Power field context fusion method and device, electronic equipment and storage medium

By mapping power business information to a vector space and introducing a context fusion method that incorporates time validity and business priority constraints, the problems of heterogeneity and insufficient timeliness of power business data are solved, enabling real-time and accurate decision-making in digital human systems.

CN122174141APending Publication Date: 2026-06-09BEIJING CHINA POWER INFORMATION TECH
View PDF 0 Cites 0 Cited by

Patent Information

Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
BEIJING CHINA POWER INFORMATION TECH
Filing Date
2026-01-20
Publication Date
2026-06-09

AI Technical Summary

Technical Problem

The existing power business data comes from diverse sources and has significantly different update frequencies. The lack of a unified context association mechanism makes it impossible for digital humans to perceive the grid operation status, changes in customer electricity consumption behavior, and marketing constraints in real time during the conversation, resulting in problems such as delayed response and inaccurate decision-making.

Method used

Power business information is mapped to a vector space, and context fusion is performed through an attention algorithm. Time validity and business priority constraints are introduced, and the score data of the fused features is calculated to determine the currently valid context.

Benefits of technology

It enables the formation of a unified, stable, and computable context state in multi-source, multi-temporal, and multi-business conflict scenarios, ensuring the real-time performance and accuracy of digital human interaction and decision-making.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN122174141A_ABST
    Figure CN122174141A_ABST
Patent Text Reader

Abstract

The application provides a power field context fusion method, comprising: mapping preset power service information to a vector space to obtain power vector information. The power service information comprises power service data from at least two data sources, and the power vector information comprises power vector data corresponding to the at least two data sources. Context fusion is performed on the at least two power vector data according to a preset attention algorithm to obtain fusion features. The fusion features are scored to obtain feature score data. The current effective context is determined according to the feature score data. Thus, the accurate fusion of the power field context is realized.
Need to check novelty before this filing date? Find Prior Art