Nuclear power industry mechanism model platform architecture for industrial internet platform integration

Abstract

This invention discloses a nuclear power industry mechanism model platform architecture for integration with an industrial internet platform, comprising a runtime environment layer, a service layer, an interface layer, a view layer, and an application layer. The runtime environment layer abstracts the hardware, provides the necessary library files for the upper-layer environment, and handles inter-cluster scheduling and communication. The service layer includes AI models, real-time computing models, cloud simulation models, industrial cloud software models, auxiliary analysis tools, and integrated development and execution tools. The interface layer standardizes and uniformly manages the interfaces of the service layer. The view layer includes mechanism model library interfaces, an operation management module, a data management module, and an engineering management module. The application layer includes artificial intelligence application scenarios, real-time computing scenarios, cloud simulation scenarios, and cloud-based industrial software scenarios. This invention provides an integrated development and execution environment for nuclear power industry mechanism models.

Description

Technical Field

[0001] This invention relates to the field of nuclear power industrial internet platform optimization technology, and in particular to a nuclear power industrial mechanism model platform architecture for industrial internet platform integration. Background Technology

[0002] With the development of new-generation information technology, the Industrial Internet has become an important enabling tool for the digital transformation of the real economy at home and abroad. Its core idea is to combine emerging information technology with traditional manufacturing, and take cyber-physical systems as the core technology system to realize comprehensive perception, dynamic transmission, real-time analysis, scientific decision-making and intelligent control of data in the enterprise's production and operation process, thereby promoting the upgrading and transformation of traditional industries.

[0003] Nuclear power operations involve a wide variety and large number of mechanism models. In order to meet the programming and debugging needs of various mechanism models and improve user experience, new requirements have been put forward for the optimization of the nuclear power industrial internet platform. Summary of the Invention

[0004] In view of this, it is necessary to provide a nuclear power industry mechanism model platform architecture for integration with industrial internet platforms, so as to effectively solve the technical problem of the lack of an integrated development environment for nuclear power industry mechanism models.

[0005] This invention provides a nuclear power industry mechanism model platform architecture for integration with industrial internet platforms, comprising an operating environment layer, a service layer, an interface layer, a view layer, and an application layer. The operating environment layer abstracts the hardware, provides the necessary library files for the upper-layer environment, and performs scheduling and communication between clusters. The service layer includes AI models, real-time computing models, cloud simulation models, industrial cloud software models, auxiliary analysis tools, and integrated development and operation tools. The interface layer standardizes and uniformly manages the interfaces of the service layer. The view layer includes mechanism model library interfaces, operation management modules, data management modules, and engineering management modules. The application layer includes artificial intelligence application scenarios, real-time computing scenarios, cloud simulation scenarios, and cloud-based industrial software scenarios.

[0006] Preferably, the runtime environment layer is equipped with a nuclear power industry mechanism model library that stores the library files. The AI ​​model, real-time computing model, cloud simulation model, and industrial cloud software model can use or upload / delete the model by calling the model service of the nuclear power industry mechanism model library.

[0007] Preferably, the service layer further includes a data management module and a model resource interface management module; the data management module obtains data resources from the nuclear power industrial internet platform, the nuclear power mechanism model library preprocesses the data resources, and uses the preprocessed data resources to train and verify the mechanism models, and analyzes and optimizes the original model resource files of the nuclear power industrial mechanism model library based on the training and verification results; the resource interface management module is used to perform read and write operations on the nuclear power industrial mechanism model library, provide a model publishing and sharing interface, uniformly publish various model resources within the platform to the service open platform, and provide a model status interface to monitor the mechanism models.

[0008] Preferably, the service layer uses Docker and Singularity to manage containerized environments for different types of models, and uses Kubernetes for container orchestration, providing three elastic scaling strategies: CA, HPA, and VPA.

[0009] Preferably, the interface layer provides a unified external service interface based on the Resful API and imposes constraints on the model library interface, sample database interface, and service interface.

[0010] Preferably, the view layer uses YS3D, VUE, React+Mobx, and Ant Design for visualization and 3D rendering.

[0011] Preferably, it also includes a hardware layer, which is equipped with an AI server cluster, a real-time computing server cluster, a cloud simulation server cluster, an industrial software server cluster, an integrated development server, and a data storage cluster.

[0012] The server cluster includes GPU servers, HPC servers, and other servers. The integrated development server includes front-end servers and back-end servers. The data storage cluster uses relational databases, object databases, and in-memory databases to store model data, configure the environment, and interact with data in the nuclear power industry mechanism model library.

[0013] Preferably, the interface layer specifically includes an authentication module interface, a platform model management interface, a model service management interface, and a unified gateway interface;

[0014] The authentication module interface is a GET interface, which is used to obtain the token of the component service. The component service first checks whether the transmitted platform username exists. If it does not exist, it recreates the user and then logs in the user and returns the token.

[0015] The platform model management interface includes interfaces for adding a model, modifying a model, deleting a model, uploading a model, debugging a model, querying model debugging results, uploading an IML model, and downloading an IML model. The add model interface is a POST interface for adding a model; the modify model interface is a GET interface for modifying a model; the delete model interface is a DELETE interface for deleting a model; the upload model interface is a POST interface for uploading a model when the component service does not provide a configuration page; the debug model interface is a POST interface for submitting a model for debugging when the component service does not provide a configuration page; the query model debugging results interface is a GET interface for querying model debugging results; the upload model IML interface is a POST interface for uploading an IML model; and the download IML model interface is a GET interface for downloading an IML model and creating the model.

[0016] The model service management interface includes a model download and run interface, a model service creation API interface, a model run information query interface, a model run interaction interface, and a run environment query interface. The model download and run interface is a POST interface used for downloading and running models; the model service creation API interface is a GET interface used for creating model services; the model run information query interface is a GET interface used for querying model run information; the model run interaction interface is a GET interface used for model run interaction commands; and the run environment query interface is a POST interface used for querying the run environment.

[0017] The unified gateway interface is a POST interface, used for authentication and authorization.

[0018] Preferably, the service layer further includes a basic security module, which includes an access control module, a transmission security module, a storage security module, an asset access security module, and a network protection module.

[0019] The permission management module assigns different functional permissions to different types of users, allowing them to access the corresponding permission pages through different web interfaces; the transmission security module provides data encryption, data decryption, and data verification services; the storage security module provides database auditing, data encryption, and data backup and recovery services; the asset open security module provides asset access control and access log auditing services; and the network protection module provides firewall, intrusion detection, and intrusion prevention services.

[0020] Preferably, the service layer further includes a business function security module, which includes a model call management module and a model transaction management module. The model call management module performs access permission authentication, usage control, and monitoring on users through model call business operations to ensure the security of model usage. The model transaction management module performs value assessment, transaction verification, encrypted model information transaction, transaction information recording and auditing, transaction result credibility assessment, and transaction process log recording on the transaction model to ensure the security of model transactions.

[0021] Compared with the prior art, the beneficial effects of the present invention are as follows: the runtime environment layer provides library files for the upper environment, the service layer integrates four main components, the interface layer performs unified interface management, the view layer realizes operation management, and the application layer integrates different scenarios. Overall, it provides an integrated development and operation environment for the nuclear power industry mechanism model, and provides an online orchestration and debugging runtime environment for the four components of the service layer, as well as service encapsulation and release functions. Attached Figure Description

[0022] The accompanying drawings, which are included to provide a further understanding of the invention and form part of this application, illustrate exemplary embodiments of the invention and, together with their description, serve to explain the invention and do not constitute an undue limitation thereof. In the drawings:

[0023] Figure 1 This is a functional architecture diagram of an embodiment of the nuclear power industry mechanism model platform architecture for industrial internet platform integration provided by the present invention.

[0024] Figure 2 yes Figure 1 The component call diagram of an embodiment of the nuclear power industry mechanism model platform architecture for industrial internet platform integration shown in the example is as follows:

[0025] Figure 3 yes Figure 1 The data architecture diagram shown is from one embodiment of the nuclear power industry mechanism model platform architecture for industrial internet platform integration.

[0026] Figure 4 yes Figure 1 The illustrated embodiment is a technical architecture diagram of an embodiment of a nuclear power industry mechanism model platform architecture for industrial internet platform integration.

[0027] Figure 5 yes Figure 1 The illustrated embodiment is a hardware deployment architecture diagram of a nuclear power industry mechanism model platform architecture for industrial internet platform integration.

[0028] Figure 6 yes Figure 1The diagram shown is a security architecture diagram of an embodiment of a nuclear power industry mechanism model platform architecture for industrial internet platform integration. Detailed Implementation

[0029] Preferred embodiments of the present invention will now be described in detail with reference to the accompanying drawings, which form part of this application and are used together with the embodiments of the present invention to illustrate the principles of the present invention, but are not intended to limit the scope of the present invention.

[0030] Example 1

[0031] The nuclear power industry mechanism model platform architecture for industrial internet platform integration in this embodiment includes an operating environment layer, a service layer, an interface layer, a view layer, and an application layer. The operating environment layer is used to abstract the hardware, provide the library files required for operation of the upper-layer environment, and perform scheduling and communication between clusters. The service layer includes AI models, real-time computing models, cloud simulation models, industrial cloud software models, auxiliary analysis tools, and integrated development and operation tools. The interface layer standardizes and uniformly manages the interfaces of the service layer. The view layer includes mechanism model library interfaces, operation management modules, data management modules, and engineering management modules. The application layer includes artificial intelligence application scenarios, real-time computing scenarios, cloud simulation scenarios, and cloud-based industrial software scenarios.

[0032] Please see Figure 1The nuclear power industry mechanism model integrated development and operation platform provided in this embodiment consists of multiple layers, mainly including five layers: the operating environment layer, the service layer, the interface layer, the view layer, and the application layer. Each layer is composed of sub-modules and components. The operating environment layer mainly abstracts the hardware, provides the library files required for the upper-layer environment to run, and performs scheduling and communication between clusters. The service layer mainly includes the service functions of four sub-modules and the functions of two tools. The four sub-modules respectively solve the integrated development and operation problems of AI models, real-time computing models, cloud simulation models, and industrial cloud software models. The two tools are an auxiliary analysis tool and a model integration development and operation tool. The auxiliary analysis tool includes a sensitivity analysis tool, an uncertainty analysis tool, and an intelligent optimization tool. The model integration development tool includes a model acquisition tool, a development and debugging tool, a model packaging tool, and a model release tool. The runtime environment layer's library files are configured for the four sub-modules of the service layer, providing AI computing environment, real-time computing environment, co-simulation environment, and industrial CAE simulation computing environment for each sub-module. Each computing environment contains different library files for the four sub-modules to call. Simultaneously, the runtime environment implements the scheduling of GPU clusters, computing clusters, and HPC clusters. The interface layer primarily provides standardized encapsulation and unified permission management for the service layer's interfaces, offering a unified interface service to the outside world. The view layer mainly includes the interface for the mechanism model library, operation management, data management, and engineering management. The application layer mainly consists of applications developed based on the functions of other layers, primarily including artificial intelligence application scenarios, real-time computing scenarios, cloud simulation scenarios, and cloud-based industrial software scenarios, with various applications available for users to choose from in each scenario.

[0033] The nuclear power industry mechanism model platform architecture provided in this embodiment for industrial internet platform integration aims to provide an integrated development and operation environment for nuclear power industry mechanism models, and to provide online orchestration and debugging operation environment for AI models, real-time computing models, simulation models and industrial cloud software, as well as service encapsulation and release services.

[0034] Specifically, the runtime environment layer is equipped with a nuclear power industry mechanism model library that stores the library files. The AI ​​model, real-time computing model, cloud simulation model, and industrial cloud software model can use or upload / delete the model by calling the model service of the nuclear power industry mechanism model library.

[0035] like Figure 2As shown, this platform mainly comprises four components. Users access the platform through the front-end service and, after unified authentication and login, can use the functions of each component. The four components are the AI ​​service component, the real-time computing service component, the cloud simulation service component, and the industrial software service component. During use, if it is necessary to use existing models on the platform or to add or remove models, the model library service of the nuclear power industry mechanism model library is invoked.

[0036] Specifically, the service layer also includes a data management module and a model resource interface management module. The data management module obtains data resources from the nuclear power industrial internet platform. The nuclear power mechanism model library preprocesses the data resources and uses the preprocessed data resources to train and verify the mechanism models. Based on the training and verification results, it analyzes and optimizes the original model resource files of the nuclear power industrial mechanism model library. The resource interface management module is used to read and write operations on the nuclear power industrial mechanism model library, provide a model publishing and sharing interface, uniformly publish various model resources within the platform to the service open platform, and provide a model status interface to monitor the mechanism models.

[0037] To achieve unified, streamlined, and standardized industrial model development and management, the model integration development platform decouples the entire lifecycle of industrial model management, connecting each element to form a relatively standardized process. This process covers the entire lifecycle of model resource management, model creation, packaging, debugging, verification, operation, production services, and monitoring management, enabling different stakeholders in the model lifecycle to clarify their responsibilities, collaborate better, and improve overall operational efficiency. The platform's data architecture is as follows: Figure 3 As shown, external data resources come from the Nuclear Power Industrial Internet Platform, and external model resources come from the Nuclear Power Industrial Mechanism Model Library. The Nuclear Power Mechanism Model Library provides industrial software, models, and related configuration information, specifically including data sample tools, an asset management system, and data annotation tools. The platform obtains data resources from the Nuclear Power Industrial Internet Platform through the data management module, and generates sample data, multimodal data, and a feature repository through the data sample tools, asset management system, and data annotation tools in the Nuclear Power Industrial Mechanism Model Library. It then uses the generated sample data, multimodal data, and feature repository to train and validate the mechanism models, and reads model resource files from the Nuclear Power Industrial Mechanism Model Library for analysis and optimization. The model resource interface management module provides read and write operations to the Nuclear Power Industrial Mechanism Model Library for use by other modules within the platform. This module also provides a model publishing and sharing interface, uniformly publishing various model resources within the platform to the service open platform according to the format requirements and security control requirements of intelligent applications. Furthermore, the model resource interface management module provides a model status interface to monitor the development, operation, and usage of the models.

[0038] Specifically, the service layer uses Docker and Singularity to manage containerized environments for different types of models, and uses Kubernetes for container orchestration, providing three elastic scaling strategies: CA, HPA, and VPA.

[0039] Specifically, the interface layer provides a unified external service interface based on the ResfulAPI and imposes constraints on the model library interface, sample database interface, and service interface.

[0040] Specifically, the view layer uses YS3D, VUE, React+Mobx, and AntDesign for visualization and 3D rendering.

[0041] Specifically, the technical architecture of the model integration development and runtime platform Figure 4 As shown, the platform's technical architecture is divided into four layers: Model Library: It uses relational database MySQL, object database OSS, and in-memory database Redis to address various needs related to model data, configuration environment, and data interaction; Platform Layer: It employs containerized environment management for different model types. Considering the needs of different model applications and services, the platform uses Docker and Singularity for containerized environment management and Kubernetes for container orchestration, providing three elastic scaling strategies: CA, HPA, and VPA; Service Interface: The interface layer provides a unified external service interface based on ResfulAPI; Front-end Display: The front-end uses YS3D, Vue, React+Mobx, and AntDesign for visualization and 3D rendering.

[0042] The platform's general technical foundation comprises three parts: hardware configuration: considering the application requirements of different models, the hardware layer requires the construction of three types of hardware devices: GPU servers, computing clusters, and HPC servers; to adapt to the needs of different models and applications, the platform uses two containerization technologies, Singularity and Docker, to handle different models; the interface design mainly constrains the model library interface, sample database interface, and service interface.

[0043] Specifically, it also includes a hardware layer, which is equipped with AI server clusters, real-time computing server clusters, cloud simulation server clusters, industrial software server clusters, integrated development servers, and data storage clusters.

[0044] The server cluster includes GPU servers, HPC servers, and other servers. The integrated development server includes front-end servers and back-end servers. The data storage cluster uses relational databases, object databases, and in-memory databases to store model data, configure the environment, and interact with data in the nuclear power industry mechanism model library.

[0045] The deployment architecture of the model integration development and runtime platform is as follows: Figure 5 As shown, the hardware layer is equipped with a database server, and the computing server includes GPU computing nodes, computing clusters and HPC clusters. In addition, task scheduling, cloud rendering and external service interfaces all use separate servers.

[0046] The core functions of a GPU server include monitoring GPU resources, obtaining POD requirements, matching PODs with GPU resources, reserving GPU resources, monitoring GPU resource usage, integrating existing schedulers, and scheduling policies.

[0047] A computing cluster is a system composed of multiple computers interconnected by a high-speed network to collaborate on computational tasks. Each computer in the cluster is called a node, and nodes can share resources, status, and data with each other. The goal of a computing cluster is to achieve efficient, high-performance computing while also possessing high availability and load balancing capabilities.

[0048] High-performance computing, or HPC for short, consists of multiple networked high-performance computer servers. It interconnects a large number of servers and storage devices through a high-performance network to build a large-scale computing cluster, and has a centralized job scheduler to manage parallel computing workloads. In other words, it breaks down large-scale computing tasks and distributes them to various servers for parallel computing, and then aggregates the computing results to obtain the final result.

[0049] Docker is an open-source application container engine that wraps applications and their dependencies into portable containers, allowing applications to run on any Docker-enabled platform. Docker containers offer high isolation and, being based on the Linux kernel, are faster than traditional virtual machines. They feature standardized units, lightweight design, high isolation, efficient networking, ease of continuous integration and continuous delivery, and a rich ecosystem. Kubernetes can also be used to manage and orchestrate service containers for elastic scaling.

[0050] Considering the actual situation of this project, PowerJob is adopted as the task scheduler for this platform.

[0051] Service management aims to run user models as a web server. Users can reuse pre-trained AI models and perform prediction tasks using REST APIs. For versatility, service management can use user-defined Docker images and execute user-specified "run commands" to run the required "model service" as a web server; alternatively, it can use existing frameworks: users only need to provide the trained model, and the system automatically matches a suitable prediction service image; or the system can assign an "access address" to successfully running prediction services, allowing users to access the model service through this address. Furthermore, service management also supports advanced capabilities required for AI service operation and maintenance, such as QPS super-resolution, prediction data collection, storage mounting, static ports, and GPU sharing.

[0052] Specifically, the interface layer includes an authentication module interface, a platform model management interface, a model service management interface, and a unified gateway interface;

[0053] The authentication module interface is a GET interface, which is used to obtain the token of the component service. The component service first checks whether the transmitted platform username exists. If it does not exist, it recreates the user and then logs in the user and returns the token.

[0054] The platform model management interface includes interfaces for adding a model, modifying a model, deleting a model, uploading a model, debugging a model, querying model debugging results, uploading an IML model, and downloading an IML model. The add model interface is a POST interface for adding a model; the modify model interface is a GET interface for modifying a model; the delete model interface is a DELETE interface for deleting a model; the upload model interface is a POST interface for uploading a model when the component service does not provide a configuration page; the debug model interface is a POST interface for submitting a model for debugging when the component service does not provide a configuration page; the query model debugging results interface is a GET interface for querying model debugging results; the upload model IML interface is a POST interface for uploading an IML model; and the download IML model interface is a GET interface for downloading an IML model and creating the model.

[0055] The model service management interface includes a model download and run interface, a model service creation API interface, a model run information query interface, a model run interaction interface, and a run environment query interface. The model download and run interface is a POST interface used for downloading and running models; the model service creation API interface is a GET interface used for creating model services; the model run information query interface is a GET interface used for querying model run information; the model run interaction interface is a GET interface used for model run interaction commands; and the run environment query interface is a POST interface used for querying the run environment.

[0056] The unified gateway interface is a POST interface, used for authentication and authorization.

[0057] Specifically, the various interfaces of the interface layer are shown in Table 1:

[0058] Table 1. Interface Layer Interface Design Table

[0059]

[0060] Specifically, IML refers to the Nuclear Power Industry Mechanism Model Library. This platform features a unified API gateway function. An API gateway is designed in distributed systems to protect internal services, providing high-performance, highly available API hosting services. This helps application service developers easily provide services externally without worrying about security controls, traffic control, log auditing, etc. Security authentication, traffic control, log auditing, blacklists / whitelists, and request combination are all implemented uniformly at the API gateway layer. The gateway provides key functions such as API publishing, management, and maintenance, making it simple, convenient, and efficient to build a new application service, allowing developers to focus more on business-related tasks.

[0061] The security architecture diagram of the model integration development and runtime platform is as follows: Figure 6 Based on the source of the demand, it is divided into two parts: platform basic security and business function security.

[0062] Specifically, the service layer also includes a basic security module, which includes an access control module, a transmission security module, a storage security module, an asset access security module, and a network protection module.

[0063] The permission management module assigns different functional permissions to different types of users, allowing them to access the corresponding permission pages through different web interfaces; the transmission security module provides data encryption, data decryption, and data verification services; the storage security module provides database auditing, data encryption, and data backup and recovery services; the asset open security module provides asset access control and access log auditing services; and the network protection module provides firewall, intrusion detection, and intrusion prevention services.

[0064] The platform's basic security measures serve to protect the overall platform, managing user roles and permissions; ensuring the security of data transmission; ensuring the security of data and model asset storage; ensuring the security of data and model assets exposed to the public; and providing network protection for the overall platform.

[0065] 1) Access Control

[0066] The platform is divided into system administrators, audit administrators, and ordinary users, with different functional permissions assigned to each. Users authenticate their identity using their account and password through different web interfaces and then access the corresponding permission pages to perform relevant functions.

[0067] 2) Transmission security

[0068] It provides data encryption services, converting raw data into unreadable ciphertext and transmitting it to the target server using the HTTPS secure communication protocol. It also provides data decryption and verification services; upon receiving encrypted data, the server uses a key to decrypt the ciphertext and verifies the data's integrity and whether it has been tampered with. This ensures the security and confidentiality of data transmission.

[0069] 3) Storage security

[0070] We provide database auditing services, recording and monitoring all database operations to ensure data storage security. We also provide data encryption services, encrypting data assets and model assets stored on the platform before storage to ensure data confidentiality. Furthermore, we offer data backup and recovery services, regularly performing comprehensive backups of all platform data and verifying the integrity and availability of the backup data. This facilitates rapid data recovery in the event of platform data anomalies, ensuring the platform's normal operation.

[0071] 4) Asset security

[0072] Provides asset access control services, allowing all asset owners to control access permissions. Other users can view and use the assets after authentication and access requests. Provides access log auditing services to record and monitor asset access information, and to restrict user access based on detected abnormal behavior.

[0073] 5) Network protection

[0074] Provides firewall services to monitor network traffic entering and leaving the platform, filtering data packets that match platform authorization to prevent unauthorized access and malicious traffic. Provides intrusion detection services to monitor network traffic in real time, performing pattern matching and statistical analysis to identify platform access behavior, thereby detecting abnormal behavior and potential malicious activity and issuing alerts. Provides intrusion prevention services to block and mitigate attacks when abnormal behavior and malicious traffic are detected, taking timely protective measures to prevent potential threats.

[0075] Specifically, the service layer also includes a business function security module, which includes a model call management module and a model transaction management module. The model call management module performs access permission authentication, usage control, and monitoring on users through model call business operations to ensure the security of model usage. The model transaction management module performs value assessment, transaction verification, encrypted model information transaction, transaction information recording and auditing, transaction result credibility assessment, and transaction process log recording on the transaction model to ensure the security of model transactions.

[0076] Business function security is applied to the security protection of platform-related business operations, including the two main business functions of model usage and model trading.

[0077] Model Usage: Users authenticate their identity through the user authentication module. After successful authentication, they can invoke business operations on the model. The model invocation management module performs access permission authentication, verifying user access rights. Only after successful authorization can users access and use authorized model services. During the access and use of model services, the model runtime management module controls and monitors model usage to ensure model security during operation. Key functions include:

[0078] 1) Identity Authentication: Users provide their username, password, and identity certificate to the platform for user identity authentication. Only after successful authentication can subsequent business operations be performed.

[0079] 2) Access Control: In conjunction with the system's user permission management, verify the functional permissions granted to users based on their assigned roles before granting them access to functional operations. Combine user model transaction information records and model access information records to verify users' asset access permissions before granting them access to model and data assets.

[0080] 3) Input Validation and Monitoring: Monitor and validate user input data during model usage. For models with validation rules, validate data format, data range, and other rules. Reject requests that do not meet the rules and display error messages. Prevent malicious input from causing model malfunctions.

[0081] 4) Vulnerability Scanning: Continuously scan for and fix security vulnerabilities during model usage to ensure the security of the model during operation.

[0082] 5) Security audit and log monitoring: Record and monitor the model's operation logs and call logs, regularly review the model itself and analyze usage logs, promptly identify vulnerabilities and abnormal operation behaviors in the model itself, and respond and handle them.

[0083] Model Trading: Users authenticate their identity through the user authentication module. After successful authentication, they can proceed with model trading operations. The model trading management module performs value assessment of the traded models, transaction verification, encrypted model information trading, transaction information recording, transaction information auditing, credibility assessment of transaction results, and transaction process log recording. Key functions include:

[0084] 1) Identity Authentication: Users provide their username, password, and identity certificate to the platform for user identity authentication. Only after successful authentication can subsequent business operations be performed.

[0085] 2) Transaction Verification: When a model is listed, the model owner provides an identity certificate to verify and authenticate the user's identity, ensuring the user's legitimacy and the model's ownership. Upon listing, the model owner selects the scope and content of the model transaction. The model buyer confirms the scope and content of the transaction to ensure the transaction parties are correct.

[0086] 3) Encrypted Information Transactions: When a model purchaser initiates a transaction and completes payment, the platform provides the encrypted content of the transaction and the decryption key for the model purchaser to decrypt the data and verify the validity and availability of the model.

[0087] 4) Transaction Records: Record user identity verification information, transaction communication information, and basic transaction information during the transaction process. The records are open to all users to ensure that transactions are open, transparent, authentic, and trustworthy.

[0088] 5) Transaction auditing: Real-time monitoring and analysis of all recorded transaction information, alerting and terminating abnormal behavior to ensure transaction security.

[0089] 6) Transaction evaluation: Based on all recorded data during the transaction process, evaluate the validity, completion rate, and completion results of the transaction to confirm that the transaction was completed normally.

[0090] 7) Log recording: Record information throughout the entire transaction process to ensure transaction traceability.

[0091] The above description is only a preferred embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any changes or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in the present invention should be included within the scope of the present invention.

Claims

1. A nuclear power industry mechanism model platform architecture oriented to industrial internet platform integration, characterized in that, The application comprises a running environment layer, a service layer, an interface layer, a view layer and an application layer; the running environment layer is used for abstracting hardware, providing a library file required for running for an upper layer environment and performing scheduling and communication between clusters, the service layer comprises an AI model, a real-time calculation model, a cloud-based simulation model, an industrial cloud-based software model, an auxiliary analysis tool and an integrated development and running tool; the interface layer performs standard encapsulation and unified management on an interface of the service layer, the view layer comprises a mechanism model library interface, an operation management module, a data management module and an engineering management module, and the application layer comprises an artificial intelligence application scenario, a real-time calculation scenario, a cloud-based simulation scenario and a cloud-based industrial software scenario; The running environment layer is provided with a nuclear power industry mechanism model library for saving the library file, and the AI model, the real-time calculation model, the cloud-based simulation model and the industrial cloud-based software model use or perform on-shelf and off-shelf operations on the model by calling a model service of the nuclear power industry mechanism model library; The service layer further comprises a data management module and a model resource interface management module; the data management module obtains data resources from a nuclear power industry internet platform, the nuclear power industry mechanism model library performs data preprocessing on the data resources, and trains and verifies a mechanism model by using the preprocessed data resources, analyzes and optimizes an original model resource file of the nuclear power industry mechanism model library based on a result of the training and verification, and the resource interface management module is used for reading and writing operations on the nuclear power industry mechanism model library, providing a model publishing and sharing interface, uniformly publishing various model resources in a platform to a service opening platform, and providing a model state interface to monitor the mechanism model; The service layer uses Docker and Singularity to perform containerized environment management on different types of models, and uses Kubernetes to perform container arrangement, and provides three elastic scaling strategies of CA, HPA and VPA; The service layer further comprises a basic security module, and the basic security module comprises a permission management module, a transmission security module, a storage security module, an asset opening security module and a network protection module; The permission management module divides different function permissions for different types of users, and enters a corresponding permission page through different WEB interfaces; the transmission security module provides data encryption services, data decryption services and data verification services; The storage security module provides database audit services, data encryption services and data backup and recovery services; The asset opening security provides asset access control services and access log audit services; and the network protection module provides firewall services, intrusion detection services and intrusion prevention services. The service layer further comprises a business function security module, the business function security module comprising a model calling management module and a model transaction management module; the model calling management module performs access permission authentication, use control and monitoring on users through a model calling business operation, ensuring the security in the process of model use, and the model transaction management module performs value assessment, transaction verification, model encryption information transaction, transaction information record and audit, transaction result credibility assessment and transaction process log record on the transaction model, ensuring the security of model transaction.

2. The nuclear power industry mechanism model platform architecture facing industrial internet platform integration according to claim 1, characterized in that, The interface layer provides a unified external service interface based on Resful API, and constrains the model library interface, sample database interface and service interface. 3.The industrial internet platform integration oriented nuclear power industry mechanism model platform architecture of claim 1, wherein, The view layer uses YS3D, VUE, React+Mobx and Ant Design for visual display and three-dimensional rendering.

4. The nuclear power industry mechanism model platform architecture facing industrial internet platform integration according to claim 1, characterized in that, Further comprising a hardware layer, the hardware layer being deployed with an AI server cluster, a real-time calculation server cluster, a clouded simulation server cluster, an industrial software server cluster, an integrated development server and a data storage cluster; The server cluster comprises GPU servers, HPC servers and other servers, the integrated development server comprises a front-end server and a back-end server, and the data storage cluster uses a relational database, an object database and an in-memory database to store model data of the nuclear power industry mechanism model library, configure an environment and interact data.

5. The nuclear power industry mechanism model platform architecture facing industrial internet platform integration according to claim 1, characterized in that, The interface layer specifically comprises an authentication module interface, a platform model management interface, a model service management interface and a unified gateway interface; The authentication module interface is a GET interface, used for obtaining a token of a component service, the component service first checks whether a platform username passed is present, if not, the user is rebuilt, and then the token is returned after the user is logged in; The platform model management interface comprises a new model interface, a model modification interface, a model deletion interface, a model uploading interface, a model debugging interface, a model debugging result query interface, an IML model uploading interface and an IML model downloading interface; the new model interface is a POST interface, used for adding a model, the model modification interface is a GET interface, used for modifying a model, the model deletion interface is a DELETE interface, used for deleting a model, the model uploading interface is a POST interface, used for uploading a model when a component service does not provide a configuration page, the model debugging interface is a POST interface, used for submitting a model for debugging when a component service does not provide a configuration page, the model debugging result query interface is a GET interface, used for querying a model debugging result, the IML model uploading interface is a POST interface, used for IML model uploading, and the IML model downloading interface is a GET interface, used for downloading an IML model and creating a model. The model service management interface comprises a model uploading and running interface, a model service API interface creation interface, a model running information query interface, a model running interaction interface and a running environment query interface; the model uploading and running interface is a POST interface, used for model uploading and running; the model service API interface creation interface is a GET interface, used for creating a model service API interface; the model running information query interface is a GET interface, used for querying model running information; the model running interaction interface is a GET interface, used for model running interaction instructions; and the running environment query interface is a POST interface, used for running environment query. The unified gateway interface is a POST interface, used for authentication and authorization.

Images