Method for monitoring operation management based on artificial intelligence platform

By determining optimization goals based on business scenarios on the artificial intelligence platform and building a single model, the problem of low recognition accuracy of artificial intelligence models caused by the complexity of power grid business requirements is solved, and the recognition accuracy and ease of use of the model are improved.

CN119783904AActive Publication Date: 2025-04-08STATE GRID HENAN INFORMATION & TELECOMM CO

Patent Information

Application Number
CN202411979998.4
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-12-31
Publication Date
2025-04-08
Estimated Expiration
2044-12-31

AI Technical Summary

Technical Problem

Due to the complexity of power grid business requirements, the lack of statistics and analysis of different artificial intelligence models in the artificial intelligence platform makes it difficult for users to effectively and quickly obtain and understand the overall situation of artificial intelligence, and thus it is difficult to identify artificial intelligence models with low recognition accuracy or long processing time, affecting the ease of use and accuracy of the model.

Method used

By determining the general model and matching business scenarios based on different business scenarios on the artificial intelligence platform, obtaining the call data and identifying changes in accuracy, determining the model optimization goals, and using the similar situations and data volume of training data to screen the optimization goals, and building a single artificial intelligence model.

Benefits of technology

The screening of matching business scenarios with low recognition accuracy is realized, the recognition accuracy of artificial intelligence models in matching business scenarios is improved, and the recognition accuracy of multiple matching business scenarios is avoided due to the general model is not satisfactory, and the impact on the recognition accuracy of other matching business scenarios is reduced.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN119783904A_ABST
    Figure CN119783904A_ABST
Patent Text Reader

Abstract

The invention provides a method for monitoring operation management based on an artificial intelligence platform, and belongs to the technical field of operation management, and the method specifically comprises the steps: determining a model optimization target in a general model based on a change condition and call data, determining a similar condition of training data between different matching business scenes in the model optimization target, and carrying out the simulation of the training data. Determining a training data similarity coefficient of the matched service scene and other matched service scenes and an optimized service scene by utilizing the similar condition, and acquiring the change condition of the training data in different divided time periods in different optimized service scenes; and in combination with the training data similarity coefficient of the optimization service scene, the optimization target in the optimization service scene is determined, and a single artificial intelligence model is established for the optimization target, so that the recognition accuracy of the artificial intelligence model is improved.
Need to check novelty before this filing date? Find Prior Art

Citation Information

Patent Citations

  • An information processing method, system and cloud platform based on artificial intelligence

    CN114611478B

  • Artificial intelligence service system and method based on cloud platform

    CN115082082A

  • Model management method and system based on business requirements

    CN117764194A

  • Intelligent traffic management optimization method and system based on target detection

    CN118153780A

  • Model optimization method and device, computer equipment and storage medium

    CN118626329A

Cited By

  • Property information management method and system based on artificial intelligence

    CN121032135A