A soil-rock dam generative intelligent design method, device, equipment and medium

Abstract

The application discloses a soil-rock dam generative intelligent design method and device, equipment and medium, and relates to the soil-rock dam scheme design field. The method comprises the following steps: constructing a heterogeneous data set for training based on the design point requirements of different dam types; inputting the heterogeneous data set into a generator of a generative adversarial network; constructing a finite element calculation data set according to a soil-rock dam design scheme output by the generator based on dam body safety specification requirements of different dam types; inputting the finite element calculation data set into an evaluator of the generative adversarial network; taking dam body safety of a soil-rock dam design scheme output by the evaluator as a constraint, and taking a minimum loss value of a working state output by the evaluator and a calculated working state as an objective of adversarial training of the generative adversarial network; and training the generator to be a soil-rock dam generative design model. The soil-rock dam generative design model is used to generate a soil-rock dam design scheme to be designed, and the application can realize efficient and high-quality design of the soil-rock dam.

Description

Technical Field

[0001] This application relates to the field of earth-rock dam design, and in particular to a generative intelligent design method, device, equipment and medium for earth-rock dams. Background Technology

[0002] Dams are the backbone of hydropower engineering construction. Earth-rock dams, as one of the important dam types, have advantages in safety, economy, energy conservation, and environmental protection. Face-faced dams and core-wall dams are two major advantageous dam types of earth-rock dams, widely used in conventional hydropower and pumped storage power stations. In recent years, intelligent design and operation and maintenance of earth-rock dams have become an inevitable choice for achieving high-quality and efficient engineering design and meeting the development needs of intelligent construction.

[0003] The design of earth-rock dam structures is currently mainly in the parametric structural design stage based on high-precision modeling algorithms, and digital design technology still needs further development. At present, conventional earth-rock dam design schemes are still dominated by manual design, with insufficient automation and intelligent design capabilities. The earth-rock dam design model, which heavily relies on engineers, faces developmental challenges such as difficulty in transferring experience and improving efficiency. Currently, the design results (drawings, documents, experience, etc.) of completed panel dams, as well as a large amount of design knowledge and experience, are difficult to effectively utilize, and the accumulated excellent experience in the field is difficult to pass on. At the same time, due to the limited productivity of engineers, the manual design model, which heavily relies on engineers, is unable to further improve design efficiency. It is evident that the design model, which over-relies on manual labor, is gradually facing difficulties in improving efficiency, transferring experience, and reusing design results, urgently requiring intelligent transformation. Therefore, developing artificial intelligence (AI) design theories, methods, technologies, and system platforms for earth-rock dams with self-learning capabilities holds promise for solving these problems. Summary of the Invention

[0004] The purpose of this application is to provide a generative intelligent design method, device, equipment and medium for earth-rock dams, which can realize efficient and high-quality design of earth-rock dams.

[0005] To achieve the above objectives, this application provides the following solution.

[0006] In the first aspect, this application provides a generative intelligent design method for earth-rock dams, comprising the following steps.

[0007] Obtain design text knowledge for different types of earth-rock dams.

[0008] Based on the design text knowledge, design requirements and dam safety specifications for different dam types are generated.

[0009] The heterogeneous data set for training is constructed based on the design key point requirements of different dam types; the heterogeneous data set is determined according to text data and image data; the text data includes dam height, geological exploration data and dam material physical and mechanical parameters; and the image data includes foundation surface images and dam body contour images.

[0010] The heterogeneous data set for training is input into a generator of a generative adversarial network, and a finite element calculation data set is constructed according to a dam design scheme output by the generator based on dam body safety specification requirements of different dam types; the finite element calculation data set includes the dam design scheme and corresponding working behavior calculated by using a finite element analysis method.

[0011] The finite element calculation data set is input into an evaluator of the generative adversarial network, the dam body safety of the dam design scheme output by the evaluator meets the dam body safety specification requirements as a constraint, and the working behavior output by the evaluator and the loss value of the working behavior calculated in the finite element calculation data set are minimized as an objective to perform adversarial training on the generative adversarial network, so as to obtain a trained generative adversarial network; wherein the trained generator is a soil and rock dam generative design model, and the trained evaluator is a dam body working behavior evaluation model.

[0012] A heterogeneous data set to be designed is constructed based on the design key point requirements required by the soil and rock dam to be designed.

[0013] The heterogeneous data set to be designed is input into the soil and rock dam generative design model to generate a soil and rock dam design scheme to be designed.

[0014] In a second aspect, the present application provides a soil and rock dam generative intelligent design device, comprising the following modules.

[0015] A design text knowledge acquisition module is configured to acquire design text knowledge of different dam types of soil and rock dams.

[0016] A requirement generation module is configured to generate design key point requirements and dam body safety specification requirements of different dam types according to the design text knowledge.

[0017] A first heterogeneous data set construction module is configured to construct a heterogeneous data set for training based on design key point requirements of different dam types; the heterogeneous data set is determined according to text data and image data; the text data includes dam height, geological exploration data and dam material physical and mechanical parameters; and the image data includes foundation surface images and dam body contour images.

[0018] A finite element calculation dataset construction module is configured to input the heterogeneous dataset for training into a generator of a generative adversarial network, and construct a finite element calculation dataset according to a design scheme of an earth-rock dam output by the generator based on dam body safety specification requirements of different dam types; the finite element calculation dataset includes the design scheme of the earth-rock dam and corresponding working conditions calculated by using a finite element analysis method.

[0019] An adversarial training module is configured to input the finite element calculation dataset into an evaluator of the generative adversarial network, and perform adversarial training on the generative adversarial network with the dam body safety of the design scheme of the earth-rock dam output by the evaluator meeting dam body safety specification requirements as a constraint and a loss value of the working conditions output by the evaluator and the working conditions calculated in the finite element calculation dataset being minimized as an objective, to obtain a trained generative adversarial network; wherein the trained generator is an earth-rock dam generative design model, and the trained evaluator is a dam body working condition evaluation model.

[0020] A second heterogeneous dataset construction module is configured to construct a heterogeneous dataset to be designed based on design point requirements required by an earth-rock dam to be designed.

[0021] An earth-rock dam design scheme generation module is configured to input the heterogeneous dataset to be designed into the earth-rock dam generative design model, and generate a design scheme of the earth-rock dam to be designed.

[0022] In a third aspect, the present application provides a computer device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, and the processor executes the computer program to implement the earth-rock dam generative intelligent design method in any one of the above aspects.

[0023] In a fourth aspect, the present application provides a computer readable storage medium having a computer program stored thereon, and the computer program is executed by a processor to implement the earth-rock dam generative intelligent design method in any one of the above aspects.

[0024] According to the specific embodiments provided in the present application, the present application has the following technical effects.

[0025] The application provides a soil and rock dam generative intelligent design method, device, equipment and medium. A heterogeneous data set for training is constructed based on design point requirements of different dam types, a generator of a generative adversarial network is trained, a finite element calculation data set is constructed based on dam body safety specification requirements of different dam types and a soil and rock dam design scheme output by the generator, an evaluator in the generative adversarial network is trained, in the training process, dam body safety of the soil and rock dam design scheme output by the evaluator meets dam body safety specification requirements as a constraint, and adversarial training is realized with the loss value of the working state output by the evaluator and the working state calculated in the finite element calculation data set being minimum as a target, so that a soil and rock dam generative design model and a dam body working state evaluation model are obtained. The application considers the working state of the soil and rock dam design scheme, adversarially trains the soil and rock dam generative design model, and on the basis of automatic design of the soil and rock dam design scheme, intelligent discrimination is performed on the working state of the generated soil and rock dam design scheme, so that it is ensured that the design generation result meets engineering safety and stability requirements, and efficient and high-quality design of the soil and rock dam is realized. BRIEF DESCRIPTION OF DRAWINGS

[0026] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the drawings needed in the embodiments will be briefly introduced below. Obviously, the drawings in the following description are only some embodiments of the present application, and other drawings can be obtained by those skilled in the art without creative labor.

[0027] Figure 1 An application environment diagram of a soil and rock dam generative intelligent design method in an embodiment of the present application.

[0028] Figure 2 A flowchart of a soil and rock dam generative intelligent design method provided in an embodiment of the present application.

[0029] Figure 3 An adversarial training process diagram provided in an embodiment of the present application.

[0030] Figure 4 A structure diagram of a generative adversarial network provided in an embodiment of the present application.

[0031] Figure 5 A functional module diagram of a soil and rock dam generative intelligent design device provided in another embodiment of the present application.

[0032] Figure 6 A structure diagram of a computer device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0033] With reference to the drawings, the technical solutions in the embodiments of the present application will be clearly and completely described below. Obviously, the described embodiments are only a part of the embodiments of the present application, but not all the embodiments of the present application. Based on the embodiments in the present application, all the other embodiments obtained by a person of ordinary skill in the art without creative work fall within the protection scope of the present application.

[0034] The above purposes, features and advantages of the present application will be more apparent and understandable. The present application will be further described in detail below with reference to the drawings and specific embodiments.

[0035] The rockfill dam generative intelligent design method provided by the embodiments of the present application can be applied to an application environment as shown in the figure. Figure 1 The terminal 102 communicates with the server 104 through a network. The data storage system can store data required to be processed by the server 104. The data storage system can be separately arranged, or integrated on the server 104, or placed on the cloud or other servers. The terminal 102 can send the design text knowledge of different dam types of earth-rockfill dams to the server 104. After receiving the design text knowledge of different dam types of earth-rockfill dams, the server generates design key point requirements and dam body safety specification requirements of different dam types according to the design text knowledge. Based on the design key point requirements of different dam types, a heterogeneous data set for training is constructed. The heterogeneous data set for training is input into a generator of a generative adversarial network. Based on the dam body safety specification requirements of different dam types, a finite element calculation data set is constructed according to the earth-rockfill dam design scheme output by the generator. The finite element calculation data set is input into an evaluator of the generative adversarial network. The dam body safety of the earth-rockfill dam design scheme output by the evaluator meets the dam body safety specification requirements as a constraint. The loss value of the working state calculated in the finite element calculation data set and the working state output by the evaluator is minimized as an objective to adversarially train the generative adversarial network, so as to obtain a trained generative adversarial network. The trained generator is used as an earth-rockfill dam generative design model, and the trained evaluator is used as a dam body working state evaluation model. Based on the design key point requirements required by the to-be-designed earth-rockfill dam, a to-be-designed heterogeneous data set is constructed. The to-be-designed heterogeneous data set is input into the earth-rockfill dam generative design model to generate a to-be-designed earth-rockfill dam design scheme.

[0036] The server 104 can feed back the obtained to-be-designed earth-rock dam design scheme to the terminal 102. In addition, in some embodiments, the earth-rock dam generative intelligent design method can also be implemented by the server 104 or the terminal 102 alone, for example, the terminal 102 can directly process the design text knowledge of different dam types of the earth-rock dam, or the server 104 can obtain the design text knowledge of different dam types of the earth-rock dam from the data storage system and process the design text knowledge of different dam types of the earth-rock dam.

[0037] The terminal 102 can be, but is not limited to, various desktop computers, notebook computers, smart phones, tablet computers, Internet of Things devices, and portable wearable devices. The Internet of Things device can be a smart speaker, a smart television, a smart air conditioner, a smart vehicle-mounted device, etc. The portable wearable device can be a smart watch, a smart bracelet, a head-mounted device, etc. The server 104 can be implemented by a stand-alone server or a server cluster composed of multiple servers, and can also be a cloud server.

[0038] In an exemplary embodiment, as shown in Figure 2 , an earth-rock dam generative intelligent design method is provided, which is executed by a computer device, specifically, can be executed by a terminal or a server or the like computer device alone, or can be executed by a terminal and a server together. In the embodiment of the present application, the method is applied to the server 104 in the Figure 1 , and includes the following steps 201 to 207.

[0039] Step 201, obtaining design text knowledge of different dam types of the earth-rock dam.

[0040] The design text knowledge includes the earth-rock dam design specification, the hydraulic structure design manual, and the typical engineering design report, etc.

[0041] Step 202, generating design point requirements and dam body safety specification requirements of different dam types according to the design text knowledge.

[0042] Step 203, constructing a heterogeneous data set for training based on the design point requirements of different dam types; the heterogeneous data set is determined according to text data and image data; the text data includes dam height, geological exploration data, and dam material physical and mechanical parameters; the image data includes foundation surface images and dam body contour images.

[0043] Step 204, inputting the heterogeneous data set for training into a generator of a generative adversarial network, and constructing a finite element calculation data set according to an earth-rock dam design scheme output by the generator based on the dam body safety specification requirements of different dam types; the finite element calculation data set includes the earth-rock dam design scheme and corresponding working behavior calculated by the finite element analysis method.

[0044] Step 205, input the finite element calculation data set into the evaluator of the generative adversarial network, output the dam safety of the earth-rock dam design scheme by the evaluator as a constraint that meets the dam safety specification requirements, and perform adversarial training on the generative adversarial network with the loss value of the working state calculated in the finite element calculation data set as the target, to obtain a trained generative adversarial network; wherein the trained generator is an earth-rock dam generative design model, and the trained evaluator is a dam working state evaluation model. The evaluator is an artificial neural network surrogate model.

[0045] Wherein, the adversarial training: input the automatically generated earth-rock dam design scheme into the evaluator, determine the loss function according to the evaluation function, and propagate back to the generator to realize the adversarial training of the earth-rock dam generative design model.

[0046] Step 206, construct a heterogeneous data set to be designed based on the design point requirements required by the earth-rock dam to be designed.

[0047] Step 207, input the heterogeneous data set to be designed into the earth-rock dam generative design model to generate an earth-rock dam design scheme to be designed.

[0048] Implementing the above steps 201 to 207 can realize efficient and high-quality design of earth-rock dams.

[0049] In another exemplary embodiment of the present application, step 202 specifically includes the following content.

[0050] (1) The natural language processing (NLP) technology is used to completely electronicize, standardize and unify the data format of the design text knowledge, and the processed data is obtained, and the purpose is to construct an earth-rock dam design field knowledge graph.

[0051] (2) The earth-rock dam design field knowledge graph is constructed according to the processed data.

[0052] (3) Based on the earth-rock dam design field knowledge graph, the large model is trained by using the prompt word engineering technology to learn the field text knowledge, so as to accurately and efficiently identify the key design text features, and realize the automatic learning of the key physical and mechanical parameters and specification requirements in the text report. Therefore, the design point requirements and dam safety specification requirements of different dam types are generated according to the learned large model.

[0053] In another exemplary embodiment of the present application, step 203 specifically includes the following content.

[0054] (1) Collecting the engineering data of the built earth-rock dam; the engineering data includes dam height, geological exploration data, foundation surface image, and dam material physical and mechanical parameters.

[0055] (2) Based on the design key point requirements of different dam types, the engineering data of the built earth-rock dam is used to construct the foundation surface-dam slope-dam height earth-rock dam contour image matrix and material parameter vector for training.

[0056] (3) The heterogeneous data set for training is determined according to the foundation surface-dam slope-dam height earth-rock dam contour image matrix and material parameter vector for training.

[0057] Subsequently, based on the heterogeneous data set, an AI model architecture is used as a generator to autonomously learn the partition characteristics of the earth-rock dam core wall, panel, cushion layer, transition layer, filter layer, main rockfill area, secondary rockfill area, drainage area, and increment area, and an AI model architecture generated by profile partition design is studied to realize the generation of corresponding design images by heterogeneous data fusion; an automatic vectorization extraction method of design images is proposed, which automatically converts image data that is difficult to use in design into profile partition design parameters required for parametric modeling and design optimization.

[0058] In another exemplary embodiment of the present application, step 204 specifically includes the following.

[0059] (1) Based on the dam body safety specification requirements of different dam types, the earth-rock dam under the earth-rock dam design scheme and dam material physical and mechanical parameters output by the generator is subjected to finite element analysis, and the working state of the corresponding earth-rock dam is calculated.

[0060] (2) The finite element calculation data set is established from the earth-rock dam design scheme output by the generator and the calculated corresponding working state; the finite element calculation data set is used to reflect the dam body seepage stability, deformation stability, seismic safety, and dam slope stability performance.

[0061] In another exemplary embodiment of the present application, steps 204 and 205 realize adversarial training, which adopts a multiple iteration method, as shown in Figure 3 For the nth iteration, the process includes the following steps.

[0062] Step 301 inputs the nth iteration of the heterogeneous data set for training into the generator in the generative adversarial network, and the generator outputs the nth iteration of the earth-rock dam design scheme.

[0063] Step 302 constructs the nth iteration of the finite element calculation data set according to the earth-rock dam design scheme output by the generator in the nth iteration.

[0064] Step 303, input the finite element calculation data set of the nth iteration into the evaluator in the generative adversarial network, and the evaluator outputs the dam safety of the earth-rock dam design scheme of the nth iteration, and calculates the loss value of the working state of the nth iteration output by the evaluator and the working state calculated in the finite element calculation data set.

[0065] Step 304, judge whether the dam safety of the earth-rock dam design scheme of the nth iteration meets the dam safety specification requirements, and whether the loss value of the nth iteration reaches the set loss value range. If yes, execute step 305, take the generator trained in the nth iteration as the earth-rock dam generative design model, and take the evaluator trained in the nth iteration as the dam working state evaluation model. Otherwise, update the iteration number and return to step 301.

[0066] In another exemplary embodiment of the present application, referring to Figure 4 , the generator G includes a profile contour generator G1 and a profile partition generator G2. The evaluator EV includes a specification constraint discriminator N and a mechanical performance evaluator E.

[0067] The profile contour generator G1 is used to generate an earth-rock dam profile contour according to a heterogeneous data set for training. The profile partition generator G2 is used to generate the earth-rock dam design scheme according to the heterogeneous data set for training and the earth-rock dam profile contour. The physical and mechanical parameters of the dam material adopted when constructing the heterogeneous data set are determined according to the material source information, and the design key points and the dam safety specification requirements are used as design conditions. The earth-rock dam design scheme includes an earth-rock dam profile design drawing, and the earth-rock dam profile partitions in the earth-rock dam profile design drawing include an earth-rock dam core wall, a face plate, a cushion layer, a transition layer, a filter layer, a main rockfill area, a secondary rockfill area, a drainage area, and a mold increasing area.

[0068] The specification constraint discriminator N is used to judge the dam safety of the earth-rock dam design scheme generated by the generator G, so as to realize the constraint of the design specification. The mechanical performance evaluator E is used to evaluate the working state of the earth-rock dam design scheme generated by the generator G, so as to realize the optimization of the mechanical performance. The working state includes stress indicators and deformation indicators, and the working state is used to reflect the dam permeation stability, deformation stability, anti-seismic safety, and dam slope stability performance.

[0069] In another example embodiment of the present application, the step 206 and the step 207 generate the to-be-designed earth-rock dam design scheme by using the earth-rock dam generative design model, and then further intelligently evaluate the to-be-designed earth-rock dam design scheme by using the dam body working state evaluation model. If the to-be-designed earth-rock dam design scheme does not meet the design point requirements of the to-be-designed earth-rock dam and the dam body safety specification requirements of the to-be-designed earth-rock dam, it indicates that the training data may be lack of the dam type of the to-be-designed earth-rock dam in the adversarial training process. Therefore, new training data needs to be added to fine-tune the earth-rock dam generative design model and the dam body working state evaluation model, so as to further improve the quality of the earth-rock dam design scheme.

[0070] In actual application, an implementation process of the above-mentioned earth-rock dam generative intelligent design method can be specifically described as follows.

[0071] (1) The prompt word engineering technology is used to learn the text knowledge in the earth-rock dam design field, and the design point requirements and the dam body safety specification requirements of different dam types are automatically generated.

[0072] (2) The earth-rock dam generative design model fused with multi-modal data is constructed.

[0073] Specifically, facing the design point requirements of the required design dam body, the AI model architecture of the earth-rock dam profile partition design generation is proposed by using the heterogeneous data fused with the design dam height, the geological exploration data, the foundation surface image and the dam material physical and mechanical parameters. The generator is trained on the design instance data set to obtain the earth-rock dam generative design model.

[0074] (3) The finite element proxy model of the earth-rock dam working state evaluation based on the artificial neural network is established.

[0075] Specifically, facing the specification requirements of the dam body safety, the finite element calculation data set reflecting the dam body seepage, deformation, anti-seismic and dam slope stability is constructed by the finite element calculation of the earth-rock dam under different profile forms and dam material physical and mechanical parameters. The dam body working state evaluation model based on the artificial neural network proxy model (i.e. the evaluator, also called the finite element calculation proxy model) is established and trained by using the data set.

[0076] In this step, during the training, the database of the earth-rock dam structure model and the stress and deformation mechanics response is constructed based on the finite element analysis; the deep neural network is constructed, the evaluator is trained by using the earth-rock dam finite element calculation database, so as to realize the evaluation of the earth-rock dam structure mechanical properties. The stress and deformation indexes under the static and dynamic conditions are predicted by using the evaluator for the AI intelligently generated earth-rock dam design profile; the feature engineering analysis is performed on the prediction results of the stress and deformation indexes, the key control indexes of the earth-rock dam mechanical properties are measured, and the AI design scheme evaluation model considering the earth-rock dam mechanical properties is constructed.

[0077] (4) Considering dam safety and stability, the dam profile generation model is trained.

[0078] The multi-dimensional index evaluation loss function of the design dam profile work state is constructed to evaluate the automatically generated dam design scheme intelligently, and the dam profile generation model is trained.

[0079] (5) Automatically generating a dam design scheme that meets the engineering safety requirements.

[0080] The dam profile generation model and the dam work state evaluation model are called to generate a dam design scheme that meets the engineering safety requirements by inputting the key requirements of the required design and the specification requirements of the dam safety.

[0081] The method of the above embodiment is a dam profile work state trained dam profile generation intelligent design method, which considers the work state of the profile, trains the dam profile generation model, and on the basis of the automatic design of the dam profile, intelligently judges the work state of the generated profile to ensure that the design generation result meets the engineering safety and stability requirements, and realizes efficient and high-quality design of the dam.

[0082] Based on the same inventive concept, the embodiments of the present application also provide a dam profile generation intelligent design device for implementing the dam profile generation intelligent design method described above. The implementation scheme for solving the problem provided by the device is similar to the implementation scheme described in the above method, so the specific limitations in one or more dam profile generation intelligent design device embodiments provided below can be referred to the limitations of the dam profile generation intelligent design method in the above text, which will not be repeated here.

[0083] In an exemplary embodiment, as shown in Figure 5 a dam profile generation intelligent design device is provided, which includes the following modules.

[0084] The design text knowledge acquisition module 501 is used to acquire the design text knowledge of different dam types of the dam.

[0085] The requirement generation module 502 is used to generate design key requirement and dam safety specification requirement of different dam types according to the design text knowledge.

[0086] The first heterogeneous data set construction module 503 is used to construct a heterogeneous data set for training based on the design key requirement of different dam types; the heterogeneous data set is determined according to text data and image data; the text data includes dam height, geological exploration data and dam material physical and mechanical parameters; the image data includes foundation surface image and dam body contour image.

[0087] The finite element calculation dataset construction module 504 is configured to input the heterogeneous dataset for training into a generator of the generative adversarial network, and construct a finite element calculation dataset according to a design scheme of the earth-rock dam output by the generator based on dam body safety specification requirements of different dam types; the finite element calculation dataset comprises the design scheme of the earth-rock dam and corresponding working behavior calculated by using a finite element analysis method.

[0088] The adversarial training module 505 is configured to input the finite element calculation dataset into an evaluator of the generative adversarial network, and perform adversarial training on the generative adversarial network with the dam body safety of the design scheme of the earth-rock dam output by the evaluator satisfying dam body safety specification requirements as a constraint and a loss value of the working behavior output by the evaluator and the working behavior calculated in the finite element calculation dataset being minimum as an objective, to obtain a trained generative adversarial network; wherein the trained generator is taken as an earth-rock dam generative design model, and the trained evaluator is taken as a dam body working behavior evaluation model.

[0089] The second heterogeneous dataset construction module 506 is configured to construct a heterogeneous dataset to be designed based on design point requirements required by the earth-rock dam to be designed.

[0090] The earth-rock dam design scheme generation module 507 is configured to input the heterogeneous dataset to be designed into the earth-rock dam generative design model, and generate a design scheme of the earth-rock dam to be designed.

[0091] In an exemplary embodiment, a computer device can be provided, which can be a server or a terminal, and an internal structure diagram thereof can be as shown in Figure 6 The computer device comprises a processor, a memory, an input / output interface (I / O) and a communication interface. The processor, the memory and the input / output interface are connected through a system bus, and the communication interface is connected to the system bus through the input / output interface. The processor of the computer device is configured to provide computing and control capabilities. The memory of the computer device comprises a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program and a database. The internal memory provides an environment for the operating system and the computer program in the non-volatile storage medium to run. The database of the computer device is configured to store design text knowledge of different dam types of earth-rock dams. The input / output interface of the computer device is configured to exchange information between the processor and external devices. The communication interface of the computer device is configured to communicate with external terminals through network connection. The computer program is executed by the processor to implement an earth-rock dam generative intelligent design method.

[0092] Those skilled in the art can understand that, Figure 6The structure shown in the figure is only a block diagram of part of the structure related to the scheme of the present application, and does not constitute a limitation on the computer device to which the scheme of the present application is applied. The specific computer device can include more or fewer components than those shown in the figure, or combine certain components, or have a different arrangement of components. In an exemplary embodiment, a computer device is provided, including a memory and a processor, the memory storing a computer program, and the processor executing the computer program to implement the steps in the above method embodiments.

[0093] In an exemplary embodiment, a computer readable storage medium is provided, storing a computer program, which is executed by a processor to implement the steps in the above method embodiments.

[0094] In an exemplary embodiment, a computer program product is provided, including a computer program, which is executed by a processor to implement the steps in the above method embodiments.

[0095] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data for analysis, stored data, displayed data, etc.) involved in the present application are all information and data authorized by the user or authorized by all parties, and the collection, use and processing of related data need to comply with relevant regulations.

[0096] Those skilled in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be completed by instructing the relevant hardware through a computer program. The computer program can be stored in a non-volatile computer readable storage medium, and when the computer program is executed, the processes of the above-mentioned embodiments of the methods can be included. Any reference to memory, database or other medium used in the embodiments provided in the present application can include at least one of non-volatile and volatile memory. The non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical storage, high-density embedded non-volatile memory, resistive memory (ReRAM), magnetoresistive random access memory (MRAM), ferroelectric memory (FRAM), phase change memory (PCM), graphene memory, etc. The volatile memory can include random access memory (RAM) or external cache memory, etc. As an illustration but not limitation, the RAM can be in various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM), etc.

[0097] The database involved in the embodiments provided in the present application can include at least one of a relational database and a non-relational database. The non-relational database can include a distributed database based on a block chain, etc., without being limited thereto. The processor involved in the embodiments provided in the present application can be a general processor, a central processing unit, a graphics processing unit, a digital signal processor, a programmable logic device, etc., without being limited thereto.

[0098] The technical features of the above embodiments can be combined arbitrarily. In order to make the description concise, all possible combinations of the technical features in the above embodiments are not described, however, as long as the combinations of the technical features do not exist contradictory, they should be considered as the scope of the present application.

[0099] The principles and implementation modes of the present application are described by using specific examples in the present application. The above-mentioned embodiments are only used to help understand the method and its core idea of the present application; at the same time, for those skilled in the art, according to the idea of the present application, the specific implementation mode and application range will be changed. In conclusion, the content of the present application should not be understood as a limitation.

Claims

1. A method for generating intelligent design of earth-rockfill dam, characterized in that, The earth-rock dam generation type intelligent design method comprises: acquiring design text knowledge of different dam types of earth-rock dams; generating design key point requirements and dam body safety specification requirements of different dam types according to the design text knowledge; constructing a heterogeneous data set for training based on the design key point requirements of different dam types; the heterogeneous data set is determined according to text data and image data; the text data comprises dam height, geological exploration data and dam material physical and mechanical parameters; the image data comprises foundation surface images and dam body contour images; inputting the heterogeneous data set for training into a generator of a generative adversarial network, constructing a finite element calculation data set according to an earth-rock dam design scheme output by the generator based on dam body safety specification requirements of different dam types; the finite element calculation data set comprises an earth-rock dam design scheme and corresponding working behavior calculated by using a finite element analysis method; inputting the finite element calculation data set into an evaluator of the generative adversarial network, taking dam body safety of an earth-rock dam design scheme output by the evaluator meeting dam body safety specification requirements as a constraint, and taking a loss value of working behavior output by the evaluator and working behavior calculated in the finite element calculation data set being minimum as an objective to perform adversarial training on the generative adversarial network, to obtain a trained generative adversarial network; wherein the trained generator is taken as an earth-rock dam generative design model, and the trained evaluator is taken as a dam body working behavior evaluation model; constructing a heterogeneous data set to be designed based on design key point requirements required by an earth-rock dam to be designed; inputting the heterogeneous data set to be designed into the earth-rock dam generative design model to generate an earth-rock dam design scheme to be designed.

2. The earth-rock dam generative intelligent design method according to claim 1, characterized in that, generating design key point requirements and dam body safety specification requirements of different dam types according to the design text knowledge, specifically comprising: completely electronically processing the design text knowledge, standardizing data formats and unifying data formats by using natural language processing technology, to obtain processed data; constructing an earth-rock dam design field knowledge graph according to the processed data; based on the earth-rock dam design field knowledge graph, learning field text knowledge of a large model by using a prompt word engineering technology, and generating design key point requirements and dam body safety specification requirements of different dam types according to the learned large model.

3. The earth-rock dam generative intelligent design method according to claim 1, characterized in that, constructing a heterogeneous data set for training based on design key point requirements of different dam types, specifically comprising: collecting engineering data of built earth-rock dams; the engineering data comprises dam height, geological exploration data, foundation surface images and dam material physical and mechanical parameters; constructing a foundation surface-dam slope-dam height earth-rock dam contour image matrix and a material parameter vector for training based on design key point requirements of different dam types and using engineering data of built earth-rock dams; determining the heterogeneous data set for training according to the foundation surface-dam slope-dam height earth-rock dam contour image matrix and the material parameter vector for training.

4. The earth-rock dam generative intelligent design method according to claim 1, characterized in that, constructing a finite element calculation data set according to an earth-rock dam design scheme output by the generator based on dam body safety specification requirements of different dam types, specifically comprising: Perform finite element analysis on the earth-rock dam under the earth-rock dam design scheme and the physical and mechanical parameters of dam materials output by the generator based on dam body safety specification requirements of different dam types, and calculate the working state of the corresponding earth-rock dam; An earth-rock dam design scheme output by the generator and a corresponding working state calculated thereby establish a finite element calculation dataset.

5. The earth-rock dam generative intelligent design method according to claim 1, characterized in that, The evaluator is an artificial neural network agent model.

6. The earth-rock dam generative intelligent design method according to claim 1, characterized in that, The generator comprises a profile contour generator and a profile partition generator. The profile contour generator is configured to generate an earth-rock dam profile contour based on a heterogeneous dataset for training, and the profile partition generator is configured to generate the earth-rock dam design scheme based on the heterogeneous dataset for training and the earth-rock dam profile contour; the earth-rock dam design scheme comprises an earth-rock dam profile design drawing, and an earth-rock dam profile partition in the earth-rock dam profile design drawing comprises an earth-rock dam core wall, a face plate, a cushion layer, a transition layer, a filter layer, a main rockfill area, a secondary rockfill area, a drainage area, and a mold area; The evaluator comprises a specification constraint discriminator and a mechanical performance evaluator; the specification constraint discriminator is configured to judge the dam body safety of the earth-rock dam design scheme generated by the generator, and the mechanical performance evaluator is configured to evaluate the working state of the earth-rock dam design scheme generated by the generator; The working state comprises stress indicators and deformation indicators, and the working state is used to reflect the dam body seepage stability, deformation stability, seismic safety, and dam slope stability performance.

7. The earth-rock dam generative intelligent design method according to claim 1, characterized in that, The design text knowledge comprises earth-rock dam design specifications, hydraulic structure design manuals, and typical engineering design reports.

8. A soil and rockfill dam generative intelligent design apparatus, characterized by, The earth-rock dam generative intelligent design apparatus comprises: A design text knowledge acquisition module configured to acquire design text knowledge of different dam types of earth-rock dams; A requirement generation module configured to generate design point requirements and dam body safety specification requirements of different dam types based on the design text knowledge; A first heterogeneous dataset construction module configured to construct a heterogeneous dataset for training based on design point requirements of different dam types; the heterogeneous dataset is determined based on text data and image data; the text data comprises dam height, geological exploration data, and physical and mechanical parameters of dam materials; and the image data comprises a foundation surface image and a dam body contour image; A finite element calculation dataset construction module configured to input the heterogeneous dataset for training into a generator of a generative adversarial network, construct a finite element calculation dataset based on dam body safety specification requirements of different dam types and an earth-rock dam design scheme output by the generator; and the finite element calculation dataset comprises an earth-rock dam design scheme and a corresponding working state calculated by a finite element analysis method. A finite element calculation dataset construction module configured to input the heterogeneous dataset for training into a generator of a generative adversarial network, construct a finite element calculation dataset based on dam body safety specification requirements of different dam types and an earth-rock dam design scheme output by the generator; and the finite element calculation dataset comprises an earth-rock dam design scheme and a corresponding working state calculated by a finite element analysis method. The adversarial training module is configured to input the finite element calculation dataset into an evaluator of the generative adversarial network, take the dam safety of the dam design scheme output by the evaluator as a constraint which meets the dam safety specification requirement, and perform adversarial training on the generative adversarial network with the loss value of the working state output by the evaluator and the working state calculated in the finite element calculation dataset as a target, to obtain a trained generative adversarial network; wherein the trained generator is used as a soil and rock dam generative design model, and the trained evaluator is used as a dam working state evaluation model. The second heterogeneous data set construction module is configured to construct a to-be-designed heterogeneous data set based on design point requirements required by the to-be-designed soil and rock dam. The soil and rock dam design scheme generation module is configured to input the to-be-designed heterogeneous data set into the soil and rock dam generative design model to generate a design scheme of the to-be-designed soil and rock dam.

9. A computer device comprising: The memory, the processor, and the computer program stored in the memory and executable on the processor are characterized in that the processor executes the computer program to implement the soil and rock dam generative intelligent design method in any one of claims 1-7.

10. A computer-readable storage medium having stored thereon a computer program, characterized in that, The computer program is executed by the processor to implement the soil and rock dam generative intelligent design method in any one of claims 1-7.

Images