A power grid material supply service system

Abstract

The application provides a power grid material supply service system, which solves the problems of low business processing efficiency and high error rate in the prior art. Specifically, it includes an infrastructure module, a data module, a service module, an application module and a front-end interaction module. The infrastructure module is used to manage the basic facilities required for system operation; the data module is used to store and manage the data required by the system; the service module is used to manage the core services in the system; the application module is used to realize the semantic understanding function of the large language model of the system and optimize the large language model business application; the front-end interaction module is used to realize the information interaction between the system and the user, which improves the business processing efficiency and also improves the user experience.

Description

Technical Field

[0001] This invention relates to the field of supply service technology, and in particular to a power grid material supply service system. Background Technology

[0002] To meet the needs of the power grid company's digital transformation and upgrading, fully implement the strategic deployment of the energy internet, actively implement the overall plan for the application of artificial intelligence technology, and promote the deep integration of various business specialties with artificial intelligence.

[0003] Currently, customer service for suppliers relies on traditional outbound calls and counter services. This primarily includes business inquiries, guidance, payment inquiries, invoice and document searches, and recorded responses for suppliers. It also includes internal staff services such as policy and regulation dissemination, policy and regulation retrieval and learning, and contract document comprehension. However, this approach suffers from issues such as repetitive, mechanical workload, low response efficiency, communication difficulties, and challenges in tracking and performance evaluation. Therefore, it is necessary to improve the overall intelligence level of the business operations.

[0004] While the AI ​​platform has accumulated a wealth of algorithmic model achievements, it mainly focuses on small-scale models for the power industry, image recognition, and speech recognition. In scenarios such as materials companies where consulting and text-based work are the main focus, there is room for improvement in the capabilities of natural language semantic understanding, text generation, and human-like question answering.

[0005] Currently, the power grid material supply service faces the following problems: First, with the expansion of business scale, material management faces increasingly complex challenges. Traditional manual processing and management methods can no longer meet the needs of business development, resulting in problems such as low processing efficiency, high error rate, and difficulty in cost control. Second, with the increase in suppliers and the diversification of needs, it is necessary to provide suppliers with a more convenient and personalized self-service experience, so that suppliers can obtain handover forms, acceptance forms, invoice information, payment status, and other information in a timely manner. Third, there is a need for tools to assist in the interpretation of rules and regulations and internal documents by employees within the material company and even at the local level, thereby improving the efficiency of internal communication. Summary of the Invention

[0006] This invention provides a power grid material supply service system to solve the above-mentioned problems.

[0007] This invention provides a power grid material supply service system, specifically including the following modules:

[0008] The infrastructure module is used to manage the basic facilities required for the system's operation.

[0009] The data module is used to store and manage the data required by the system;

[0010] The service module is used to manage the core services in the system;

[0011] The application module is used to implement the semantic understanding function of the system's large language model and optimize the business applications of the large language model;

[0012] The front-end interaction module is used to enable information exchange between the system and the user.

[0013] The basic infrastructure required for the operation of the system includes servers, network / security equipment, virtualization equipment, storage devices, container cloud, load balancing equipment, etc.

[0014] The core support services include large language model service, vector model service, load balancing service, indicator monitoring service, log collection service, speech recognition and synthesis service, business caching service, authentication and authorization service, distributed configuration service, risk control and rate limiting service, log early warning service, link monitoring service, and other core services.

[0015] Among them, the semantic understanding functions of the large language model include single-turn question answering, key information extraction, online learning, transfer learning, data collection, continuous multi-turn question answering, summary generation, feedback loop, and anthropomorphic question answering.

[0016] Preferably, the infrastructure module specifically includes the following sub-modules:

[0017] The server management submodule is used to manage the operation of the server;

[0018] The network / security device management submodule is used to manage the operation of network / security devices.

[0019] The container cloud management submodule is used to manage the operation of the container cloud;

[0020] The virtualization device management submodule is used to manage the operation of virtualization devices;

[0021] The storage device management submodule is used to manage the operation of storage devices;

[0022] The load balancing device management submodule is used to manage the operation of load balancing devices.

[0023] Preferably, the data module specifically includes the following sub-modules:

[0024] The database management submodule is used to store and manage structured data;

[0025] The caching submodule is used to temporarily store frequently accessed data;

[0026] The vector retrieval submodule is used for fast retrieval and similarity matching of vector data;

[0027] The data security / privacy protection submodule is used to ensure data compliance and user privacy security.

[0028] The full-text search submodule is used for searching text content;

[0029] The System / Business Log submodule is used to record system operation and business operation logs;

[0030] The data backup / recovery submodule is used to ensure data reliability and disaster recovery capabilities;

[0031] The Big Data and Business Intelligence Engine submodule is used for big data analysis and business applications;

[0032] The data quality management submodule is used to monitor and maintain the consistency and accuracy of data.

[0033] Preferably, the service module specifically includes the following sub-modules:

[0034] The large language model service submodule is used to provide natural language processing and semantic understanding services;

[0035] The vector model service submodule is used for vectorized data processing and similarity calculation.

[0036] The load balancing service submodule is used to dynamically allocate service requests and optimize resource usage;

[0037] The metrics monitoring service submodule is used to monitor system performance and service status in real time.

[0038] The log collection service submodule is used for centralized collection and management of log data;

[0039] The speech recognition and synthesis service submodule is used to convert speech to text and vice versa;

[0040] The business caching service submodule is used to cache business data;

[0041] The authentication and authorization service submodule is used to manage user authentication and access control;

[0042] The distributed configuration service submodule is used to uniformly manage the configuration information of the distributed system;

[0043] The risk-limiting service submodule is used to prevent system overload;

[0044] The log alert service submodule is used to trigger an alert mechanism based on log analysis;

[0045] The link monitoring service submodule is used to monitor the status of business links.

[0046] Preferably, the application module specifically includes the following sub-modules:

[0047] The model semantic understanding submodule is used to parse the semantic intent of user input;

[0048] The data collection submodule is used to collect business data;

[0049] The Model Business Application Submodule is used to embed large language models into business processes;

[0050] The key information extraction submodule is used to extract key fields from unstructured text;

[0051] The summary generation submodule is used to generate summaries from text.

[0052] The vendor information management submodule is used to maintain basic supplier information, qualifications, and cooperation records;

[0053] The contract search submodule is used to quickly find and locate contract documents;

[0054] The contract extraction submodule is used to extract structured data from contracts;

[0055] The contract classification submodule is used to classify contracts.

[0056] The online learning submodule is used to update and iterate the large language model online based on new data generated during system operation;

[0057] The feedback loop submodule is used to optimize the large language model based on the collected user satisfaction with the large language model's answers;

[0058] The transfer learning submodule is used to reuse pre-trained models to solve new business scenario tasks, reducing training costs.

[0059] The anthropomorphic question-and-answer submodule is used to provide a more natural dialogue experience by adjusting the answer style and interaction logic;

[0060] The statistics report submodule is used to generate business data visualization reports;

[0061] The human Q&A submodule is used to transfer complex questions to human customer service when the AI ​​assistant cannot handle them;

[0062] The single-round question-and-answer submodule is used to handle one-time questions from users;

[0063] A multi-turn question-and-answer submodule is used to support context-dependent continuous dialogue;

[0064] The transaction bill query submodule is used to query transaction records and bill details by criteria such as time and type.

[0065] The Auto-Answer submodule is used for automatically answering questions.

[0066] The business guidance submodule is used to provide operational guidance for business steps;

[0067] The proactive information notification submodule is used to push important reminders;

[0068] The document traceability submodule is used to record the document modification history.

[0069] Preferably, the front-end interaction module includes an APP, a PC application, and a web page; wherein,

[0070] The app provides a basic interactive interface that allows users to check bills, contracts, and contact customer service.

[0071] The web page provides a fully functional web-based interactive interface;

[0072] The PC application provides a fully functional interactive interface.

[0073] Compared with the prior art, the present invention has the following obvious and prominent substantive features and significant advantages:

[0074] This invention provides a power grid material supply service system that solves the problems of low business processing efficiency and high error rate in existing technologies. It has the following advantages:

[0075] 1. Highly modular architecture

[0076] Each module is designed independently, supporting on-demand expansion or upgrades, avoiding a cascading effect where a change in one part affects the whole system. Modules interact through standardized interfaces, making fault location and repair more efficient and reducing overall system maintenance costs.

[0077] 2. Data lifecycle management capabilities

[0078] Through technologies such as data encryption, privacy protection, and access control, we ensure the compliance and leak prevention capabilities of sensitive information. Mechanisms including data backup / recovery, caching systems, and load balancing guarantee business continuity and prevent data loss or service interruption due to single points of failure. Combined with big data and business intelligence engines and full-text search systems, we support rapid analysis and decision support for massive amounts of data.

[0079] 3. Deep integration of AI technology and business

[0080] The AI-powered large language model service supports natural language interaction, lowering the barrier to entry for users and improving service efficiency. Functions such as key information extraction, contract extraction and classification, and summary generation replace manual processing of repetitive tasks, reducing error rates and increasing speed by over 50%. Through feedback loops, online learning, and transfer learning, the system can dynamically optimize the model based on user behavior and business changes, adapting to new scenario requirements.

[0081] 4. Multi-level stability and risk control

[0082] Link monitoring services, metric monitoring, and log alerts track system health status in real time, quickly pinpointing bottlenecks and triggering circuit breakers and rate limiting to ensure service continuity. Container cloud, virtualization devices, and load balancing technologies work together to dynamically allocate computing and storage resources to handle traffic peaks.

[0083] 5. User experience and interaction diversity

[0084] It supports multiple entry points, including mobile apps, PC websites, and live agent services, catering to users' operational habits in different scenarios. Features such as single / multi-round Q&A, business guidance, and proactive notifications provide a natural and seamless service experience. Document tracking and statistical reports enhance user trust in the business process.

[0085] 6. Dual optimization of cost and efficiency

[0086] Virtualization devices, caching systems, and other technologies reduce hardware redundancy and lower IT infrastructure investment costs. AI models replace manual processing of high-frequency tasks, freeing up human resources to focus on high-value work.

[0087] Through modular architecture, deep AI integration, end-to-end stability assurance, and user experience optimization, the system achieves efficient business processing, intelligent operation and maintenance management, and intensive resource utilization, providing safe, reliable, and agile digital service support for power grid material supply services. Attached Figure Description

[0088] The accompanying drawings, which constitute a part of this invention, are used to provide a further understanding of the invention. The illustrative embodiments of the invention and their descriptions are used to explain the invention and do not constitute an undue limitation of the invention. In the drawings:

[0089] Figure 1 This is a structural diagram of a power grid material supply service system according to a preferred embodiment of the present invention.

[0090] Figure 2 This is a structural diagram of the infrastructure modules of a power grid material supply service system according to a preferred embodiment of the present invention.

[0091] Figure 3 This is a data module structure diagram of a power grid material supply service system according to a preferred embodiment of the present invention.

[0092] Figure 4 This is a service module structure diagram of a power grid material supply service system according to a preferred embodiment of the present invention.

[0093] Figure 5 This is a structural diagram of an application module of a power grid material supply service system according to a preferred embodiment of the present invention.

[0094] Figure 6 This is a structural diagram of the front-end interactive module of a power grid material supply service system according to a preferred embodiment of the present invention.

[0095] Legend:

[0096] 100. Infrastructure Module; 200. Data Module; 300. Service Module; 400. Application Module; 500. Front-end Interaction Module;

[0097] 101. Server Management Submodule; 102. Network / Security Device Management Submodule; 103. Container Cloud Management Submodule; 104. Virtualization Device Management Submodule; 105. Storage Device Management Submodule; 106. Load Balancing Device Management Submodule;

[0098] 201. Database Management Submodule; 202. Caching Submodule; 203. Vector Retrieval Submodule; 204. Data Security / Privacy Protection Submodule; 205. Full-Text Search Submodule; 206. System / Business Log Submodule; 207. Data Backup / Recovery Submodule; 208. Big Data and Business Intelligence Engine Submodule; 209. Data Quality Management Submodule;

[0099] 301. Large Language Model Service Submodule; 302. Vector Model Service Submodule; 303. Load Balancing Service Submodule; 304. Metric Monitoring Service Submodule; 305. Log Collection Service Submodule; 306. Speech Recognition and Synthesis Service Submodule; 307. Business Caching Service Submodule; 308. Authentication and Authorization Service Submodule; 309. Distributed Configuration Service Submodule; 310. Risk Rate Limiting Service Submodule; 311. Log Early Warning Service Submodule; 312. Link Monitoring Service Submodule;

[0100] 401. Model Semantic Understanding Submodule; 402. Data Collection Submodule; 403. Model Business Application Submodule; 404. Key Information Extraction Submodule; 405. Summary Generation Submodule; 406. Vendor Information Management Submodule; 407. Contract Search Submodule; 408. Contract Extraction Submodule; 409. Contract Classification Submodule; 410. Online Learning Submodule; 411. Feedback Loop Submodule; 412. Transfer Learning Submodule; 413. Human-like Question and Answer Submodule; 414. Statistical Report Submodule; 415. Human Question and Answer Submodule; 416. Single-Turn Question and Answer Submodule; 417. Multi-Turn Question and Answer Submodule; 418. Transaction Bill Inquiry Submodule; 419. Automatic Reply Submodule; 420. Business Guidance Submodule; 421. Proactive Information Notification Submodule; 422. Document Traceability Submodule. Detailed Implementation

[0101] This invention provides a power grid material supply service system. To make the objectives, technical solutions, and effects of this invention clearer and more explicit, the invention will be further described in detail below with reference to the accompanying drawings and examples. It should be understood that the specific embodiments described herein are only for explaining the invention and are not intended to limit the invention.

[0102] It should be noted that the terms "first," "second," etc., in the specification, claims, and accompanying drawings of this invention are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be used interchangeably where appropriate. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion. For example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to these processes, methods, products, or apparatuses.

[0103] Example:

[0104] like Figures 1-6 As shown in the figure, the power grid material supply service system described in this embodiment specifically includes an infrastructure module 100, a data module 200, a service module 300, an application module 400, and a front-end interaction module 500.

[0105] 1. Infrastructure module 100, used to manage the basic facilities required for system operation; wherein, the basic facilities required for system operation include infrastructure such as servers, network / security equipment, virtualization equipment, storage equipment, container cloud, and load balancing equipment.

[0106] The infrastructure module 100 specifically includes a server management submodule 101, a network / security device management submodule 102, a container cloud management submodule 103, a virtualization device management submodule 104, a storage device management submodule 105, and a load balancing device management submodule 106.

[0107] Server management submodule 101 is used to manage the operation of the server, provide computing resources, and support system operation.

[0108] The network / security device management submodule 102 is used to manage the operation of network / security devices and ensure network communication and data security, including devices such as firewalls and routers.

[0109] The container cloud management submodule 103 is used to manage the operation of the container cloud, provide a containerized deployment environment, and support the elastic scaling and rapid iteration of services.

[0110] The virtualization device management submodule 104 is used to manage the operation of virtualization devices, improve the utilization of physical resources through virtualization technology, and achieve flexible resource allocation.

[0111] The storage device management submodule 105 is used to manage the operation of storage devices, providing high-capacity, high-performance data storage capabilities and supporting both structured and unstructured data.

[0112] The load balancing device management submodule 106 is used to manage the operation of the load balancing device, dynamically allocate traffic, and ensure high system availability and service stability.

[0113] 2. Data module 200, used to store and manage the data required by the system.

[0114] The data module 200 specifically includes a database management submodule 201, a caching submodule 202, a vector retrieval submodule 203, a data security / privacy protection submodule 204, a full-text retrieval submodule 205, a system / business log submodule 206, a data backup / recovery submodule 207, a big data and business intelligence engine submodule 208, and a data quality management submodule 209.

[0115] The database management submodule 201 is used to store and manage structured data, enabling the storage, querying, and management of structured data, such as MySQL and Oracle.

[0116] The caching submodule 202 is used to temporarily store frequently accessed data (such as Redis), thereby reducing database load and improving response speed.

[0117] The vector retrieval submodule 203 is used for fast retrieval and similarity matching of vector data, supporting efficient retrieval and similarity matching of vectorized data (such as AI model features).

[0118] The data security / privacy protection submodule 204 is used to ensure data compliance and user privacy security through encryption, de-identification and other technologies;

[0119] Full-text search submodule 205 is used for text content retrieval, and it enables efficient search and analysis of text content based on tools such as Elasticsearch.

[0120] The System / Business Log submodule 206 is used to record system operation and business operation logs. By recording the system operation status and business operation logs, it is convenient for subsequent system fault troubleshooting and log security auditing.

[0121] The data backup / recovery submodule 207 is used to back up data regularly and support rapid recovery, ensuring data reliability and disaster recovery capabilities.

[0122] The Big Data and Business Intelligence Engine submodule 208 is used for big data analysis and business applications, supporting massive data analysis and visualization, and driving business decisions (such as Hadoop and Tableau).

[0123] The data quality management submodule 209 is used to monitor and maintain the consistency and accuracy of data, and to clean up redundant or erroneous data.

[0124] 3. Service module 300 is used to manage the core services in the system; wherein, the core support services include large language model service, vector model service, load balancing service, indicator monitoring service, log collection service, speech recognition and synthesis service, business caching service, authentication and authorization service, distributed configuration service, risk control and rate limiting service, log early warning service, link monitoring service and other core services.

[0125] Specifically, the service module 300 includes a large language model service submodule 301, a vector model service submodule 302, a load balancing service submodule 303, an indicator monitoring service submodule 304, a log collection service submodule 305, a speech recognition and synthesis service submodule 306, a business caching service submodule 307, an authentication and authorization service submodule 308, a distributed configuration service submodule 309, a risk rate limiting service submodule 310, a log early warning service submodule 311, and a link monitoring service submodule 312.

[0126] The Large Language Model Service Submodule 301 is used to provide Natural Language Processing (NLP) and Semantic Understanding services, such as intent recognition and text generation.

[0127] The vector model service submodule 302 is used for vectorized data processing and similarity calculation, converting text, images and other data into vectors, and supporting similarity calculation and recommendation.

[0128] The load balancing service submodule 303 is used to dynamically allocate service requests, avoid single point of overload, and optimize resource utilization.

[0129] The metrics monitoring service submodule 304 is used to monitor system performance and service status in real time, including key performance metrics such as CPU, memory, and request latency.

[0130] The log collection service submodule 305 is used to centrally collect and manage log data, and supports unified analysis and storage (such as the ELK stack).

[0131] The speech recognition and synthesis service submodule 306 is used to implement speech-to-text (ASR) and text-to-speech (TTS) functions.

[0132] The business caching service submodule 307 is used to cache business data, reducing the access pressure on the database.

[0133] The authentication and authorization service submodule 308 is used to manage user authentication, permission allocation, and access control.

[0134] The distributed configuration service submodule 309 is used to uniformly manage the configuration information of the distributed system and supports dynamic updates.

[0135] The risk-limiting service submodule 310 is used to prevent system overload by using circuit breaking mechanisms and rate limiting strategies to avoid system overload.

[0136] The log alert service submodule 311 is used to trigger alerts based on log anomaly detection, such as a sudden increase in error rate or server crash.

[0137] Link monitoring service submodule 312 is used to monitor the status of business links.

[0138] 4. Application module 400 is used to realize the semantic understanding function of the system's large language model and optimize the business application of the large language model. Among them, the semantic understanding function of the large language model includes functions such as single-turn question answering, key information extraction, online learning, transfer learning, data collection, continuous multi-turn question answering, summary generation, feedback loop and anthropomorphic question answering.

[0139] The application module 400 includes a model semantic understanding submodule 401, a data collection submodule 402, a model business application submodule 403, a key information extraction submodule 404, a summary generation submodule 405, a vendor information management submodule 406, a contract search submodule 407, a contract extraction submodule 408, a contract classification submodule 409, an online learning submodule 410, a feedback loop submodule 411, a transfer learning submodule 412, a human-like question-and-answer submodule 413, a statistical report submodule 414, a human question-and-answer submodule 415, a single-turn question-and-answer submodule 416, a multi-turn question-and-answer submodule 417, a transaction bill query submodule 418, an automatic reply submodule 419, a business guidance submodule 420, a proactive information notification submodule 421, and a document traceability submodule 422. The model semantic understanding submodule 401 is used to parse the semantic intent of user input, such as mapping "query bill" to a specific business window.

[0140] The data collection submodule 402 is used to collect business data (such as user behavior logs) and for the training and optimization of large language models.

[0141] Model Business Application Submodule 403 is used to embed large language models into business processes, such as automatically generating contract summaries.

[0142] The key information extraction submodule 404 is used to extract key fields from unstructured text, such as contracts, and to extract key fields such as amount and date from contracts.

[0143] The summary generation submodule 405 is used to generate summaries from text, improving the efficiency of information retrieval.

[0144] The vendor information management submodule 406 is used to maintain basic supplier information, qualifications, and cooperation records.

[0145] Contract search submodule 407 is used to quickly find and locate contract documents.

[0146] The contract extraction submodule 408 is used to extract structured data from contracts, such as contract terms and signatories.

[0147] The contract classification submodule 409 is used to classify contracts. It allows for customized automatic classification of contracts based on dimensions such as type and status.

[0148] The online learning submodule 410 is used to update and iterate the large language model online based on new data generated during system operation, without retraining the entire large language model.

[0149] Feedback loop submodule 411 is used to optimize the large language model based on the collected user satisfaction with the large language model's answers.

[0150] The transfer learning submodule 412 is used to reuse pre-trained models to solve new business scenario tasks and reduce training costs.

[0151] The anthropomorphic question-and-answer submodule 413 is used to provide a more natural dialogue experience by adjusting the answer style and interaction logic.

[0152] The statistical report submodule 414 is used to generate business data visualization reports, such as monthly transaction trends and supplier ratings.

[0153] The human Q&A submodule 415 is used to transfer the call to a human customer service representative when the AI ​​assistant cannot handle complex questions.

[0154] The single-round question-and-answer submodule 416 is used to handle one-time user questions, such as "What is the current electricity balance?"

[0155] Multi-turn question-and-answer submodule 417 is used to support context-dependent continuous dialogue, such as "There is a problem with last month's bill, please check it for me".

[0156] The transaction bill query submodule 418 is used to query transaction records and bill details by time, type, and other conditions.

[0157] The Auto-Answer submodule 419 is used for automatic answers to questions, usually for common questions, such as "How do I change my password?".

[0158] The Business Guidance Submodule 420 is used to provide operational guidance for business steps, such as "How to apply for new supplier onboarding".

[0159] The proactive information notification submodule 421 is used to push important reminders, such as contract expiration and billing expectations.

[0160] The document traceability submodule 422 is used to record the document modification history, and also supports version comparison and operation traceability.

[0161] 5. Front-end interaction module 500, used to realize information interaction between the system and the user.

[0162] The front-end interaction module 500 includes an APP, a PC application, and a web page. The APP provides a basic interactive interface, supporting user operations such as querying bills, contracts, and contacting customer service; the web page provides a fully functional web-based interactive interface; and the PC application provides a fully functional interactive interface.

[0163] The specific embodiments of the present invention have been described in detail above, but they are merely examples, and the present invention is not limited to the specific embodiments described above. For those skilled in the art, any equivalent modifications and substitutions to the present invention are also within the scope of the present invention. Therefore, all equivalent transformations and modifications made without departing from the spirit and scope of the present invention should be covered within the scope of the present invention.

Claims

1. A power grid material supply service system, characterized in that, Specifically, it includes the following modules: The infrastructure module is used to manage the basic facilities required for the system's operation. The data module is used to store and manage the data required by the system; The service module is used to manage the core services in the system; The application module is used to implement the semantic understanding function of the system's large language model and optimize the business applications of the large language model; The front-end interaction module is used to enable information exchange between the system and the user.

2. The power grid material supply service system according to claim 1, characterized in that, The basic infrastructure required for the system to operate includes servers, network / security equipment, virtualization equipment, storage equipment, container cloud, and load balancing equipment.

3. The power grid material supply service system according to claim 1, characterized in that, The core support services include large language model service, vector model service, load balancing service, indicator monitoring service, log collection service, speech recognition and synthesis service, business caching service, authentication and authorization service, distributed configuration service, risk control and rate limiting service, log early warning service, and link monitoring service.

4. The power grid material supply service system according to claim 1, characterized in that, The semantic understanding functions of the large language model include single-turn question answering, key information extraction, online learning, transfer learning, data collection, continuous multi-turn question answering, summary generation, feedback loop, and anthropomorphic question answering.

5. A power grid material supply service system according to claim 1, characterized in that, The infrastructure module specifically includes the following sub-modules: The server management submodule is used to manage the operation of the server; The network / security device management submodule is used to manage the operation of network / security devices. The container cloud management submodule is used to manage the operation of the container cloud; The virtualization device management submodule is used to manage the operation of virtualization devices; The storage device management submodule is used to manage the operation of storage devices; The load balancing device management submodule is used to manage the operation of load balancing devices.

6. The power grid material supply service system according to claim 1, characterized in that, The data module specifically includes the following sub-modules: The database management submodule is used to store and manage structured data; The caching submodule is used to temporarily store frequently accessed data; The vector retrieval submodule is used for fast retrieval and similarity matching of vector data; The data security / privacy protection submodule is used to ensure data compliance and user privacy security. The full-text search submodule is used for searching text content; The System / Business Log submodule is used to record system operation and business operation logs; The data backup / recovery submodule is used to ensure data reliability and disaster recovery capabilities; The Big Data and Business Intelligence Engine submodule is used for big data analysis and business applications; The data quality management submodule is used to monitor and maintain the consistency and accuracy of data.

7. A power grid material supply service system according to claim 1, characterized in that, The service module specifically includes the following sub-modules: The large language model service submodule is used to provide natural language processing and semantic understanding services; The vector model service submodule is used for vectorized data processing and similarity calculation. The load balancing service submodule is used to dynamically allocate service requests and optimize resource usage; The metrics monitoring service submodule is used to monitor system performance and service status in real time. The log collection service submodule is used for centralized collection and management of log data; The speech recognition and synthesis service submodule is used to convert speech to text and vice versa; The business caching service submodule is used to cache business data; The authentication and authorization service submodule is used to manage user authentication and access control; The distributed configuration service submodule is used to uniformly manage the configuration information of the distributed system; The risk-limiting service submodule is used to prevent system overload; The log alert service submodule is used to trigger an alert mechanism based on log analysis; The link monitoring service submodule is used to monitor the status of business links.

8. A power grid material supply service system according to claim 1, characterized in that, The application module specifically includes the following sub-modules: The model semantic understanding submodule is used to parse the semantic intent of user input; The data collection submodule is used to collect business data; The Model Business Application Submodule is used to embed large language models into business processes; The key information extraction submodule is used to extract key fields from unstructured text; The summary generation submodule is used to generate summaries from text. The vendor information management submodule is used to maintain basic supplier information, qualifications, and cooperation records; The contract search submodule is used to quickly find and locate contract documents; The contract extraction submodule is used to extract structured data from contracts; The contract classification submodule is used to classify contracts. The online learning submodule is used to update and iterate the large language model online based on new data generated during system operation; The feedback loop submodule is used to optimize the large language model based on the collected user satisfaction with the large language model's answers; The transfer learning submodule is used to reuse pre-trained models to solve new business scenario tasks, reducing training costs. The anthropomorphic question-and-answer submodule is used to provide a more natural dialogue experience by adjusting the answer style and interaction logic; The statistics report submodule is used to generate business data visualization reports; The human Q&A submodule is used to transfer complex questions to human customer service when the AI ​​assistant cannot handle them; The single-round question-and-answer submodule is used to handle one-time questions from users; A multi-turn question-and-answer submodule is used to support context-dependent continuous dialogue; The transaction bill query submodule is used to query transaction records and bill details by time and type. The Auto-Answer submodule is used for automatically answering questions. The business guidance submodule is used to provide operational guidance for business steps; The proactive information notification submodule is used to push important reminders; The document traceability submodule is used to record the document modification history.

9. A power grid material supply service system according to claim 1, characterized in that, The front-end interaction module includes an APP, a PC application, and a web page; among them... The app provides a basic interactive interface that allows users to check bills, contracts, and contact customer service. The web page provides a fully functional web-based interactive interface; The PC application provides a fully functional interactive interface.

Images