Product life cycle management system interaction method and device based on large model context protocol, medium and equipment

Abstract

The invention discloses a product life cycle management system interaction method and device based on a large model context protocol, a medium and equipment, and belongs to the technical field of man-machine interaction. The method comprises the following steps: initializing configuration of a system; obtaining an original instruction of each tenant by using the management system; performing semantic analysis on the original instruction of each tenant to generate a standardized tool calling sequence; the standardized tool call sequence is converted into an execution preview described by a natural language based on a large model context protocol for corresponding tenants to confirm; according to the confirmation result of the corresponding tenant, output information is obtained, and the output information meets the security level requirement of the corresponding tenant. The device, the medium and the equipment can be used for realizing the product life cycle management system interaction method based on the large model context protocol. The system can effectively solve the problems of low intelligent level, data isolated island and large model application safety of a traditional product life cycle management system.

Description

Technical Field

[0001] This invention relates to the field of human-computer interaction technology, and in particular to an interaction method, apparatus, medium and equipment for a product lifecycle management system based on a large model context protocol. Background Technology

[0002] As a core platform for digital transformation in the manufacturing industry, the Product Lifecycle Management System (PLifecycle Management System) spans the entire product lifecycle, from concept design to end-of-life recycling, playing a crucial role in improving R&D efficiency and data consistency. With sustainable development standards increasingly influencing corporate strategies, PLifecycle Management Systems have become a key driver of sustainable product development. However, as the manufacturing industry accelerates its transformation towards intelligent, personalized, and service-oriented approaches, the shortcomings of traditional PLifecycle Management Systems in terms of architecture and functionality are becoming increasingly apparent. Summary of the Invention

[0003] In view of this, the present invention provides an interaction method, apparatus, medium and device for a product lifecycle management system based on a large model context protocol, which can effectively solve the problems of low system intelligence, data silos and security of large model applications in traditional product lifecycle management systems, thus making it more suitable for practical use.

[0004] To achieve the first objective mentioned above, the technical solution of the product lifecycle management system interaction method based on the large model context protocol provided by this invention is as follows: The product lifecycle management system interaction method based on the large model context protocol provided by this invention includes the following steps: Load multi-tenant configuration information; based on the multi-tenant configuration information, establish exclusive environment parameters corresponding to each tenant, and enable the secure sandbox environment of the application container engine to isolate the tool execution process from the host system; initialize the audit system, establish a full-link operation log recording mechanism, and complete the initial configuration of the product lifecycle management system based on the large model context protocol. The product lifecycle management system based on the large model context protocol establishes an independent configuration context for each tenant in the multi-tenant system, and data access and business operations establish independent configuration contexts through the tenant identifiers of each tenant in the multi-tenant system. The original instructions of each tenant are obtained using the product lifecycle management system based on the large model context protocol, which has already completed initial configuration. The original instructions of each tenant are semantically parsed using a large language model to generate standardized tool call sequences; The standardized tool call sequence is based on the large model context protocol and is converted into an execution preview described in natural language for confirmation by the corresponding tenant. Based on the confirmation result of the corresponding tenant, the tool call is executed sequentially in the security sandbox environment of the application container engine, and the returned result is processed for permission adaptation by the data desensitization engine to obtain output information, wherein the output information meets the confidentiality requirements of the corresponding tenant.

[0005] The product lifecycle management system interaction method based on the large model context protocol provided by this invention can also be further implemented using the following technical measures.

[0006] Preferably, the process involves loading multi-tenant configuration information; establishing unique environment parameters for each tenant based on the multi-tenant configuration information, and enabling the application container engine's secure sandbox environment to isolate the tool execution process from the host system; initializing the audit system and establishing a full-link operation log recording mechanism. During the initial configuration steps of the product lifecycle management system based on the large model context protocol, the isolation strategy between multi-tenants includes the following isolation modes: Physical isolation mode allocates independent database cluster instances to different tenants in a multi-tenant system, thus isolating each tenant from the others. Logical isolation mode, in a shared database environment, uses the tenant's identity field to achieve logical separation of data, thus isolating each tenant in a multi-tenant system from each other; The hybrid isolation mode uses physical isolation for core business data and logical isolation for general business data.

[0007] Preferably, multi-tenant configuration information is loaded; based on the multi-tenant configuration information, exclusive environment parameters are established for each tenant, and the secure sandbox environment of the application container engine is enabled to isolate the tool execution process from the host system; the audit system is initialized, and a full-link operation log recording mechanism is established. In the process of completing the initial configuration steps of the product lifecycle management system based on the large model context protocol, a hierarchical structure of files is established for the file storage of the same tenant, and the storage path adopts a hierarchical structure of "tenant identifier / project identifier / file identifier".

[0008] Preferably, the process involves loading multi-tenant configuration information; establishing unique environment parameters for each tenant based on the multi-tenant configuration information, and enabling the application container engine's secure sandbox environment to isolate the tool execution process from the host system; initializing the audit system and establishing a full-link operation log recording mechanism; and completing the initial configuration steps of the product lifecycle management system based on the large model context protocol. Each tenant in a multi-tenant system is assigned a unique encryption key; It also includes a cross-isolation mode data access middleware, which provides a unified data access interface for upper-layer applications and shields the complexity of the underlying isolation implementation.

[0009] As a preferred option The large language model and service components of the product lifecycle management system based on the large model context protocol are all deployed on the organization's intranet. The large model context protocol server runs within a strictly network-isolated application container engine security sandbox. The product lifecycle management system based on the large model context protocol only allows communication between systems on the whitelist.

[0010] Preferably, for time-consuming tasks, the product lifecycle management system interaction method based on the large model context protocol includes the following steps: The original instructions of each tenant are obtained using the product lifecycle management system based on the large model context protocol, which has already completed initial configuration. Based on the original instructions from each tenant, immediately return an acceptance response to the corresponding tenant; Once the time-consuming task is completed, a task completion notification and task result will be sent to the corresponding tenant.

[0011] Preferably, the product lifecycle management system based on the large model context protocol includes a two-tier management view architecture, comprising a platform-level management view and a user-level management view, wherein... The platform-level management view provides system administrators with cross-tenant global monitoring and management capabilities, enabling real-time monitoring of the system health status of all tenants, including service availability, API response time, and error rate. The platform-level management view also provides a unified configuration function for resource quotas, supporting the allocation and adjustment of computing resources, storage space, and concurrent request limits by tenant. The tenant-level management view is for administrators of each tenant, providing complete autonomous management functions, including organization user management, supporting role assignment and permission configuration; toolset configuration, allowing the enabling or disabling of specific large model context protocol tools; data anonymization rule settings, supporting sensitive information processing strategies based on field level; and cache policy management, customizing cache expiration and memory allocation. The tenant-level management view also provides internal audit log query and analysis functions for tenants.

[0012] To achieve the second objective mentioned above, the technical solution of the product lifecycle management system interaction device based on the large model context protocol provided by the present invention is as follows: The product lifecycle management system interaction device based on the large model context protocol provided by this invention includes: The initialization configuration module is used to load multi-tenant configuration information; based on the multi-tenant configuration information, it establishes a one-to-one exclusive environment parameter for each tenant and enables the application container engine's security sandbox environment to isolate the tool execution process from the host system; it initializes the audit system, establishes a full-link operation log recording mechanism, and completes the initialization configuration of the product lifecycle management system based on the large model context protocol. The product lifecycle management system based on the large model context protocol establishes an independent configuration context for each tenant in the multi-tenant system, and data access and business operations establish independent configuration contexts through the tenant identifiers of each tenant in the multi-tenant system. The raw instruction acquisition module is used to acquire the raw instructions of each tenant using the product lifecycle management system based on the large model context protocol, which has already completed initialization and configuration. The semantic parsing module is used to perform semantic parsing on the original instructions of each tenant using a large language model, and generate standardized tool call sequences; The Natural Language Description Conversion Module is used to convert the standardized tool call sequence into an execution preview of a natural language description based on the Large Model Context Protocol for confirmation by the corresponding tenant. The information output module is used to sequentially execute tool calls in the secure sandbox environment of the application container engine according to the confirmation result of the corresponding tenant, and perform permission adaptation processing on the returned result through the data desensitization engine to obtain output information, wherein the output information meets the confidentiality requirements of the corresponding tenant.

[0013] To achieve the third objective mentioned above, the technical solution of the computer-readable storage medium provided by the present invention is as follows: The computer-readable storage medium provided by the present invention is characterized in that it stores an interactive program for a product lifecycle management system based on the large model context protocol, and when the interactive program for the product lifecycle management system based on the large model context protocol is executed by a processor, it implements the steps of the interactive method for the product lifecycle management system based on the large model context protocol provided by the present invention.

[0014] To achieve the fourth objective mentioned above, the technical solution for the electronic device provided by this invention is as follows: The electronic device provided by the present invention includes a memory and a processor. The memory stores an interactive program for a product lifecycle management system based on the large model context protocol. When the interactive program for a product lifecycle management system based on the large model context protocol is executed by the processor, it implements the steps of the interactive method for a product lifecycle management system based on the large model context protocol provided by the present invention.

[0015] The product lifecycle management system interaction method, apparatus, medium, and device provided by this invention, based on the large model context protocol, receives user natural language commands through an intelligent interaction layer. The large language model parses these commands to generate a large model context protocol tool call sequence and generates an execution preview for user confirmation. Through the large model context protocol intelligent central layer, tool calls are securely routed to the corresponding tenant's backend product lifecycle management system based on the tenant's identity, and a permission mapping engine performs real-time permission verification. A dynamic data anonymization engine processes the results according to tenant policies, and a cache management module improves query performance. A unified management portal enables platform-level and tenant-level management, realizing a smart product lifecycle management system service with natural language interaction, cross-domain knowledge fusion, multi-tenant data isolation, and secure control. This effectively solves the problems of low intelligence levels, data silos, and large model application security in traditional product lifecycle management systems. Attached Figure Description

[0016] Various other advantages and benefits will become apparent to those skilled in the art upon reading the following detailed description of preferred embodiments. The accompanying drawings are for illustrative purposes only and are not intended to limit the invention. Furthermore, the same reference numerals denote the same parts throughout the drawings. In the drawings: Appendix Figure 1 A flowchart illustrating the overall steps of the product lifecycle management system interaction method based on the large model context protocol provided in this embodiment of the invention. Appendix Figure 2 A schematic diagram illustrating the signal flow relationship between functional modules in the interactive device of the product lifecycle management system based on the large model context protocol provided in this embodiment of the invention. Appendix Figure 3 A schematic diagram of the interactive device structure of a product lifecycle management system based on a large model context protocol, which is provided for the hardware operating environment of the embodiments of the present invention. Detailed Implementation

[0017] To address the problems existing in the prior art, this invention provides an interaction method, apparatus, medium, and device for a product lifecycle management system based on a large model context protocol. This method effectively solves the problems of low system intelligence, data silos, and security issues in large model applications in traditional product lifecycle management systems, making it more suitable for practical use.

[0018] Multi-tenancy: Multi-tenancy is a software architecture that allows multiple users or organizations (tenants) to share the same system instance while ensuring data and resource isolation between tenants.

[0019] Through arduous and persistent efforts, the inventors discovered that the main shortcomings of traditional product lifecycle management systems in terms of architecture and functionality are: First, the system's intelligence level is limited, and human-computer interaction efficiency is low. Existing product lifecycle management interfaces are complex with multiple menu levels, requiring extensive training for users to become proficient. Although some manufacturers have introduced artificial intelligence technology to implement intelligent auxiliary functions such as design recommendations and simulation optimization in local modules, significantly improving design reusability and simulation accuracy, these applications are still limited to single-point optimization. They lack digital assistant systems that can understand natural language and support complex cross-module tasks, and the "human-adapted system" interaction mode has not fundamentally changed.

[0020] Secondly, the problem of data silos is severe, and the capabilities for knowledge reuse and decision support are weak. Although the product lifecycle management concept emphasizes the integration of data throughout the entire lifecycle, significant data barriers still exist between systems such as design, production, and operation and maintenance in practical applications. Data is scattered across multiple heterogeneous databases, making it difficult to build a coherent product digital thread. A large amount of tacit knowledge, such as design intent, process experience, and troubleshooting strategies, is buried in unstructured documents or engineers' experience, failing to achieve effective mining and intelligent reuse, thus hindering the development of cross-domain knowledge integration and decision support capabilities.

[0021] Third, security and multi-tenancy requirements bring new challenges. With the promotion of the Industrial Internet and cloud deployment models, product lifecycle management systems face multiple threats such as data breaches, privacy protection, and cyberattacks. Unauthorized access to core enterprise intellectual property, design drawings, and process parameters must be prevented. At the same time, large corporations and software service providers urgently need to provide shared yet isolated product lifecycle management services for different subsidiaries or customers. Existing systems still have significant room for improvement in supporting multi-tenant data isolation and flexible functional customization, especially in achieving secure and controllable intelligent services.

[0022] Furthermore, the integration of large language models faces real-world bottlenecks. Although large language models have made significant progress in natural language understanding and task planning, bringing new opportunities for the intelligent upgrading of product lifecycle management systems, their application in industrial environments still faces several challenges: calling public cloud APIs may lead to the leakage of sensitive data; model illusions and misoperations may cause production accidents; the lack of effective access control and auditing mechanisms makes it difficult to meet enterprise compliance requirements; and the complexity of integrating heterogeneous systems also requires a standardized and scalable protocol framework.

[0023] It is worth noting that the Model Context Protocol (MTP) provides a novel solution to the aforementioned problems. By standardizing the interaction mechanism between large language models and external tools and data sources, this protocol effectively separates model inference from tool execution, aiming to achieve seamless interaction between AI models and external tools and resources, breaking down data silos and promoting interoperability between different systems. Developers can build MTP servers for systems such as CAD, CAE, and CAPP, encapsulating their functionality as standard tools and data as resources, significantly reducing system integration complexity. The MTP natively supports security sandboxes and auditing mechanisms, effectively preventing data leakage and unauthorized operations, laying a technological foundation for the secure, controllable, and efficient integration of large language models in industrial product lifecycle management.

[0024] In summary, the field of industrial product lifecycle management is currently at a critical stage of intelligent upgrading. Although enabling technologies such as cloud computing and artificial intelligence have made some progress, the industry still lacks a comprehensive solution that can systematically achieve natural language interaction, deep knowledge fusion, enterprise-level security, multi-tenant isolation, and open ecosystem integration. Therefore, there is an urgent need for an innovative architecture to achieve deep, secure, and controllable integration of large language models and product lifecycle management systems, systematically addressing security, confidentiality, and multi-tenant requirements, and truly driving the transformation of product lifecycle management from a "data management" to an "intelligent empowerment" paradigm.

[0025] To further illustrate the technical means and effects adopted by the present invention to achieve its intended purpose, the following, in conjunction with the accompanying drawings and preferred embodiments, details the specific implementation methods, structures, features, and effects of the product lifecycle management system interaction method, apparatus, medium, and device based on the large model context protocol proposed by the present invention. In the following description, different "an embodiment" or "an embodiment" do not necessarily refer to the same embodiment. Furthermore, features, structures, or characteristics in one or more embodiments can be combined in any suitable form.

[0026] In this article, the term "and / or" is merely a description of the relationship between related objects, indicating that there can be three relationships, such as A and / or B. Specifically, it can mean that A and B can be included at the same time, A can exist alone, or B can exist alone, and any of the above three situations can be met.

[0027] Interaction Methods for Product Lifecycle Management System Based on Large Model Context Protocol See appendix Figure 1 The product lifecycle management system interaction method based on the large model context protocol provided by this invention includes the following steps: Step S1: Load multi-tenant configuration information; Based on the multi-tenant configuration information, establish exclusive environment parameters for each tenant, and enable the secure sandbox environment of the application container engine to isolate the tool execution process from the host system; Initialize the audit system, establish a full-link operation log recording mechanism, and complete the initial configuration of the product lifecycle management system based on the large model context protocol. The product lifecycle management system based on the large model context protocol establishes an independent configuration context for each tenant in the multi-tenant system. Data access and business operations establish independent configuration contexts through the tenant identifiers of each tenant in the multi-tenant system. Step S2: Obtain the original instructions of each tenant using the product lifecycle management system based on the large model context protocol, which has already been initialized and configured; Step S3: Use a large language model to perform semantic parsing on the original instructions of each tenant and generate standardized tool call sequences; Step S4: The standardized tool call sequence is converted into an execution preview in natural language based on the large model context protocol for confirmation by the corresponding tenant; Step S5: Based on the confirmation result of the corresponding tenant, execute the tool call sequentially in the security sandbox environment of the application container engine, and perform permission adaptation processing on the returned result through the data desensitization engine to obtain the output information, wherein the output information meets the confidentiality requirements of the corresponding tenant.

[0028] The product lifecycle management system interaction method based on the large model context protocol provided by this invention receives user natural language commands through an intelligent interaction layer. The large language model parses these commands to generate a large model context protocol tool call sequence and produces an execution preview for user confirmation. Through the large model context protocol intelligent hub layer, tool calls are securely routed to the corresponding tenant's backend product lifecycle management system based on the tenant's identity, and a permission mapping engine performs real-time permission verification. A dynamic data anonymization engine processes the results according to tenant policies, and a cache management module improves query performance. A unified management portal enables platform-level and tenant-level management, realizing a smart product lifecycle management system service with natural language interaction, cross-domain knowledge fusion, multi-tenant data isolation, and secure control. This effectively solves the problems of low intelligence levels, data silos, and large model application security in traditional product lifecycle management systems.

[0029] The process includes loading multi-tenant configuration information; establishing unique environment parameters for each tenant based on the configuration information and enabling the application container engine's secure sandbox environment to isolate the tool execution process from the host system; initializing the audit system and establishing a full-link operation log recording mechanism; and completing the initial configuration of the product lifecycle management system based on the large model context protocol. The isolation strategies between multi-tenants include the following isolation modes: Physical isolation mode allocates independent database cluster instances to different tenants in a multi-tenant system, thus isolating each tenant from the others. Logical isolation mode, in a shared database environment, uses the tenant's identity field to achieve logical separation of data, thus isolating each tenant in a multi-tenant system from each other; The hybrid isolation mode uses physical isolation for core business data and logical isolation for general business data.

[0030] This process includes loading multi-tenant configuration information; establishing unique environment parameters for each tenant based on the multi-tenant configuration information and enabling the application container engine's secure sandbox environment to isolate the tool execution process from the host system; initializing the audit system and establishing a full-link operation log recording mechanism; and completing the initial configuration steps of the product lifecycle management system based on the large model context protocol. During this process, for file storage of the same tenant, a hierarchical structure for files is established, with the storage path adopting a hierarchical structure of "tenant identifier / project identifier / file identifier".

[0031] This process includes loading multi-tenant configuration information; establishing unique environment parameters for each tenant based on the multi-tenant configuration information and enabling the application container engine's secure sandbox environment to isolate the tool execution process from the host system; initializing the audit system and establishing a full-link operation log recording mechanism; and completing the initial configuration steps of the product lifecycle management system based on the large model context protocol. During this process, each tenant in the multi-tenant system is assigned an independent encryption key. A cross-isolation mode data access middleware is also provided to offer a unified data access interface to upper-layer applications, shielding the complexity of the underlying isolation implementation.

[0032] Among them, the large language model and service components of the product lifecycle management system based on the large model context protocol are all deployed on the unit's intranet; the large model context protocol server runs in a strictly network-isolated application container engine security sandbox; and the interaction of the product lifecycle management system based on the large model context protocol only allows communication between systems on the whitelist.

[0033] Specifically, for time-consuming tasks, the interaction method of the product lifecycle management system based on the large model context protocol includes the following steps: The original instructions of each tenant are obtained by utilizing the product lifecycle management system based on the large model context protocol, which has already been initialized and configured. Based on the original instructions from each tenant, immediately return an acceptance response to the corresponding tenant; Once the time-consuming task is completed, a task completion notification and task result will be sent to the corresponding tenant.

[0034] The product lifecycle management system based on the Large Model Context Protocol (LAMP) comprises a two-tiered management view architecture: a platform-level management view and a tenant-level management view. The platform-level management view provides system administrators with cross-tenant global monitoring and management capabilities, enabling real-time monitoring of the system health status of all tenants, including service availability, API response time, and error rate. It also provides unified resource quota configuration, supporting the allocation and adjustment of computing resources, storage space, and concurrent request limits by tenant. The tenant-level management view, geared towards individual tenant administrators, provides complete autonomous management functions, including organization and user management, supporting role assignment and permission configuration; toolset configuration, allowing the enabling or disabling of specific LAMP tools; data anonymization rule settings, supporting field-level sensitive information processing strategies; and caching policy management, allowing customization of cache expiration and memory allocation. The tenant-level management view also provides internal audit log query and analysis functions for each tenant.

[0035] Interactive device for product lifecycle management system based on large model context protocol See appendix Figure 2 The product lifecycle management system interaction device based on the large model context protocol provided in this embodiment of the invention includes: The initialization configuration module is used to load multi-tenant configuration information; based on the multi-tenant configuration information, it establishes unique environment parameters for each tenant and enables the application container engine's secure sandbox environment, thus isolating the tool execution process from the host system; it initializes the audit system, establishes a full-link operation log recording mechanism, and completes the initialization configuration of the product lifecycle management system based on the large model context protocol. The product lifecycle management system based on the large model context protocol establishes an independent configuration context for each tenant in the multi-tenant system, and data access and business operations establish independent configuration contexts through the tenant identifiers of each tenant in the multi-tenant system. The raw instruction acquisition module is used to acquire raw instructions from each tenant using the product lifecycle management system based on the large model context protocol, which has already been initialized and configured. The semantic parsing module is used to perform semantic parsing on the original instructions of each tenant using a large language model, and generate standardized tool call sequences; The Natural Language Description Conversion Module is used to convert standardized tool call sequences into natural language descriptions based on the Large Model Context Protocol for execution previews to be confirmed by the corresponding tenants. The information output module is used to execute tool calls sequentially in the secure sandbox environment of the application container engine based on the confirmation result of the corresponding tenant, and to perform permission adaptation processing on the returned result through the data desensitization engine to obtain output information. The output information conforms to the confidentiality requirements of the corresponding tenant.

[0036] The product lifecycle management system interaction device based on the large model context protocol provided by this invention receives user natural language commands through an intelligent interaction layer. The large language model parses these commands to generate a large model context protocol tool call sequence and generates an execution preview for user confirmation. Through the large model context protocol intelligent hub layer, tool calls are securely routed to the corresponding tenant's backend product lifecycle management system based on the tenant's identity, and a permission mapping engine performs real-time permission verification. A dynamic data anonymization engine processes the results according to tenant policies, and a cache management module improves query performance. A unified management portal enables platform-level and tenant-level management, realizing a smart product lifecycle management system service with natural language interaction, cross-domain knowledge fusion, multi-tenant data isolation, and secure control. This effectively solves the problems of low intelligence levels, data silos, and large model application security in traditional product lifecycle management systems.

[0037] Computer-readable storage media The present invention provides a computer-readable storage medium storing an interactive program for a product lifecycle management system based on the large model context protocol. When the interactive program for a product lifecycle management system based on the large model context protocol is executed by a processor, it implements the steps of the interactive method for a product lifecycle management system based on the large model context protocol provided by the present invention.

[0038] The computer-readable storage medium provided by this invention receives user natural language commands through an intelligent interaction layer. A large language model parses these commands to generate a large model context protocol tool call sequence and produces an execution preview for user confirmation. Through the large model context protocol intelligent hub layer, tool calls are securely routed to the corresponding tenant's backend product lifecycle management system based on the tenant's identity, and a permission mapping engine performs real-time permission verification. A dynamic data anonymization engine processes the results according to tenant policies, and a cache management module improves query performance. A unified management portal enables platform-level and tenant-level management, realizing a smart product lifecycle management system service with natural language interaction, cross-domain knowledge fusion, multi-tenant data isolation, and secure control. This effectively solves the problems of low intelligence levels, data silos, and security issues in traditional product lifecycle management systems.

[0039] electronic devices The electronic device provided by the present invention includes a memory and a processor. The memory stores an interactive program for a product lifecycle management system based on the large model context protocol. When the interactive program for a product lifecycle management system based on the large model context protocol is executed by the processor, it implements the steps of the interactive method for a product lifecycle management system based on the large model context protocol provided by the present invention.

[0040] The electronic device provided by this invention receives user natural language commands through an intelligent interaction layer. A large language model parses these commands to generate a large model context protocol tool call sequence and produces an execution preview for user confirmation. Through the large model context protocol intelligent hub layer, tool calls are securely routed to the corresponding tenant's backend product lifecycle management system based on the tenant's identity, and a permission mapping engine performs real-time permission verification. A dynamic data anonymization engine processes the results according to tenant policies, and a cache management module improves query performance. A unified management portal enables platform-level and tenant-level management, realizing a smart product lifecycle management system service with natural language interaction, cross-domain knowledge fusion, multi-tenant data isolation, and secure control. This effectively solves the problems of low intelligence levels, data silos, and security issues in traditional product lifecycle management systems.

[0041] Reference Figure 3 , Figure 3 This is a schematic diagram of the interactive device structure of a product lifecycle management system based on a large model context protocol in the hardware operating environment of the embodiment of the present invention.

[0042] like Figure 3 As shown, the interactive device of the product lifecycle management system based on the Big Model Context Protocol may include: a processor 1001, such as a central processing unit (CPU), a communication bus 1002, a user interface 1003, a network interface 1004, and a memory 1005. The communication bus 1002 is used to implement communication between these components. The user interface 1003 may include a display screen or an input unit such as a keyboard; optionally, the user interface 1003 may also include a standard wired interface or a wireless interface. The network interface 1004 may optionally include a standard wired interface or a wireless interface (such as a Wi-Fi interface). The memory 1005 may be a high-speed random access memory (RAM) or a stable non-volatile memory (NVM), such as a disk drive. Optionally, the memory 1005 may also be a storage device independent of the aforementioned processor 1001.

[0043] Those skilled in the art will understand that Figure 3 The structure shown does not constitute a limitation on the interactive device of the product lifecycle management system based on the large model context protocol, and may include more or fewer components than shown, or combine certain components, or have different component arrangements.

[0044] like Figure 3As shown, the memory 1005, which serves as a storage medium, may include an operating system, a data storage module, a network communication module, a user interface module, and an interactive program for a product lifecycle management system based on a large model context protocol.

[0045] exist Figure 3 In the product lifecycle management system interaction device based on the large model context protocol shown, the network interface 1004 is mainly used for data communication with the network server; the user interface 1003 is mainly used for data interaction with the user; the processor 1001 and memory 1005 in the product lifecycle management system interaction device based on the large model context protocol of the present invention can be set in the product lifecycle management system interaction device based on the large model context protocol. The product lifecycle management system interaction device based on the large model context protocol calls the product lifecycle management system interaction program based on the large model context protocol stored in the memory 1005 through the processor 1001, and executes the product lifecycle management system interaction method based on the large model context protocol provided in the embodiment of the present invention.

[0046] Example The construction steps of the product lifecycle management system interaction method based on the large model context protocol provided by this invention are as follows: 1) Construct an intelligent interaction layer The intelligent interaction layer adopts a modern front-end architecture based on Vue and is deeply integrated with the Monaco editor to provide intelligent code completion and syntax highlighting, which significantly improves the user experience and efficiency when performing complex data queries and writing business rules.

[0047] In terms of communication mechanism, the system adopts a two-way communication protocol to establish and maintain a long connection, ensuring that the dialogue and interaction between the user and the intelligent assistant are real-time and continuous, effectively supporting the natural progression of multi-round question and answer and complex task flow.

[0048] In terms of security hardening, the system fully integrates SM series national commercial cryptographic algorithms and encrypted socket protocol layer certificates for end-to-end transmission encryption, ensuring the confidentiality and integrity of data during transmission. Simultaneously, a real-time sensitive word filtering engine is deployed at the front end to instantly scan and clean user input commands, preventing the leakage of sensitive information and the injection of malicious commands from the source.

[0049] For high-risk business operations, the system is designed with an execution preview confirmation mechanism. This mechanism converts the tool call sequence planned by the large language model into a clear and easy-to-read natural language step description, and clearly shows the impact and potential risks of the operation to the user. The actual execution can only be triggered after the user actively confirms, thereby effectively preventing erroneous operations caused by model "illusion" or misjudgment.

[0050] 2) Constructing a context protocol intelligent hub layer The context protocol server, as the core processing unit, encapsulates a standard toolset and resource set for the entire product lifecycle management business. The toolset covers core business operations such as bill of materials management, engineering changes, and process design, including standardized interfaces such as bill of materials query and initiating engineering change notifications. The resource set uniformly manages heterogeneous data sources such as database connection pools, document library access, and CAD file resources, providing a unified resource access abstraction layer for upper layers.

[0051] The permission mapping engine establishes a dynamic permission verification mechanism. It forwards user credentials and metadata carried in tool call requests in real time to the product lifecycle management system of the corresponding user's tenant, invoking its native permission service to complete fine-grained permission verification. This mechanism ensures that all intelligent operations comply with the enterprise's existing role-based access control system, achieving seamless integration of large language model capabilities with the product lifecycle management permission system.

[0052] The tenant routing module is designed based on a multi-tenant architecture. By parsing the tenant identity embedded in the request and combining it with a pre-configured tenant routing table, it accurately routes tool call requests to the dedicated backend resource cluster of that tenant. It supports multiple routing strategies based on domain name, database sharding, and application interface gateways to ensure physical or logical data isolation between different tenants.

[0053] The data masking rule engine implements a dynamic data masking strategy, performing real-time masking processing on the output results based on the current user's security level and the tenant's specific data security policy. The engine supports various masking algorithms, including field replacement, numerical obfuscation, and data truncation, ensuring that sensitive information is only visible to authorized users and effectively preventing data leakage risks.

[0054] The end-to-end audit logger constructs a complete operation audit system, using structured logging technology to record detailed operations from user command reception, large language model parsing, tool invocation to system response. Log information includes key fields such as timestamps, tenant identifiers, user identities, operation types, request parameters, and permission verification results, supporting security incident tracing and compliance review.

[0055] The cache management module implements an intelligent cache optimization mechanism, establishing a multi-level cache system for frequently accessed static data and query results. By creating an independent cache namespace for each tenant, it strictly ensures the isolation of cached data between tenants. The module supports configurable cache invalidation policies and memory management mechanisms, significantly improving system response performance and concurrency processing capabilities.

[0056] 3) Build the business and data layer for product lifecycle management. The system provides a multi-layered data isolation architecture, employing differentiated isolation strategies for different business scenarios and security requirements.

[0057] At the database level, the system supports three isolation modes: physical isolation mode allocates independent database cluster instances to different tenants to ensure absolute security isolation of core business data; logical isolation mode achieves logical separation of data through tenant identity fields in a shared database environment, balancing resource utilization and isolation requirements; and hybrid isolation mode uses physical isolation for highly sensitive core data and logical isolation for ordinary business data to achieve the optimal combination of security and economy.

[0058] At the file storage level, the system establishes a hierarchical storage architecture based on tenant identity. By creating an independent storage namespace for each tenant, physical or logical isolation of files is achieved. The storage path adopts a hierarchical structure of "tenant identifier / project identifier / file identifier," coupled with tenant-based access control, to ensure complete isolation of file access across tenants.

[0059] This architecture also includes a unified key management system, assigning independent encryption keys to each tenant for end-to-end encrypted storage of sensitive data. The system also establishes a cross-isolation mode data access middleware, providing a unified data access interface to upper-layer applications, shielding them from the complexity of the underlying isolation implementation, and ensuring the generality and portability of business code.

[0060] In addition, the system has built a complete multi-tenant and security confidentiality system, which ensures data security and system performance through a three-layer protection mechanism.

[0061] At the tenant isolation level, the system establishes an independent configuration context for each tenant. All data access and business operations are precisely routed through the tenant's identity identifier, achieving complete isolation of data at the physical or logical level and ensuring data invisibility between different tenants.

[0062] In terms of security and confidentiality, the system employs multi-level protection measures. A front-end security filter processes sensitive words and verifies intent in user commands to prevent malicious input. Access control strictly adheres to the principle of "separation of capabilities and permissions," with the large language model only responsible for task planning, while actual permission verification is handled by the original product lifecycle management system. For high-risk operations, the system converts the operation sequence planned by the large language model into a natural language description of the execution preview, which must be explicitly confirmed by the user before execution. During the output phase, a data anonymization engine dynamically hides sensitive information according to preset rules. The entire system adopts a private deployment model, with the large language model and service components all deployed within the enterprise intranet, eliminating the risk of data leaving the country. The context protocol server runs within a strictly network-isolated container security sandbox, allowing communication only with systems on the whitelist.

[0063] In terms of performance optimization, the system implements an asynchronous processing mechanism, immediately returning a response for time-consuming tasks and notifying users upon completion. Simultaneously, an intelligent caching system is established to isolate and cache frequently accessed read-only data according to tenant identity namespaces, significantly improving system response speed and concurrent processing capabilities.

[0064] 4) Build a unified management portal The system provides a two-tier management view architecture, enabling refined operation and maintenance management at both the platform-wide and tenant-internal levels.

[0065] The platform-level management view provides system administrators with cross-tenant global monitoring and management capabilities, enabling real-time monitoring of the system health status of all tenants, including key performance indicators such as service availability, application programming interface response time, and error rate. This view also provides unified resource quota configuration capabilities, supporting the allocation and adjustment of computing resources, storage space, and concurrent request limits by tenant, ensuring the rational allocation and efficient utilization of platform resources.

[0066] The tenant-level management view is for each tenant's administrators, providing complete autonomous management capabilities. This includes organization user management, supporting role assignment and permission configuration; toolset configuration, allowing the enabling or disabling of specific context protocol tools; data anonymization rule settings, supporting field-level sensitive information processing policies; and caching policy management, allowing customization of cache expiration and memory allocation. This view also provides tenant-internal audit log query and analysis capabilities, meeting the independent security audit and compliance requirements of each tenant.

[0067] This two-tier management architecture ensures both centralized control and operation capabilities for the platform and provides each tenant with ample room for independent configuration, thus achieving an organic combination of unified platform management and personalized tenant needs.

[0068] Although preferred embodiments of the invention have been described, those skilled in the art, upon learning the basic inventive concept, can make other changes and modifications to these embodiments. Therefore, the appended claims are intended to be interpreted as including both the preferred embodiments and all changes and modifications falling within the scope of the invention.

[0069] Obviously, those skilled in the art can make various modifications and variations to this invention without departing from its spirit and scope. Therefore, if these modifications and variations fall within the scope of the claims of this invention and their equivalents, this invention also intends to include these modifications and variations.

Claims

1. A large model context protocol based product lifecycle management system interaction method, characterized in that, The method comprises the following steps: loading multi-tenant configuration information; establishing one-to-one exclusive environment parameters for each tenant according to the multi-tenant configuration information, enabling the security sandbox environment of the application container engine; initializing the audit system, establishing a full-link operation log recording mechanism, and completing the initialization configuration of the product life cycle management system based on the large model context protocol, wherein the product life cycle management system based on the large model context protocol establishes an independent configuration context for each tenant in the multi-tenant, and data access and business operation establishes an independent configuration context through the tenant identifier of each tenant in the multi-tenant; using the product life cycle management system based on the large model context protocol which has completed the initialization configuration to obtain the original instructions of each tenant; using a large language model to perform semantic analysis on the original instructions of each tenant to generate a standardized tool calling sequence; the standardized tool calling sequence is converted into a natural language description execution preview based on the large model context protocol for the corresponding tenant to confirm; according to the confirmation result of the corresponding tenant, the tool calling is executed in the security sandbox environment of the application container engine in sequence, and the return result is processed through the data desensitization engine to obtain output information, wherein the output information meets the secret level requirement of the corresponding tenant.

2. The large model context protocol based product lifecycle management system interaction method of claim 1, wherein, loading multi-tenant configuration information; establishing one-to-one exclusive environment parameters for each tenant according to the multi-tenant configuration information, and enabling the security sandbox environment of the application container engine, so that the tool execution process is isolated from the host system; during the step process of initializing the audit system, establishing a full-link operation log recording mechanism, and completing the initialization configuration of the product life cycle management system based on the large model context protocol, the isolation strategy between the multi-tenants includes that the isolation mode includes: a physical isolation mode, in which independent database cluster instances are allocated to different tenants in the multi-tenant, so that each tenant in the multi-tenant is isolated from each other; a logical isolation mode, in which data logical separation is realized by using the identity field of the tenant in a shared database environment, so that each tenant in the multi-tenant is isolated from each other; a hybrid isolation mode, in which physical isolation is adopted for core business data and logical isolation is adopted for ordinary business data.

3. The large model context protocol based product lifecycle management system interaction method of claim 1, wherein, loading multi-tenant configuration information; establishing one-to-one exclusive environment parameters for each tenant according to the multi-tenant configuration information, and enabling the security sandbox environment of the application container engine, so that the tool execution process is isolated from the host system; during the step process of initializing the audit system, establishing a full-link operation log recording mechanism, and completing the initialization configuration of the product life cycle management system based on the large model context protocol, a hierarchical structure of files is established for the file storage of the same tenant, and the storage path adopts a hierarchical structure of "tenant identifier / project identifier / file identifier".

4. The large model context protocol based product lifecycle management system interaction method of claim 1, wherein, Load multi-tenant configuration information; based on the multi-tenant configuration information, establish unique environment parameters for each tenant, and enable the application container engine's secure sandbox environment to isolate the tool execution process from the host system; initialize the audit system, establish a full-link operation log recording mechanism, and complete the initialization configuration steps of the product lifecycle management system based on the large model context protocol. Each tenant in a multi-tenant system is assigned a unique encryption key; It also includes a cross-isolation mode data access middleware, which provides a unified data access interface for upper-layer applications and shields the complexity of the underlying isolation implementation.

5. The product lifecycle management system interaction method based on the large model context protocol according to claim 1, characterized in that, The large language model and service components of the product lifecycle management system based on the large model context protocol are all deployed on the organization's intranet. The large model context protocol server runs within a strictly network-isolated application container engine security sandbox. The product lifecycle management system based on the large model context protocol only allows communication between systems on the whitelist.

6. The large model context protocol based product lifecycle management system interaction method of claim 1, wherein, For time-consuming tasks, the product lifecycle management system interaction method based on the large model context protocol includes the following steps: The original instructions of each tenant are obtained using the product lifecycle management system based on the large model context protocol, which has already completed initial configuration. Based on the original instructions from each tenant, immediately return an acceptance response to the corresponding tenant; Once the time-consuming task is completed, a task completion notification and task result will be sent to the corresponding tenant.

7. The large model context protocol based product lifecycle management system interaction method of claim 1, wherein, The product lifecycle management system based on the large model context protocol includes a two-tier management view architecture, comprising a platform-level management view and a group / user-level management view. The platform-level management view provides system administrators with cross-tenant global monitoring and management capabilities, enabling real-time monitoring of the system health status of all tenants, including service availability, API response time, and error rate. The platform-level management view also provides a unified configuration function for resource quotas, supporting the allocation and adjustment of computing resources, storage space, and concurrent request limits by tenant. The tenant-level management view is for administrators of each tenant, providing complete autonomous management functions, including organization user management, supporting role assignment and permission configuration; toolset configuration, allowing the enabling or disabling of specific large model context protocol tools; data anonymization rule settings, supporting sensitive information processing strategies based on field level; and cache policy management, customizing cache expiration and memory allocation. The tenant-level management view also provides internal audit log query and analysis functions for tenants.

8. A product lifecycle management system interaction apparatus based on large model context protocol, characterized in that, include: The initialization configuration module is used to load multi-tenant configuration information; based on the multi-tenant configuration information, it establishes a one-to-one exclusive environment parameter for each tenant and enables the application container engine's security sandbox environment to isolate the tool execution process from the host system; it initializes the audit system, establishes a full-link operation log recording mechanism, and completes the initialization configuration of the product lifecycle management system based on the large model context protocol. The product lifecycle management system based on the large model context protocol establishes an independent configuration context for each tenant in the multi-tenant system, and data access and business operations establish independent configuration contexts through the tenant identifiers of each tenant in the multi-tenant system. The raw instruction acquisition module is used to acquire the raw instructions of each tenant using the product lifecycle management system based on the large model context protocol, which has already completed initialization and configuration. The semantic parsing module is used to perform semantic parsing on the original instructions of each tenant using a large language model, and generate standardized tool call sequences; The Natural Language Description Conversion Module is used to convert the standardized tool call sequence into an execution preview of a natural language description based on the Large Model Context Protocol for confirmation by the corresponding tenant. The information output module is used to sequentially execute tool calls in the secure sandbox environment of the application container engine according to the confirmation result of the corresponding tenant, and perform permission adaptation processing on the returned result through the data desensitization engine to obtain output information, wherein the output information meets the confidentiality requirements of the corresponding tenant.

9. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores an interactive program for a product lifecycle management system based on the large model context protocol. When executed by a processor, the interactive program implements the steps of the interactive method for a product lifecycle management system based on the large model context protocol as described in any one of claims 1-7.

10. An electronic device, comprising: The system includes a memory and a processor. The memory stores an interactive program for a product lifecycle management system based on the large model context protocol. When the interactive program is executed by the processor, it implements the steps of the interactive method for a product lifecycle management system based on the large model context protocol as described in any one of claims 1-7.

Images