Business process processing method and system based on large language model

By building parameter optimization of multi-relationship networks and large language models, the problem that traditional methods are difficult to capture the complex relationships of power grid services is solved, and the intelligent and automated processing of power grid business processes is realized.

CN119003704AActive Publication Date: 2024-11-22CHENGDU POWER SUPPLY COMPANY OF STATE GRID SICHUAN ELECTRIC POWER

Patent Information

Application Number
CN202411440853.7
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-10-16
Publication Date
2024-11-22
Estimated Expiration
2044-10-16

AI Technical Summary

Technical Problem

Traditional manual processing and data processing methods are difficult to fully capture the complex relationships, diversity, heterogeneity, and dynamics in power grid business processes, resulting in inaccurate and comprehensive processing results.

Method used

Using a business process processing method based on a large language model, a multi-relationship network of power grid business description content is constructed, session events and dependencies are extracted, embedded representations and vector fusion is performed, sample network member vectors are generated, and parameters are made to initialize the large language model. Optimize and learn to generate target large language models.

Benefits of technology

It significantly improves the intelligence and automation level of power grid business processes, realizes accurate and efficient business process processing, and can better adapt to the dynamic changes of power grid business.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN119003704A_ABST
    Figure CN119003704A_ABST
Patent Text Reader

Abstract

The invention provides a business process processing method and system based on a large language model, relates to the technical field of artificial intelligence, and can accurately capture a power grid business session event and a dependency relationship thereof by constructing a multivariate relationship network of power grid business description content, thereby providing a comprehensive data basis for analysis of complex power grid business. Furthermore, the business event, the linkage event and the session knowledge tag are converted into vector representation of a high-dimensional space through an embedded representation and vector fusion technology, so that multi-dimensional information is effectively integrated, and the understanding ability of the model to a power grid business scene is enhanced. According to the method, the initial large language model is subjected to linkage parameter adaptation optimization and parameter learning by using the sample network member vector, and the finally generated target large language model is not only high in accuracy, but also can better adapt to the dynamic change of the power grid business, so that the precision, high efficiency and intelligentization of the business process processing are realized.
Need to check novelty before this filing date? Find Prior Art

Citation Information

Patent Citations

  • Feasibility study evaluation system integrating grid multi-source information

    CN108345991A

  • Risk conduction prediction method, device, equipment and medium

    CN115034596A

  • Generation method, system and equipment of power grid fault handling aid decision and medium

    CN117709461A

  • Data mining method and system applying text informatization system

    CN118037250A

  • Network security feedback analysis method and system based on artificial intelligence

    CN118174962A

Cited By

  • Digital electric power analysis method and device cooperating with multiple production elements

    CN119228322A

  • A digital power analysis method and device for coordinating multiple production factors

    CN119228322B