial Intelligence (AREA)
- General Engineering & Computer Science (AREA)
- Computing Systems (AREA)
- Biomedical Technology (AREA)
- Molecular Biology (AREA)
- Data Mining & Analysis (AREA)
- Computational Linguistics (AREA)
- Biophysics (AREA)
- Mathematical Physics (AREA)
- General Health & Medical Sciences (AREA)
- Life Sciences & Earth Sciences (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Medical Informatics (AREA)
- Computer Hardware Design (AREA)
- Health & Medical Sciences (AREA)
- Computer Graphics (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention provides a design method and a system based on an ERNIE-DPCNN model, wherein the method comprises the steps of obtaining the structural information of an object to be designed and matching a corresponding target three-dimensional model; acquiring historical design cases, constructing a case database according to the historical design cases, and generating a training sample set according to the case database; inputting the training sample set into an ERNIE-DPCNN model for coupling training, and constructing a first three-dimensional calculation model and a second three-dimensional calculation model through Midas GTS; and performing simulation design calculation on the first three-dimensional calculation model and the second three-dimensional calculation model to generate original parameters, and performing coupling analysis on the original parameters through the trained ERNIE-DPCNN model to obtain target design parameters. By the method, the three-dimensional model can be automatically designed, so that the design time of the three-dimensional model can be greatly shortened.
Description
Technical Field
The invention relates to the technical field of three-dimensional model design, in particular to a design method and a system based on an ERNIE-DPCNN model.
Background
Three-dimensional models are polygonal representations of objects, typically displayed by a computer or other video device. The displayed object may be a real-world entity or a fictional object.
At present, the three-dimensional model can also be used in the design of the slide-resistant pile, however, the structural design and the scheme selection of the three-dimensional model are still in the stage of manual operation and control, and for the design of a complex structure, the three-dimensional model is completed from the design to the modeling, so that the time cost consumed is large, and the work efficiency is greatly reduced.
Disclosure of Invention
Based on this, the invention aims to provide a design method and a system based on an ERNIE-DPCNN model, so as to solve the problems that in the prior art, the structural design and the scheme selection of a three-dimensional model are still in the stage of manual operation and control, and for the design of a complex structure, the time cost required for completing the design and the modeling of the three-dimensional model is large.
The first aspect of the embodiment of the invention provides a design method based on an ERNIE-DPCNN model, which comprises the following steps:
obtaining structural information of a current object to be designed, and matching a target three-dimensional model corresponding to the object to be designed according to the structural information;
acquiring historical design cases, and constructing a corresponding case database according to the historical design cases so as to generate a corresponding training sample set according to the case database;
inputting the training sample set into a preset ERNIE-DPCNN model to perform coupling training on the ERNIE-DPCNN model, and constructing a first three-dimensional calculation model corresponding to the current object to be designed and a second three-dimensional calculation model corresponding to the target three-dimensional calculation model through Midas GTS;
and performing simulation design calculation on the first three-dimensional calculation model and the second three-dimensional calculation model in the Midas GTS to generate corresponding original parameters, and performing coupling analysis on the original parameters through a trained ERNIE-DPCNN model to obtain target design parameters corresponding to the target three-dimensional model.
The invention has the beneficial effects that: firstly, obtaining structural information of a current object to be designed, and matching a target three-dimensional model corresponding to the object to be designed according to the structural information; then, acquiring historical design cases, and constructing a corresponding case database according to the historical design cases so as to generate a corresponding training sample set according to the case database; further, inputting the training sample set into a preset ERNIE-DPCNN model to perform coupling training on the ERNIE-DPCNN model, and constructing a first three-dimensional calculation model corresponding to the current object to be designed and a second three-dimensional calculation model corresponding to the target three-dimensional model through Midas GTS; and finally, performing simulation design calculation on the first three-dimensional calculation model and the second three-dimensional calculation model in a Midas GTS to generate corresponding original parameters, and performing coupling analysis on the original parameters through the trained ERNIE-DPCNN model to obtain target design parameters corresponding to the target three-dimensional model. According to the mode, the automatic design of the three-dimensional model can be completed on the premise of not needing manual design and calculation, so that the design time of the three-dimensional model can be greatly shortened, the design efficiency of the three-dimensional model is improved, and the method is suitable for large-scale popularization and use.
Preferably, the step of inputting the training sample set into a preset ERNIE-DPCNN model to perform coupled training on the ERNIE-DPCNN model includes:
when the training sample set is obtained, inputting the training sample set into a pre-training layer in the ERNIE-DPCNN model, and performing coupling calculation on the training sample set through a bidirectional Transformer encoder in the pre-training layer to complete coupling training on the ERNIE-DPCNN model.
Preferably, the step of performing coupling analysis on the original parameters through the trained ERNIE-DPCNN model to obtain target design parameters corresponding to the target three-dimensional model includes:
when the original parameters are acquired, inputting the original parameters into a normalized Softmax function, so that the normalized Softmax function outputs corresponding target design threshold values according to the original parameters;
inputting the target design threshold into a convolution layer in a trained ERNIE-DPCNN model, and performing coupling analysis on the target design threshold in the convolution layer to generate a corresponding parameter matrix;
and acquiring target design parameters corresponding to the target three-dimensional model according to the parameter matrix.
Preferably, the convolutional layer includes an equal-length convolution function, and an expression of the equal-length convolution function is:
wherein E represents the equal-length convolution function, n represents the target design threshold, x (m) represents an initial convolution function, and w (n) represents a window function.
Preferably, after the step of performing coupling analysis on the original parameters through the trained ERNIE-DPCNN model to obtain target design parameters corresponding to the target three-dimensional model, the method further includes:
and generating a corresponding target design report according to the target design parameters, and transmitting the target design report to a display terminal so as to display the target design report on the display terminal in real time.
The second aspect of the embodiment of the present invention provides a design system based on an ERNIE-DPCNN model, where the system includes:
the acquisition module is used for acquiring the structural information of the current object to be designed and matching a target three-dimensional model corresponding to the object to be designed according to the structural information;
the construction module is used for acquiring historical design cases, constructing a corresponding case database according to the historical design cases, and generating a corresponding training sample set according to the case database;
the training module is used for inputting the training sample set into a preset ERNIE-DPCNN model so as to perform coupling training on the ERNIE-DPCNN model, and constructing a first three-dimensional calculation model corresponding to the current object to be designed and a second three-dimensional calculation model corresponding to the target three-dimensional calculation model through Midas GTS;
and the calculation module is used for performing simulation design calculation on the first three-dimensional calculation model and the second three-dimensional calculation model in the Midas GTS to generate corresponding original parameters, and performing coupling analysis on the original parameters through a trained ERNIE-DPCNN model to obtain target design parameters corresponding to the target three-dimensional model.
In the design system based on the ERNIE-DPCNN model, the training module is specifically configured to:
when the training sample set is obtained, inputting the training sample set into a pre-training layer in the ERNIE-DPCNN model, and performing coupling calculation on the training sample set through a bidirectional Transformer encoder in the pre-training layer to complete coupling training on the ERNIE-DPCNN model.
In the design system based on the ERNIE-DPCNN model, the calculation module is specifically configured to:
when the original parameters are acquired, inputting the original parameters into a normalized Softmax function, so that the normalized Softmax function outputs corresponding target design threshold values according to the original parameters;
inputting the target design threshold into a convolutional layer in a trained ERNIE-DPCNN model, and performing coupling analysis on the target design threshold in the convolutional layer to generate a corresponding parameter matrix;
and acquiring target design parameters corresponding to the target three-dimensional model according to the parameter matrix.
In the design system based on the ERNIE-DPCNN model, the convolutional layer includes an equal-length convolution function, and an expression of the equal-length convolution function is as follows:
wherein E represents the equal-length convolution function, n represents the target design threshold, x (m) represents the initial convolution function, and w (n) represents the window function.
In the design system based on the ERNIE-DPCNN model, the design system based on the ERNIE-DPCNN model further includes a display module, and the display module is specifically configured to:
and generating a corresponding target design report according to the target design parameters, and transmitting the target design report to a display terminal so as to display the target design report on the display terminal in real time.
Additional aspects and advantages of the invention will be set forth in part in the description which follows and, in part, will be obvious from the description, or may be learned by practice of the invention.
Drawings
Fig. 1 is a flowchart of a design method based on ERNIE-DPCNN model according to a first embodiment of the present invention;
fig. 2 is a structural block diagram of a design system based on the ERNIE-DPCNN model according to a second embodiment of the present invention.
The following detailed description will further illustrate the invention in conjunction with the above-described figures.
Detailed Description
To facilitate an understanding of the invention, the invention will now be described more fully with reference to the accompanying drawings. Several embodiments of the invention are presented in the drawings. This invention may, however, be embodied in many different forms and should not be construed as limited to the embodiments set forth herein. Rather, these embodiments are provided so that this disclosure will be thorough and complete.
It will be understood that when an element is referred to as being "secured to" another element, it can be directly on the other element or intervening elements may also be present. When an element is referred to as being "connected" to another element, it can be directly connected to the other element or intervening elements may also be present. The terms "vertical," "horizontal," "left," "right," and the like are used herein for purposes of illustration only.
Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs. The terminology used in the description of the invention herein is for the purpose of describing particular embodiments only and is not intended to be limiting of the invention. As used herein, the term "and/or" includes any and all combinations of one or more of the associated listed items.
Referring to fig. 1, a design method based on ERNIE-DPCNN provided in the first embodiment of the present invention is shown, and the design method based on ERNIE-DPCNN provided in this embodiment can complete automatic design of a three-dimensional model without manual design and calculation, so that the design time of the three-dimensional model can be greatly shortened, and the design efficiency of the three-dimensional model is further improved, and the method is suitable for wide popularization and use.
Specifically, the designing method based on the ERNIE-DPCNN model provided in this embodiment specifically includes the following steps:
step S10, obtaining the structural information of the current object to be designed, and matching a target three-dimensional model corresponding to the object to be designed according to the structural information;
specifically, in this embodiment, it should be noted that the design method based on the ERNIE-DPCNN model provided in this embodiment is specifically applied to a scene where a three-dimensional model needs to be set.
In addition, it should be noted that the design method based on the ERNIE-DPCNN model provided in this embodiment is implemented based on a server disposed in the background, and meanwhile, a plurality of algorithms are preset in the server, so that the design efficiency of the three-dimensional model can be effectively improved.
Therefore, in this step, in order to accurately design the three-dimensional model of the object to be designed, the server first obtains the structural information corresponding to the object to be designed, and preferably, in this embodiment, the structural information may include information such as rigidity, size, and material.
Further, in this step, after the server acquires the structure information in real time, the current server may match a target three-dimensional model corresponding to the object to be designed according to the structure information acquired in real time inside the server, that is, match a corresponding three-dimensional model type.
S20, acquiring historical design cases, and constructing a corresponding case database according to the historical design cases so as to generate a corresponding training sample set according to the case database;
further, in this step, it should be noted that after the server acquires the target three-dimensional model, the current server further acquires a required historical design case from a historical case database preset in the current server, on the basis, a corresponding case database is constructed according to the currently acquired historical design case, and further, a corresponding training sample set is generated according to the current case database.
Step S30, inputting the training sample set into a preset ERNIE-DPCNN model to perform coupling training on the ERNIE-DPCNN model, and constructing a first three-dimensional calculation model corresponding to the current object to be designed and a second three-dimensional calculation model corresponding to the target three-dimensional calculation model through Midas GTS;
specifically, in this step, it should be noted that, after the server acquires the training sample set, the current server immediately inputs the acquired training sample set into a preset ERNIE-DPCNN model to perform coupling training on the ERNIE-DPCNN model.
It should be noted that the ERNIE-DPCNN model is a composite model, has a powerful algorithm function, and can perform effective coupling analysis on received data.
Specifically, in this embodiment, it should be noted that the ERNIE-DPCNN model provided in this embodiment includes a convolutional layer and a pooling layer, where the convolutional layer includes a convolutional neural network that can recognize n-gram with predictive property without specifying a word embedding vector for each possible n-gram in advance. The convolution structure can be extended to hierarchical convolutional layers, each layer effectively looking at the longer n-grams in the sentence, making the model sensitive to non-continuous n-grams. In order to avoid the degradation problem caused by the increment of convolution layers, the ERNIE-DPCNN model-based data asset classification and security grading only reserves the convolution function with the convolution kernel of 3 in the first convolution module of one DPCNN model, and replaces zero padding with copy padding to be used as a padding mode in convolution.
In addition, the purpose of the pooling layer is to focus on the most important "features" in the sentence, neglecting their position, and the gradient returned from the network during the training process is used to adjust the parameters of the filter function so that it reinforces the more important part of the data for the network task. Maximum pooling effectively suppresses mean shift due to estimation errors by scanning the maximum value in the matrix region. The method adopts a maximum pooling mode, sets the pooling size to be 3, sets the step length to be 2, and simultaneously fixes the number of the characteristic graphs, so that the calculation time and the data size of each pooling layer of the DPCNN model are both shortened to 1/2 of the original calculation time and data size, and the calculation efficiency is obviously improved.
On the basis, the server further constructs a first three-dimensional calculation model corresponding to the current object to be designed and a second three-dimensional calculation model corresponding to the target three-dimensional model through a preset Midas GTS.
The method is characterized in that the Midas GTS is computer three-dimensional simulation software, the whole process of model damage can be simulated by a finite element strength reduction method arranged in the Midas GTS, the influence of various model size effects on the model stability can be considered, the damage surface of the model does not need to be assumed by human subjectivity when the model is analyzed for stability, the stress and deformation of the model can be calculated based on a physical and mechanical equation built in the software, and all information of displacement, stress, strain and deformation can be obtained from the calculation result.
In this step, it should be noted that the step of inputting the training sample set into a preset ERNIE-DPCNN model to perform coupling training on the ERNIE-DPCNN model includes:
when the training sample set is obtained, inputting the training sample set into a pre-training layer in the ERNIE-DPCNN model, and performing coupling calculation on the training sample set through a bidirectional Transformer encoder in the pre-training layer to complete coupling training on the ERNIE-DPCNN model.
Step S40, performing simulation design calculation on the first three-dimensional calculation model and the second three-dimensional calculation model in the Midas GTS to generate corresponding original parameters, and performing coupling analysis on the original parameters through the trained ERNIE-DPCNN model to obtain target design parameters corresponding to the target three-dimensional model.
Finally, in this step, it should be noted that, the server performs simulation design calculation processing on the first three-dimensional calculation model and the second three-dimensional calculation model in the Midas GTS to generate corresponding original parameters, and further performs coupling analysis on the original parameters through the trained ERNIE-DPCNN model to finally obtain target design parameters corresponding to the target three-dimensional model.
In this step, it should be noted that the step of performing coupling analysis on the original parameters through the trained ERNIE-DPCNN model to obtain target design parameters corresponding to the target three-dimensional model includes:
when the original parameters are acquired, inputting the original parameters into a normalized Softmax function, so that the normalized Softmax function outputs corresponding target design threshold values according to the original parameters;
inputting the target design threshold into a convolution layer in a trained ERNIE-DPCNN model, and performing coupling analysis on the target design threshold in the convolution layer to generate a corresponding parameter matrix;
and acquiring target design parameters corresponding to the target three-dimensional model according to the parameter matrix.
In this embodiment, it should be noted that, the convolutional layer includes an equal-length convolution function, and an expression of the equal-length convolution function is:
wherein E represents the equal-length convolution function, n represents the target design threshold, x (m) represents the initial convolution function, and w (n) represents the window function.
In this embodiment, it should be further noted that, after the step of performing coupling analysis on the original parameters through the trained ERNIE-DPCNN model to obtain the target design parameters corresponding to the target three-dimensional model, the method further includes:
and generating a corresponding target design report according to the target design parameters, and transmitting the target design report to a display terminal so as to display the target design report on the display terminal in real time, thereby enabling a worker to observe the required target design parameters in real time.
When the method is used, the structural information of the current object to be designed is obtained, and a target three-dimensional model corresponding to the object to be designed is matched according to the structural information; acquiring historical design cases, constructing a corresponding case database according to the historical design cases, and generating a corresponding training sample set according to the case database; further, inputting the training sample set into a preset ERNIE-DPCNN model to perform coupling training on the ERNIE-DPCNN model, and constructing a first three-dimensional calculation model corresponding to the current object to be designed and a second three-dimensional calculation model corresponding to the target three-dimensional model through Midas GTS; and finally, performing simulation design calculation on the first three-dimensional calculation model and the second three-dimensional calculation model in a Midas GTS to generate corresponding original parameters, and performing coupling analysis on the original parameters through the trained ERNIE-DPCNN model to obtain target design parameters corresponding to the target three-dimensional model. According to the mode, the automatic design of the three-dimensional model can be completed on the premise that manual design and calculation are not needed, so that the design time of the three-dimensional model can be greatly shortened, the design efficiency of the three-dimensional model is improved, and the method is suitable for large-scale popularization and use.
It should be noted that the above implementation process is only for illustrating the applicability of the present application, but this does not represent that the designing method based on the ERNIE-DPCNN model of the present application has only one implementation flow, and on the contrary, the designing method based on the ERNIE-DPCNN model of the present application can be incorporated into the feasible embodiments of the present application as long as it can be implemented.
In summary, the designing method based on the ERNIE-DPCNN model provided in the embodiments of the present invention can complete the automatic design of the three-dimensional model without manual design and calculation, so as to greatly shorten the design time of the three-dimensional model, thereby improving the design efficiency of the three-dimensional model, and being suitable for large-scale popularization and use.
Referring to fig. 2, a design system based on ERNIE-DPCNN model according to a second embodiment of the present invention is shown, the system including:
the acquisition module 12 is configured to acquire structural information of a current object to be designed, and match a target three-dimensional model corresponding to the object to be designed according to the structural information;
the construction module 22 is configured to obtain historical design cases, construct a corresponding case database according to the historical design cases, and generate a corresponding training sample set according to the case database;
the training module 32 is configured to input the training sample set into a preset ERNIE-DPCNN model, so as to perform coupling training on the ERNIE-DPCNN model, and construct a first three-dimensional computation model corresponding to the current object to be designed and a second three-dimensional computation model corresponding to the target three-dimensional model through Midas GTS;
a calculating module 42, configured to perform simulation design calculation on the first three-dimensional calculation model and the second three-dimensional calculation model in the Midas GTS to generate corresponding original parameters, and perform coupling analysis on the original parameters through the trained ERNIE-DPCNN model to obtain target design parameters corresponding to the target three-dimensional model.
In the design system based on the ERNIE-DPCNN model, the training module 32 is specifically configured to:
when the training sample set is obtained, inputting the training sample set into a pre-training layer in the ERNIE-DPCNN model, and performing coupling calculation on the training sample set through a bidirectional Transformer encoder in the pre-training layer to complete coupling training on the ERNIE-DPCNN model.
In the design system based on the ERNIE-DPCNN model, the calculating module 42 is specifically configured to:
when the original parameters are acquired, inputting the original parameters into a normalized Softmax function, so that the normalized Softmax function outputs corresponding target design threshold values according to the original parameters;
inputting the target design threshold into a convolutional layer in a trained ERNIE-DPCNN model, and performing coupling analysis on the target design threshold in the convolutional layer to generate a corresponding parameter matrix;
and acquiring target design parameters corresponding to the target three-dimensional model according to the parameter matrix.
In the design system based on the ERNIE-DPCNN model, the convolutional layer includes an equal-length convolution function, and an expression of the equal-length convolution function is as follows:
wherein E represents the equal-length convolution function, n represents the target design threshold, x (m) represents the initial convolution function, and w (n) represents the window function.
In the design system based on the ERNIE-DPCNN model, the design system based on the ERNIE-DPCNN model further includes a display module 52, and the display module 52 is specifically configured to:
and generating a corresponding target design report according to the target design parameters, and transmitting the target design report to a display terminal so as to display the target design report on the display terminal in real time.
In summary, the designing method and system based on the ERNIE-DPCNN model provided in the embodiments of the present invention can complete the automatic design of the three-dimensional model without manual design and calculation, so as to greatly shorten the design time of the three-dimensional model, further improve the design efficiency of the three-dimensional model, eliminate the potential safety hazard, and are suitable for large-scale popularization and use.
The above modules may be functional modules or program modules, and may be implemented by software or hardware. For a module implemented by hardware, the above modules may be located in the same processor; or the modules can be respectively positioned in different processors in any combination.
The logic and/or steps represented in the flowcharts or otherwise described herein, such as an ordered listing of executable instructions that can be considered to implement logical functions, can be embodied in any computer-readable medium for use by or in connection with an instruction execution system, apparatus, or device, such as a computer-based system, processor-containing system, or other system that can fetch the instructions from the instruction execution system, apparatus, or device and execute the instructions. For the purposes of this description, a "computer-readable medium" can be any means that can contain, store, communicate, propagate, or transport the program for use by or in connection with the instruction execution system, apparatus, or device.
More specific examples (a non-exhaustive list) of the computer-readable medium would include the following: an electrical connection (electronic device) having one or more wires, a portable computer diskette (magnetic device), a Random Access Memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber device, and a portable compact disc read-only memory (CDROM). Additionally, the computer-readable medium could even be paper or another suitable medium upon which the program is printed, as the program can be electronically captured, via for instance optical scanning of the paper or other medium, then compiled, interpreted or otherwise processed in a suitable manner if necessary, and then stored in a computer memory.
It should be understood that portions of the present invention may be implemented in hardware, software, firmware, or a combination thereof. In the above embodiments, the various steps or methods may be implemented in software or firmware stored in memory and executed by a suitable instruction execution system. For example, if implemented in hardware, as in another embodiment, any one or combination of the following techniques, which are known in the art, may be used: a discrete logic circuit having a logic gate circuit for implementing a logic function on a data signal, an application specific integrated circuit having an appropriate combinational logic gate circuit, a Programmable Gate Array (PGA), a Field Programmable Gate Array (FPGA), or the like.
In the description herein, references to the description of the term "one embodiment," "some embodiments," "an example," "a specific example," or "some examples," etc., mean that a particular feature, structure, material, or characteristic described in connection with the embodiment or example is included in at least one embodiment or example of the invention. In this specification, the schematic representations of the terms used above do not necessarily refer to the same embodiment or example. Furthermore, the particular features, structures, materials, or characteristics described may be combined in any suitable manner in any one or more embodiments or examples.
The above-mentioned embodiments only express several embodiments of the present invention, and the description thereof is specific and detailed, but not to be understood as limiting the scope of the present invention. It should be noted that various changes and modifications can be made by those skilled in the art without departing from the spirit of the invention, and these changes and modifications are all within the scope of the invention. Therefore, the protection scope of the present patent shall be subject to the appended claims.
Claims (10)
1. A design method based on ERNIE-DPCNN model, which is characterized in that the method comprises:
obtaining structural information of a current object to be designed, and matching a target three-dimensional model corresponding to the object to be designed according to the structural information, wherein the structural information comprises material, rigidity and size;
acquiring a historical design case, and constructing a corresponding case database according to the historical design case to generate a corresponding training sample set according to the case database, wherein the training sample set comprises the type, the design standard and the quality standard of a three-dimensional model;
inputting the training sample set into a preset ERNIE-DPCNN model to perform coupling training on the ERNIE-DPCNN model, and constructing a first three-dimensional calculation model corresponding to the current object to be designed and a second three-dimensional calculation model corresponding to the target three-dimensional model through Midas GTS, wherein the first three-dimensional calculation model and the second three-dimensional calculation model are unique;
and performing simulation design calculation on the first three-dimensional calculation model and the second three-dimensional calculation model in the Midas GTS to generate corresponding original parameters, and performing coupling analysis on the original parameters through a trained ERNIE-DPCNN model to obtain target design parameters corresponding to the target three-dimensional model, wherein the target design parameters comprise model materials, model proportion and model size.
2. The ERNIE-DPCNN model-based design method of claim 1, wherein: the step of inputting the training sample set into a preset ERNIE-DPCNN model to perform coupled training on the ERNIE-DPCNN model comprises the following steps:
when the training sample set is obtained, inputting the training sample set into a pre-training layer in the ERNIE-DPCNN model, and performing coupling calculation on the training sample set through a bidirectional Transformer encoder in the pre-training layer to complete coupling training on the ERNIE-DPCNN model.
3. The ERNIE-DPCNN model-based design method of claim 1, wherein: the step of performing coupling analysis on the original parameters through the trained ERNIE-DPCNN model to obtain target design parameters corresponding to the target three-dimensional model comprises the following steps:
when the original parameters are acquired, inputting the original parameters into a normalized Softmax function, so that the normalized Softmax function outputs corresponding target design threshold values according to the original parameters;
inputting the target design threshold into a convolutional layer in a trained ERNIE-DPCNN model, and performing coupling analysis on the target design threshold in the convolutional layer to generate a corresponding parameter matrix;
and acquiring target design parameters corresponding to the target three-dimensional model according to the parameter matrix.
4. The ERNIE-DPCNN model-based design method of claim 3, characterized in that: the convolutional layer comprises an equal length convolution function, and the expression of the equal length convolution function is as follows:
wherein E represents the equal-length convolution function, n represents the target design threshold, x (m) represents the initial convolution function, and w (n) represents the window function.
5. The ERNIE-DPCNN model-based design method of claim 1, wherein: after the step of performing coupling analysis on the original parameters through the trained ERNIE-DPCNN model to obtain target design parameters corresponding to the target three-dimensional model, the method further includes:
and generating a corresponding target design report according to the target design parameters, and transmitting the target design report to a display terminal so as to display the target design report on the display terminal in real time.
6. A design system based on ERNIE-DPCNN model, the system comprising:
the system comprises an acquisition module, a calculation module and a display module, wherein the acquisition module is used for acquiring the structural information of a current object to be designed and matching a target three-dimensional model corresponding to the object to be designed according to the structural information, and the structural information comprises material, rigidity and size;
the construction module is used for acquiring historical design cases, constructing a corresponding case database according to the historical design cases, and generating a corresponding training sample set according to the case database, wherein the training sample set comprises the type, the design standard and the quality standard of a three-dimensional model;
the training module is used for inputting the training sample set into a preset ERNIE-DPCNN model so as to perform coupling training on the ERNIE-DPCNN model, and constructing a first three-dimensional calculation model corresponding to the current object to be designed and a second three-dimensional calculation model corresponding to the target three-dimensional model through Midas GTS, wherein the first three-dimensional calculation model and the second three-dimensional calculation model are unique;
a calculating module, configured to perform simulation design calculation on the first three-dimensional calculation model and the second three-dimensional calculation model in the Midas GTS to generate corresponding original parameters, and perform coupling analysis on the original parameters through a trained ERNIE-DPCNN model to obtain target design parameters corresponding to the target three-dimensional model, where the target design parameters include a model material, a model proportion, and a model size.
7. The ERNIE-DPCNN model-based design system of claim 6, wherein: the training module is specifically configured to:
when the training sample set is obtained, inputting the training sample set into a pre-training layer in the ERNIE-DPCNN model, and performing coupling calculation on the training sample set through a bidirectional Transformer encoder in the pre-training layer to complete coupling training on the ERNIE-DPCNN model.
8. The ERNIE-DPCNN model-based design system of claim 6, wherein: the calculation module is specifically configured to:
when the original parameters are obtained, inputting the original parameters into a normalized Softmax function so that the normalized Softmax function outputs corresponding target design threshold values according to the original parameters;
inputting the target design threshold into a convolution layer in a trained ERNIE-DPCNN model, and performing coupling analysis on the target design threshold in the convolution layer to generate a corresponding parameter matrix;
and acquiring target design parameters corresponding to the target three-dimensional model according to the parameter matrix.
9. The ERNIE-DPCNN model-based design system of claim 8, wherein: the convolutional layer comprises an equal length convolution function, and the expression of the equal length convolution function is as follows:
wherein E represents the equal-length convolution function, n represents the target design threshold, x (m) represents an initial convolution function, and w (n) represents a window function.
10. The ERNIE-DPCNN model-based design system of claim 6, wherein: the design system based on the ERNIE-DPCNN model further comprises a display module, and the display module is specifically used for:
and generating a corresponding target design report according to the target design parameters, and transmitting the target design report to a display terminal so as to display the target design report on the display terminal in real time.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211205559.9A CN115270645B (en) | 2022-09-30 | 2022-09-30 | Design method and system based on ERNIE-DPCNN model |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211205559.9A CN115270645B (en) | 2022-09-30 | 2022-09-30 | Design method and system based on ERNIE-DPCNN model |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN115270645A true CN115270645A (en) | 2022-11-01 |
| CN115270645B CN115270645B (en) | 2022-12-27 |
Family
ID=83757913
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202211205559.9A Active CN115270645B (en) | 2022-09-30 | 2022-09-30 | Design method and system based on ERNIE-DPCNN model |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115270645B (en) |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106919763A (en) * | 2017-03-07 | 2017-07-04 | 上海波客实业有限公司 | A kind of dimensionally-optimised method of product structure |
| CN110765677A (en) * | 2019-08-26 | 2020-02-07 | 西安理工大学 | Modeling method of high-precision and rapid three-dimensional geological model finite element model |
| CN112562069A (en) * | 2020-12-24 | 2021-03-26 | 北京百度网讯科技有限公司 | Three-dimensional model construction method, device, equipment and storage medium |
| US20210201182A1 (en) * | 2020-09-29 | 2021-07-01 | Beijing Baidu Netcom Science And Technology Co., Ltd. | Method and apparatus for performing structured extraction on text, device and storage medium |
| CN114239939A (en) * | 2021-12-08 | 2022-03-25 | 东北大学 | Method for criminal prediction by incorporating auxiliary knowledge |
| CN114780724A (en) * | 2022-04-11 | 2022-07-22 | 湖南工商大学 | Case classification method and device, computer equipment and storage medium |
| CN115114749A (en) * | 2022-07-15 | 2022-09-27 | 重庆三航新材料技术研究院有限公司 | Template, flow and automatic model design method for pipeline system |
-
2022
- 2022-09-30 CN CN202211205559.9A patent/CN115270645B/en active Active
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106919763A (en) * | 2017-03-07 | 2017-07-04 | 上海波客实业有限公司 | A kind of dimensionally-optimised method of product structure |
| CN110765677A (en) * | 2019-08-26 | 2020-02-07 | 西安理工大学 | Modeling method of high-precision and rapid three-dimensional geological model finite element model |
| US20210201182A1 (en) * | 2020-09-29 | 2021-07-01 | Beijing Baidu Netcom Science And Technology Co., Ltd. | Method and apparatus for performing structured extraction on text, device and storage medium |
| CN112562069A (en) * | 2020-12-24 | 2021-03-26 | 北京百度网讯科技有限公司 | Three-dimensional model construction method, device, equipment and storage medium |
| CN114239939A (en) * | 2021-12-08 | 2022-03-25 | 东北大学 | Method for criminal prediction by incorporating auxiliary knowledge |
| CN114780724A (en) * | 2022-04-11 | 2022-07-22 | 湖南工商大学 | Case classification method and device, computer equipment and storage medium |
| CN115114749A (en) * | 2022-07-15 | 2022-09-27 | 重庆三航新材料技术研究院有限公司 | Template, flow and automatic model design method for pipeline system |
Non-Patent Citations (2)
| Title |
|---|
| YUXUAN LUO等: "Text Matching application for integration between industry and education", 《2022 14TH INTERNATIONAL CONFERENCE ON MEASURING TECHNOLOGY AND MECHATRONICS AUTOMATION (ICMTMA)》 * |
| 牛玉婷等: "基于改进ERNIE-DPCNN模型的中文文本分类", 《江苏师范大学学报(自然科学版)》 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN115270645B (en) | 2022-12-27 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111681162B (en) | Defect sample generation method and device, electronic equipment and storage medium | |
| CN108959794B (en) | Structural frequency response dynamic model correction method based on deep learning | |
| CN110059330B (en) | Method and system for authoring simulated scenes | |
| CN111401472B (en) | Infrared target classification method and device based on deep convolutional neural network | |
| CN113158292B (en) | Component matching method, engineering amount calculating device and electronic equipment | |
| CN113569852A (en) | Training method and device of semantic segmentation model, electronic equipment and storage medium | |
| CN112420125A (en) | Molecular attribute prediction method and device, intelligent equipment and terminal | |
| CN116229066A (en) | Portrait segmentation model training method and related device | |
| Ovseenko et al. | The possibility of artificial neural network application in prototyping in instrument making industry | |
| US20200394275A1 (en) | Design of microstructures using generative adversarial networks | |
| CN115270645B (en) | Design method and system based on ERNIE-DPCNN model | |
| Zimmerling et al. | A meta-model based approach for rapid formability estimation of continuous fibre reinforced components | |
| CN112926052A (en) | Deep learning model security vulnerability testing and repairing method, device and system based on genetic algorithm | |
| CN117217020A (en) | Industrial model construction method and system based on digital twin | |
| CN111581869A (en) | Method, device and storage medium for establishing bolt connection | |
| EP3499468A1 (en) | Systems and methods for finite element mesh repair | |
| US20230215093A1 (en) | Method and system for 3d modeling based on irregular-shaped sketch | |
| WO2023056501A1 (en) | Harmonizing diffusion tensor images using machine learning | |
| CN113516670B (en) | Feedback attention-enhanced non-mode image segmentation method and device | |
| Xin et al. | Prediction of the buckling mode of cylindrical composite shells with imperfections using FEM-based deep learning approach | |
| CN111241614B (en) | Engineering structure load inversion method based on condition generation confrontation network model | |
| Monaco et al. | Simulation of waves propagation into composites thin shells by FEM methodologies for training of deep neural networks aimed at damage reconstruction | |
| CN113743437A (en) | Training classification models using distinct training sources and inference engine employing same | |
| CN117456611B (en) | Virtual character training method and system based on artificial intelligence | |
| CN116311478B (en) | Training method of face binding model, face binding method, device and equipment |
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 |